Notung-2.

9: A Manual

Roseanna Alderson, Dave Danicic, Dannie Durand,

Aiton Goldman, Han Lai, Annette McLeod,
Maureen Stolzer, Benjamin Vernot, and Minli Xu

Date: March 3, 2020

Contents

1 Introduction to Notung 6
1.1 How to cite Notung . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2 Using This Manual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2 Downloading Notung 10

3 Getting Started 12
3.1 Gene and Species Trees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.2 The Graphical User Interface . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.3 File Menu and Opening and Saving Trees . . . . . . . . . . . . . . . . . . . . 15
3.4 General Tree Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.5 Parameter Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

4 Theory 25
4.1 Transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.1.1 Multiple Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.2 Non-Binary Trees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.2.1 Fitting Binary Gene Trees to Non-Binary Species Trees . . . . . . . . 30
4.2.2 Fitting a Non-Binary Gene Tree to a Binary Species Tree . . . . . . . 35
4.3 Practical considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

5 Reconciliation Mode 40
5.1 Labeling and Pruning Gene and Species Trees . . . . . . . . . . . . . . . . . 41
5.2 Reconciling Binary Trees with Transfers . . . . . . . . . . . . . . . . . . . . 42
5.2.1 Temporal Feasibility . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.2.2 Multiple Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.2.3 Rearrange and Resolve with Transfers . . . . . . . . . . . . . . . . . 44
5.3 Reconciling Non-Binary Trees . . . . . . . . . . . . . . . . . . . . . . . . . . 44
5.4 Using Reconciliation Commands . . . . . . . . . . . . . . . . . . . . . . . . . 45
5.5 Additional Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.5.1 Event Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.6 Inferring Orthologs, Paralogs, and Xenologs . . . . . . . . . . . . . . . . . . 53
5.6.1 Homology Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . 54

2
 5.6.2 Interactive Homology Mode . . . . . . . . . . . . . . . . . . . . . . . 57
5.6.3 Homology Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

6 Rooting Mode 68
6.0.1 Rooting with Transfers . . . . . . . . . . . . . . . . . . . . . . . . . . 71

7 Rearrange Mode 73
7.1 Alternate Optimal Hypotheses . . . . . . . . . . . . . . . . . . . . . . . . . . 74
7.2 Rearrangement Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78

8 Resolve Mode 81

9 History 85

10 Annotations 87

11 Changing the Appearance of the Tree Panel 93

11.1 Display Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
11.2 Zoom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
11.3 Changing Font Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

12 Command Line Options and Batch Processing 98

12.1 Opening and Using a Command Window/Terminal . . . . . . . . . . . . . . 99
12.2 Running Notung from the command line . . . . . . . . . . . . . . . . . . . . 102
12.3 Running Notung from a Batch File . . . . . . . . . . . . . . . . . . . . . . . 105
12.4 Running a phylogenomic analysis . . . . . . . . . . . . . . . . . . . . . . . . 107
12.5 Saving PNG Images of Trees . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
12.6 Inferring Losses when Reconciling with Non-Binary Species Trees . . . . . . 114
12.7 Rooting trees from the command line with the DTL model . . . . . . . . . . 115
12.8 Analyzing non-binary gene trees with the DTL model . . . . . . . . . . . . . 117
12.9 Command line options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

A File Formats 126

A.1 Newick File Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
A.2 NHX File Format - New Hampshire eXtended . . . . . . . . . . . . . . . . . 128
A.3 Notung File Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
A.4 Specifying the Species Associated with Each Gene . . . . . . . . . . . . . . . 129
A.5 Punctuation in Species Names . . . . . . . . . . . . . . . . . . . . . . . . . . 130
A.6 Location of Edge Weight Values . . . . . . . . . . . . . . . . . . . . . . . . . 131

B Building a Species Tree 134

C Glossary 136
D Keystroke Shortcuts 139

E Worked Examples 140

E.1 Exercise 1 - Reconciling a gene tree with a species tree . . . . . . . . . . . . 140
E.2 Exercise 2 - Rooting an unrooted tree . . . . . . . . . . . . . . . . . . . . . . 144
E.3 Exercise 3 - Rearranging a gene tree . . . . . . . . . . . . . . . . . . . . . . . 146
E.4 Exercise 4 - Inferring duplications and losses in a binary gene tree with a
non-binary species tree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
E.5 Exercise 5 - Non-binary gene tree with a binary species tree . . . . . . . . . 157
E.6 Exercise 6 - Inferring transfers in a binary gene tree with a non-binary species
tree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171

F Troubleshooting 177

4
c Copyright 2005 - 2019 by the Notung Development Team.

The Notung software package is provided “as is” without warranty of any kind. In no event
shall the authors or their employers be held responsible for any damage or inconvenience
resulting from the use of this software.

Development of certain features, including horizontal transfer inference, was supported in

part by NSF Grant DBI1262593. Any opinions, findings, and conclusions or recommendations
expressed in this material are those of the author(s) and do not necessarily reflect the views
of the National Science Foundation.

5
Chapter 1

Introduction to Notung

Notung offers a unified framework for incorporating event-inference parsimony into phylo-
genetic tasks. This parsimony principle asserts that gene duplication, transfer and loss are
rare events. Notung’s functions embody the assumption that, in the absence of information
from other sources, the phylogenetic hypothesis that requires the fewest number of events to
explain the data is preferred.

Notung provides a graphical interface for tree manipulation and visualization and offers a
command line option that can be used for automated analysis of a large number of trees.
Notung can:

• reconcile a gene tree with a species tree;

• root an unrooted gene tree by minimizing events;

• rearrange rooted trees that have weakly-supported edges to minimize events;

• resolve non-binary nodes in a non-binary gene tree based on event parsimony.

Notung is used in a broad range of applications. It can assist scientists who wish to bring gene
event parsimony models to bear on gene tree construction; evolutionary biologists studying
the history of a gene family; and experimental biologists interested in incorporating evolu-
tionary insights into questions of function and structure.

Notung can be used with either a duplication-loss (DL) model or a duplication-transfer-loss

(DTL) event model. Several features distinguish Notung from other reconciliation programs
that infer transfers. In an event model with transfers, there may be more than one most
parsimonious combination of events. Notung infers all most parsimonious event histories.

6
Chapter 1. Introduction to Notung

Further, Notung tests all candidate event histories for temporal feasibility and screens out
biologically unrealistic histories that imply a transfer that travels backwards in time. Notung
reports only those event histories that are temporally feasible. Notung’s transfer algorithm
makes no assumptions about speciation times and does not require a species tree with branch
lengths.

Notung reconciliations are detailed event histories. In addition to the reconciliation score,
Notung reports specific events and the gene and species lineages in which those events oc-
curred. Specifically, Notung reports gene duplications, and estimates upper and lower bounds
on the time of duplication. Gene losses are inferred explicitly, including the location of the
loss in the gene tree and the ancestral or contemporary species in which they occurred. The
donor and recipient species are reported for every gene transfer.

Notung was the first reconciliation software to reconcile and root non-binary gene trees
with binary species trees and binary gene trees with non-binary species trees, in addition
to traditional analysis with binary gene trees and binary species trees. Notung is unique in
its capability to reconcile and root binary gene trees with non-binary species trees under an
event model that simultaneously minimizes transfers with duplications and losses. Another
novel feature is Notung’s ability to rearrange and resolve non-binary gene trees. The specific
functions that Notung can perform on each combination of inputs are given in Table 1.1.

Gene Tree Species Tree Events Reconcile Root Rearrange Resolve

Binary Binary DTL yes yes CL N/A
Binary Binary DL yes yes yes N/A
Non-Binary Binary DTL CL no CL CL
Non-Binary Binary DL yes yes yes yes
Binary Non-Binary DTL yes yes no N/A
Binary Non-Binary DL yes yes no N/A
Table 1.1: Notung’s main functions on binary and non-binary trees, with or without transfers. CL indicates the function is
only available via the command-line interface.

Large-scale phylogenetic analyses can be carried out using Notung’s command line interface.
The phylogenomics option supports integrated analysis of all gene families drawn from a set
of genomes. In addition to reporting the events inferred for individual trees, this function
aggregates and summarizes results across all reconciled gene trees and all nodes/branches of
the species tree. For example, it reports ancestral gene content and the combined set of events
of each type associated with each node/branch in the species tree. The integrated results
from all reconciled gene families can be used to infer genome-scale evolutionary trends, such
as whole genome duplications, highways of lateral transfer activity, and the timing of genome
expansions or contractions, relative to the species tree.

Notung utilizes novel, efficient algorithms [2, 7, 16, 25, 30] for reconstructing the history of

7
 Chapter 1. Introduction to Notung

gene duplications, transfers and losses, for rooting gene trees based on event parsimony and
for the rearrangement of weakly supported areas of gene trees based on event parsimony.

More information about Notung can be found at:

http://www.cs.cmu.edu/~durand/Notung

More information about other Durand Lab projects can be found at:

http://www.cs.cmu.edu/~durand/Lab

The graphical user interface was partially constructed using the tree visualization library
provided by FORESTER (version 1.92) [32]. Cycle detection, used when inferring transfers,
utilizes the jGraph library.

1.1 How to cite Notung

D. Durand, B. V. Halldorsson, B. Vernot. A Hybrid Micro-Macroevolutionary Approach to

Gene Tree Reconstruction. Journal of Computational Biology, 13(2):320-335, 2006.

B. Vernot, M. Stolzer, A. Goldman, and D. Durand. Reconciliation with non-binary species

trees. Journal of Computational Biology, 15(8):981-1006, 2008. Also appeared in Compu-
tational Systems Bioinformatics: CSB2007 Conference Proceedings, Imperial College Press:
441-452.

M. Stolzer, H. Lai, M. Xu, D. Sathaye, B. Vernot and D. Durand. Inferring Duplications,

Losses, Transfers, and Incomplete Lineage Sorting with Non-Binary Species Trees. Bioinfor-
matics, 28: i409-i415, 2012.

C.A. Darby, M. Stolzer, P.J. Ropp, D. Barker and D. Durand. Xenolog Classification. Bioin-
formatics, 33(5):640-649, 2016.

H. Lai, M. Stolzer, and D. Durand. Fast Heuristics for Resolving Weakly Supported Branches
Using Duplication, Transfers, and Losses. Proceedings for RECOMB International Workshop
on Comparative Genomics: 298-320, 2017.

8
Chapter 1. Introduction to Notung

1.2 Using This Manual

This manual provides a detailed description of Notung, and gives step-by-step instructions
for Notung’s tasks and visualization features. It assumes familiarity with basic concepts of
phylogeny reconstruction. For more information on these subjects, refer to basic textbooks,
such as [10, 21]. A Glossary is provided on page 136. Additional sources are provided in the
Bibliography on page 183.

The manual is organized into numbered chapters by topic. Each chapter begins with para-
graphs describing the topic, followed by a list of step-by-step commands for operations asso-
ciated with the topic. Figures showing the Notung graphical user interface (GUI) have been
included to illustrate program displays and command results. Detailed Worked Examples
are provided on page 140 to guide the user through commonly used functions and features.

Instructions for downloading Notung for various operating systems is provided in Chapter
2. A basic introduction to the Notung GUI is provided in Chapter 3. Theoretical issues
related to inferring reconciliation with horizontal transfer and non-binary trees are discussed
in Chapter 4. Notung’s six task modes are described in Chapter 5 - Chapter 10. Chapter 11
discusses the options for changing the appearance of the tree. Detailed information about
running analyses through the command-line interface, including rearrangement when infer-
ring transfers, as well as batch processing of trees, is located in Chapter 12. More detailed
information about input/output and tree file formats are given in the Appendix A. Tips
regarding troubleshooting can be found in Appendix F.

9
Chapter 2

Downloading Notung

The Notung package can be downloaded from the Notung website as the file Notung-2.9.zip.
When the file is unzipped, it will create a folder called Notung-2.9 that includes: this manual;
a folder of sample trees which contains a folder of a sample batch run; and the Notung program
file, Notung-2.9.jar.

Notung is supported on Windows XP, Windows Vista and Windows 7; Mac OS X 10.5 and
above; and Linux. To run Notung, Java must be installed on your computer. Notung has
been tested under Java 1.5–7, but should work for newer versions of Java.

To download Notung-2.9:

Go to http://http://goby.compbio.cs.cmu.edu/Notung/download29.html

To unzip Notung-2.9.zip:

On Windows:

• If you are running Windows XP or newer, double click on Notung-2.9.zip.

• If you are running Windows 2000 or earlier, use jZip to open Notung-2.9.zip and extract
its contents. If you do not have jZip, go to http://www.jzip.com and download the
jZip application.

On Mac:

• Double click on Notung-2.9.zip.

10
Chapter 2. Downloading Notung

On Linux :

• Run the command:

unzip Notung-2.9.zip

If you do not know if you have Java:

• Go to http://www.java.com/en/download/help/testvm.jsp, or, in a terminal or

command window, type

java -version

Notung requires at least Java 1.5.

To get Java (if you do not have it):

• For Windows and Linux, go to: http://java.sun.com/webapps/getjava/BrowserRedirect

• For Mac, use the Software Update application to install or update Java.

11
Chapter 3

Getting Started

Notung is a tool for comparing gene and species trees. Notung takes tree files as input and
allows users to refine and manipulate them. The modified trees can be saved as output. The
following subsections introduce basic input and output in Notung, general tree statistics, the
graphical user interface, and the parameter values used in Notung’s tree refinement tasks.

3.1 Gene and Species Trees

To perform its functions, Notung requires a gene tree and a species tree. The species tree
must contain all the species from which genes in the gene tree were sampled. The species tree
may contain additional species as well - these can be ignored. A correspondence between the
leaves of the species and gene trees is determined by comparing the leaf labels in the gene
and species trees: each leaf label in the gene tree must include a substring or an NHX species
field that specifies the species from which the gene was sampled. Trees may be provided in
Newick, NHX, or Notung format. See Appendix A - File Formats on page 126 for further
information.

NOTE: If your gene tree contains elements from species you do not wish to
consider, the command line can be used to prune these taxa. See Chapter 12 -
Command Line Options and Batch Processing on page 98 for more information.

Notung can operate on a non-binary gene tree or a non-binary species tree. However, its
functions cannot be performed when both the gene tree and corresponding species tree are
non-binary. In the GUI, transfer inference is limited to cases where the gene tree is binary.
For a complete summary of functions that Notung can perform, see Table 1.1 on page 7.

12
Chapter 3. Getting Started

NOTE: If you are interested in using Notung to analyze non-binary trees or infer
transfers, you may wish to see Chapter 4 - Theory on page 25 for more a more
detailed and theoretical discussion.

Species Trees

The species tree must be rooted, with leaf nodes labeled with species names. Internal nodes
may be given taxonomic labels (e.g., “tetrapoda”), but this is not required. If the internal
nodes are not labeled, Notung will assign alphanumeric labels (such as n1, n2, etc.). If the
species tree has edge weights or branch lengths, this information will be displayed, but will
be ignored for event inference. For more information on species names, see Appendix A.4 -
Specifying the Species Associated with Each Gene on page 129.

The tasks that Notung performs are based on the assumption that the user has selected a
species tree that is a reliable representation of the true species relationships. Using Notung
with an incorrect species tree will give incorrect results. For more information on selecting
an appropriate species tree, see Chapter B - Building a Species Tree on page 134.

Gene Trees

In order to perform its reconcile, rearrange and resolve functions, Notung requires a rooted
gene tree. If the gene tree is not rooted, Notung can be used to root the gene tree. See
Chapter 6 - Rooting Mode on page 68 for more information. The leaf nodes in the gene
tree must be labeled with a unique identifier specifying the gene, as well as the species from
which the gene was sampled. See Appendix A.4 - Specifying the Species Associated with
Each Gene on page 129 for more information. The internal nodes may be labeled. If the
internal nodes are not labeled, Notung will assign alphanumeric labels (e.g., n5, n6, etc.).

In Rearrangement mode, Notung requires that the tree have edge weights. These are used
to identify edges that are weakly supported and may be rearranged. These weights may
be bootstrap values, posterior probabilities, edge lengths, or any other weighting scheme
selected by the user. Several different fields in the Newick and NHX formats may be used to
store edge weights. See Appendix A - File Formats on page 126 for a detailed explanation of
these formats and how to indicate to Notung which field is being used for edge weights in a
particular input tree.

Unrooted binary gene trees Many tree reconstruction programs represent an unrooted
binary tree as a mostly binary tree, with a single trifurcation at the root. Unless a root is

13
 Chapter 3. Getting Started

selected for these trees (in Notung or another program), Notung will incorrectly treat them
as rooted non-binary trees. If such a tree is actually an unrooted binary tree, failing to
root it will affect Notung’s diagnostics. See Chapter 6 - Rooting Mode on page 68 for more
information on rooting gene trees.

3.2 The Graphical User Interface

Notung’s graphical interface facilitates tree visualization and manipulation, enabling the
user to inspect duplicated, transferred and lost nodes in a tree, view orthologs, paralogs, and
xenologs, visualize alternate optimal trees, and color annotate genes for visual differentiation
or presentation.

To run Notung:

Using the graphical user interface on Windows or Mac:

• Unzip the downloaded file.

• Double-click on the file Notung-2.9.jar.

Using the graphical user interface on Linux:

• Unzip the downloaded file and move into the Notung directory.

• Enter in the command line:

java -jar Notung-2.9.jar.

- OR -

• Enter in the command line:

java -jar "PATH_TO_NOTUNG"

In addition, Notung can perform many of its operations from the command line without
launching the GUI. (See Chapter 12 - Command Line Options and Batch Processing on page
98 for a description of the command line interface.)

14
Chapter 3. Getting Started

When Notung is first launched, the program window will be blank. Figure 3.1(a) and Figure
3.1(b) show Notung’s graphical interface once a gene tree and species tree have been opened.
Notung’s graphical user interface has the following components:

Tree panel: The tree that is currently selected appears in the tree panel. Trees are rendered
with the root at left and leaf nodes at right. Nodes are denoted by small blue squares in the
tree. Edge weights and leaf node names appear in the tree by default. Notung fits the whole
tree in the tree panel by default. The size of the tree and tree labels can be modified using
the Zoom and Fonts menus, respectively. See Chapter 11 - Changing the Appearance of the
Tree Panel on page 93.

Although multiple trees can be open in Notung at once, Notung operates on only one tree at
a time. To facilitate working with many trees, Notung marks each open tree with a tab at
the top of the tree panel. Clicking on a tab selects the corresponding tree. Tabs are labeled
with the file name and special icons to identify them as a gene or species tree - a DNA helix
for gene trees, and a cartoon of the evolution of humankind for species trees (see Figure 3.2).

Task panel: Operations on the tree are performed in the task panel (highlighted in blue
in Figure 3.1). Tabs at the top of the task panel correspond to the various tasks that Notung
can perform. Clicking on a tab puts Notung in the corresponding task mode, revealing the
buttons that control tasks specific to that mode. If a gene tree is selected, six modes are
available: History, Reconciliation, Rooting, Rearrange, Resolve, and Annotations. Only the
History and Annotations modes can be used when a species tree is selected.

Parameter values: When a gene tree is selected, a box displaying the Edge Weight Thresh-
old and Costs/Weights for Duplications, Transfers, Co-Divergences, and Losses appears in
the bottom-right corner of the program window. These values can be changed by clicking the
“Edit Values” button directly below them. A value for transfers can only be edited if the
“Infer Transfers” option is selected. Note that when a species tree is selected, the program
window will not display the parameter values.

3.3 File Menu and Opening and Saving Trees

Notung can read and save tree files in Newick, NHX, and Notung file formats. NHX and
Notung file formats are extensions of Newick; see Appendix A - File Formats on page 126
for details. Notung can also save the image in the tree panel as a Portable Network Graphic
(PNG) file.

To open trees:

15
 Chapter 3. Getting Started

(a)

(b)

Figure 3.1: Notung’s graphical user interface displaying (a) a gene tree, and (b) a species tree. The tree panel is highlighted
in red, the task panel in blue, and the parameters panel in yellow. Only the tree panel and the task panel are
applicable to species trees.

1. Click “File → Open Gene Tree” or “File → Open Species Tree.”

16
Chapter 3. Getting Started

Figure 3.2: Tree tabs for a gene tree (left) and a species tree (right)

2. In the Open dialog box, select a tree file and click “Open.”

NOTE: Notung cannot distinguish gene trees from species trees automati-
cally. If a gene tree is opened as a species tree, or a species tree is opened as
a gene tree, reconciliation will produce incorrect results.

To save trees:

1. Click “File → Save As.”

2. In the drop-down menu, “Files of Type,” select one of the following formats:

• Notung File Format

• Newick File Format
• NHX File Format

3. Click “Save.”

NOTE: The default format for saving trees is the Notung File Format. If you
have modified the tree in Notung and wish to reopen this tree in Notung,
it may be best to save the tree in Notung format. If you wish to reopen
the modified tree in another tree program, Newick format may be a better
option. See Appendix A - File Formats on page 126 for more information.

To view text formatted trees in a dialog box:

1. Click “File → View Tree in Text Format.”

2. In the drop-down menu, select one of the following formats:

• Tree in Notung Format

• Tree in Newick Format
• Tree in NHX Format

To copy this information, click the “Copy to clipboard” button. This text can then
be pasted in any text editor.

3. When finished reviewing this information, close this window to continue using Notung.

17
 Chapter 3. Getting Started

To save the current view of a tree as a PNG file:

• Click “File → Save Current View as Image (PNG).”

NOTE: This option saves only the image currently visible in the tree panel.
If you have zoomed in on a tree, the PNG will save only the section in view.

To save an image of the whole tree as a PNG file:

• Click “File → Save Whole Tree as Image (PNG).”

NOTE: This option saves a “pretty print” version of the entire tree. Cur-
rently, display options set in Notung will not affect the output of this tree.
More options for saving tree images are available via the command line, as
discussed in Section 12.5 - Saving PNG Images of Trees

To print an image of a tree:

1. Click “File → Print Current View.”

2. The print dialog box will appear. Change the settings as necessary and click “Print.”

NOTE: For most printers the default page layout will be portrait; however,
the landscape layout is usually preferred for printing trees from Notung. You
may wish to change your printer settings before printing.

3. A red rectangle will appear in the tree panel. Only the view inside this rectangle will
be printed.

4. To proceed with printing, click “Print.”

5. If you wish to change the printer’s settings or the size of the tree, click “Cancel.” The
red rectangle will disappear and the appearance of the tree can be manipulated.

NOTE: Printing a view of the tree that shows exactly what you want may
be difficult as it may be necessary to change both the printer’s settings (i.e.
page layout, margins, etc.) and the appearance of the tree so that the desired
print area fits within the red rectangle. See Chapter 11.2 - Zoom on page 95
for more information on zooming in and out of the tree. It may be easier to
obtain the desired view by first saving the tree as a PNG image, and then
editing and printing that image using another program.

18
Chapter 3. Getting Started

To reload a tree:

• Click “File → Reload from file.”

NOTE: If the tree has been modified, a dialog box will be displayed. The
dialog box will offer you one of the three following options: “Save tree”;
“Reload tree without saving”; “Cancel reload”.

To export color annotations to a file:

1. Click “File → Export Annotations.”

2. Provide a file name and click “Save.”

NOTE: Exported annotations can be imported into other trees, or loaded on

the command line using the option --annotationfile. For more information
about color annotations, see Chapter 10 - Annotations on page 87.

To import color annotations from a file:

1. Click “File → Import Annotations.”

2. Select the desired annotations file and click “Open.”

NOTE: Annotations can be imported from previously exported annotations

files. Additionally, selecting a Notung format tree which contains annotations
will import annotations from that tree. Annotations can also be loaded via
the command line using the option --annotationfile. For more information
about color annotations, see Chapter 10 - Annotations on page 87.

To close trees:

1. Select the tree to close.

2. Click “File → Close.”

To quit Notung:

• Click “File → Exit.”

19
 Chapter 3. Getting Started

3.4 General Tree Statistics

Notung compiles information on tree characteristics, such as height, number of leaves, number
of nodes, etc. Notung reports this information is reported in the general tree statistics pop-up
box under the “About This Tree” menu. The properties examined depend on whether the
given tree is a gene tree or a species tree, and whether the gene tree has been reconciled or
not. A description of the possible information displayed is described below.

For all trees

Statistics for: the file name for the given tree.

Total nodes: the total number of nodes.

Internal nodes: the total number of internal nodes (Total nodes minus Leaf nodes).

Leaf nodes: the total number of leaves.

Polytomies: the total number of polytomies in the tree. This number will be zero if the
tree is binary.

Size of largest polytomy: the number of children of the largest polytomy in the tree. This
number will be zero if the tree is binary.

Height: the maximum path length from a leaf node to the root.

Figure 3.3 shows an example of the tree statistics provided for a species tree.

For gene trees, but not for species trees

Edge Weight Range: the range of edge weights in the gene tree in the form, [minimum
edge weight, maximum edge weight].

For reconciled gene trees

If the tree has been reconciled and the current solution is feasible, additional information on
the number of inferred events is provided under the heading Reconciliation Information:

20
Chapter 3. Getting Started

Figure 3.3: General tree statistics for a species tree.

Duplications: the total number of duplications in the reconciled gene tree.

Transfers: the total number of transfers in the reconciled gene tree. This number will be 0
if only duplications and losses are optimized, or if no transfers are inferred.

Co-Divergences: the number of co-divergences in the reconciled gene tree. This number
will be zero if the associated species tree is binary or there are no co-divergences. See
Chapter 4.2 - Non-Binary Trees on page 29 for more information on co-divergences.

Losses: the total number of losses in the reconciled gene tree.

This is followed by statistics on the topology of the reconciled tree (number of leaf nodes,
number of internal nodes, etc.) with and without losses, and then by statistics on the
topology of the associated species tree. If the tree has been reconciled, but there are no
feasible solutions, a message stating such will be reported and include the file name for the
associated gene or species tree along with the current event costs.

Figure 3.4 shows an example of the tree statistics displayed for a reconciled gene tree.

To get general statistics for a tree:

• Click “About This Tree → General Tree Statistics.”

21
 Chapter 3. Getting Started

Figure 3.4: General tree statistics for a reconciled gene tree.

A window will appear containing information on the tree’s characteristics,

as described above. To copy this information into your favorite text editor,
click the “Copy to Clipboard” button, and paste in the text editor.

NOTE: Information on duplications, transfers, and losses can also be gathered

through the “About This Tree Menu” with Event Summary. For more infor-
mation, see Chapter 12.2 - Duplication Bounds and Loss Information on page
104.

3.5 Parameter Values

The parameter values used in Notung — the Edge Weight Threshold, Duplication Cost,
Transfer Cost, Co-Divergence Cost, and Loss Cost — can be specified by the user. These
values influence the results produced by Notung’s tasks.

Notung uses a Duplication/Transfer/Loss (DTL) Score to score reconciled trees and evaluate
alternate hypotheses. The DTL Score is defined to be: cL L+cD D+cC C +cT T where L is the
number of losses, D is the number of duplications, T is the number of transfers, and C is the
number of co-divergences implied by the current reconciliation. The loss cost, cL , duplication
cost, cD , transfer cost cT , and co-divergence cost, cC reflect the relative importance of losses,

22
Chapter 3. Getting Started

duplications, transfers and co-divergences in scoring the tree. The cost of co-divergences is
only relevant when reconciling a gene tree with a non-binary species tree (see Chapter 4.2 -
Non-Binary Trees on page 29). Likewise, the cost of transfers is only relevant when including
transfers in the optimization. The default cost values are 1.0 for losses, 1.5 for duplications,
3.0 for transfers, and no cost for co-divergences, but these values can be changed by the
user. Notung displays the DTL Score of a reconciled tree, as well as the number of losses,
duplications, transfers and co-divergences, in the bottom-left corner of the program window
(see Figure 3.5).

Figure 3.5: If the gene tree has been reconciled, the DTL Score, the number of duplications, transfers, co-divergences and
losses, the number of optimal solutions, and the species tree used to reconcile it appear at the bottom of the
program window.

The Edge Weight Threshold is a parameter used to define the set of strong edges in the gene
tree. In Rearrange mode, edges weighted below the Edge Weight Threshold are considered
weak and may be rearranged (for more information about rearrangement, see Chapter 7 -
Rearrange Mode on page 73). Edges with no weight specified are assigned an edge weight of
zero, and are considered to be weak. The default threshold is 90% of the highest edge weight
in the gene tree file. If no edge weights are found, the threshold is set to one. The user may
change this cutoff if a different threshold is desired for the current data set.

NOTE: For some sources of edge weights, such as bootstrap values, setting the
threshold to a percentage of the highest edge weight works well. For other sources,
such as branch lengths, where a single very large value could cause all other edges
in the tree to be weak, it may be better to set the threshold with a fixed, minimum
value.

To change the parameter values:

1. Click the “Edit Values” button. A dialog box appears.

2. Enter the appropriate values in the text field, and then click “Apply Changes.”

NOTE: This will change the value settings only for the gene tree that is cur-
rently selected. Also, each history state saves the parameter values used at
that state; when moving through the history, parameter values may change
depending on the state and tree viewed. For more information on history

23
 Chapter 3. Getting Started

states, see Chapter 9 - History on page 85.

NOTE: Transfer costs can only be edited if the “Infer Transfers” option
is selected.

24
Chapter 4

Theory

If you plan to infer transfers or to analyze either non-binary gene trees or non-binary species
trees using Notung, this chapter is recommended. If you will be working solely with binary
trees and a duplication-loss event model, you may skip ahead to the chapters describing the
specific tasks you wish to perform.

Reconciliation relies on the observation that discordance between gene and species trees is
evidence that genes diverged through processes other than speciation, including gene dupli-
cation, gene loss, and horizontal gene transfer. Reconciliation establishes a correspondence
between the evolutionary histories of genes and species. This correspondence, called a map-
ping, is constructed between nodes in the trees, revealing which ancestral entities coexisted
and where gene and species evolution did not proceed in a coordinated manner. The events
that best explain such incongruence are then inferred. In a parsimony framework, these are
defined as the events that minimize the Event Score, a weighted sum:

π = cD D + cL L + cT T + . . . (4.1)

The reconciliation of a gene tree with a species tree results in an annotated gene tree, in
which every internal node is annotated with the species that contained the gene and the
event that caused the divergence. In addition, edges in the gene tree are annotated with the
genes lost, labeled with the species in which the loss occurred. In a parsimony framework,
the event inference problem can be stated formally as follows:

The Event Inference Reconciliation Problem

Input: A rooted species tree; a rooted gene tree; the leaf mapping, indicating from
which species the genes were sampled; an event model.
Output: All reconciliations that minimize the weighted sum of the inferred events.

25
 Chapter 4. Theory

Notung’s default mode is reconciliation of a binary gene tree and a binary species tree under
a duplication-loss (DL) event model. It minimizes the DL score,

π = cD D + cL L. (4.2)

Algorithms designed for Notung utilize this DL Score to reduce uncertainty in gene trees by
highlighting promising roots [2] and rearranging weakly supported areas [7]. See Chapter
6 - Rooting Mode on page 68 and See Chapter 7 - Rearrange Mode on page 73 for more
information.

Alternatively, users can select a duplication-transfer-loss (DTL) model, which minimizes a

DTL Score,
π = cD D + cL L + cT T. (4.3)
This model is discussed further in Section 4.1.

NOTE: Event costs are supplied by the user. Higher costs for transfers, com-
pared to duplication and loss, should be used in situations where duplications are
thought to be the dominant event, while the opposite should be utilized in cases
where transfers are the dominant event (e.g., in bacteria).

The user may also opt to consider either a non-binary species tree or a non-binary gene tree
(but not both at the same time). These non-binary tree functions are based on algorithms
we developed for Notung [25, 30] and are discussed further in Section 4.2.

NOTE: Notung can only infer transfers for a binary gene tree. For a complete
listing of the functions that Notung is able to perform on binary and non-binary
trees, see Table 1.1 on page 7. For more information on all the algorithms im-
plemented in Notung, please see our work listed in Chapter 1.1 - How to Cite
Notung.

4.1 Transfers

Horizontal gene transfer, the transmission of genetic material from an organism in one species
to the genome of an organism in another species, is a common phenomenon in prokaryotes.
While the extent and importance of horizontal transfer in eukaryotes is less well-understood, a
growing body of evidence for HGT in eukaryotes (e.g., [1,31]) and duplications in prokaryotes
(e.g., [24]) calls for models that capture both events [9]. Most event-inference reconciliation
algorithms consider either gene duplication or HGT [6, 19, 20], but not both. Our algorithm

26
Chapter 4. Theory

for binary species trees is inspired by the Duplication-Transfer (DT) models of Tofigh et
al. [29], but differs in that it includes losses in the optimization criterion. In our model, the
event history that best explains the disagreement between gene and species tree is the set of
duplications, transfers and losses that minimizes the DTL Score.

In Notung, transfers are represented by a node and an edge. If a transfer is the cause of a
divergence (internal node) in the gene tree, one of the children of that divergence remains
in the donor, while the other is copied into the recipient. The edge leading to the node
representing the recipient’s copy is annotated as a transfer edge; the parent node is assigned
the transfer event.

The inclusion of transfers in the event model introduces some interesting complexities, out-
lined below. First, when the event model includes transfers, there can be more than one most
parsimonious event history. This stands in contrast to reconciliation under the DL model,
where the most parsimonious reconciliation of a given gene and species tree is unique. In
addition, it is possible to construct an event history with transfers that requires traveling
backward in time. Care must be taken not to infer such temporally infeasible reconciliations,
since they obviously cannot represent the actual evolutionary history of a gene family. Both
of these issues are discussed in greater detail below.

4.1.1 Multiple Solutions

More than one series of duplications, transfers and losses may result in the same pattern of
incongruence because, with transfers, the gene tree is no longer confined to the structure of
the species tree. When in a model with duplications, but no transfers, the ancestral species
associated with a given node in the gene tree is uniquely determined by the species associated
with the children of that node. However, transfers jump from one species to another, allowing
gene tree nodes to map to distantly related species. In this case, the mapping from ancestral
genes to ancestral species is not unique. Further, there may be two or more combinations of
events with the same DTL score.

Multiple optimal reconciliations are a frequent occurrence, especially in data sets where
transfer is the dominant process. To assist the user in considering these alternate hypothe-
ses, Notung presents all temporally feasible, optimal solutions. In the GUI, the user may
explore other optimal solutions through a point-and-click interface described in Chapter 5
- Reconciliation Mode. In the command line version, the user may ask to see several or all
multiple optimal solutions by setting a flag.

27
 Chapter 4. Theory

Temporal Infeasibility

In order to be biologically valid, an event history that includes transfers must be temporally
feasible. There is no known constructive algorithm for finding an optimal DTL-reconciliation
that is temporally feasible (the problem of finding such a reconciliation is NP-complete [13]).
Instead, Notung generates candidate optimal reconciliations and then checks each candidate
for temporal feasibility. If a candidate reconciliation is feasible, then it is also guaranteed to
be optimal; i.e., to have minimum DTL score.

Stated technically, a reconciliation is temporally feasible if it is possible to assign a temporal

ordering to nodes in the species tree such that all timing constraints imposed by the species
tree and the inferred transfers are satisfied. The species tree implies temporal constraints
because ancestral nodes must have existed before the nodes that descended from them.
Transfers imply two types of constraints. First, an inferred transfer could only have occurred
if the donor and recipient species co-existed. Second, if one transfer occurred before another
in the gene tree, then the donor and recipient of the first transfer must have occurred before
or at the same time as the donor and recipient second transfer. These constraints are modeled
using a directed graph, called a timing graph.

A history with two or more transfers could result in a situation where there is no temporal
ordering that makes sense. If this is the case, the timing graph will contain a cycle. Otherwise,
the graph is acyclic, meaning there exists a temporally feasible ordering of nodes.

Notung does not report infeasible solutions. We test each history generated by our algorithm
to determine whether there is a feasible solution. The problem of finding the next best
solution, when all the optimal solutions turn out to be impossible, has been shown to be NP-
complete [13]. We currently do not have a solution to this problem. If all inferred optimally
scored solutions are infeasible, Notung only provides a warning and outputs no histories.

NOTE: Transfers can be inferred in polynomial time under a restricted model

where only transfers between contemporaneous species are considered. This
model, reviewed in [6, 15], requires estimates of speciation times, which are fre-
quently not known. In addition, this approach may fail to recognize transfers if
they involve a taxon missing from the data set [15, 19]. We make no assumptions
about which donor and recipient species co-existed before transfers are inferred,
except to prohibit transfers between ancestors and descendants.

28
Chapter 4. Theory

4.2 Non-Binary Trees

A non-binary, or multifurcating, tree is a tree in which at least one node has more than two
children. Such nodes are referred to as polytomies, or non-binary nodes. A polytomy can
have several meanings [17]. A hard polytomy represents the true, simultaneous divergence
of all its children. A soft polytomy, on the other hand, refers to the situation where the
true pattern of divergence is binary, but the branching pattern is unknown. A tree with
soft polytomies is sometimes referred to as an “unresolved tree.” The indication is that the
polytomies can be “resolved,” meaning replaced with a series of binary divergences, to the
true binary branching pattern.

In Notung, polytomies are represented as vertical edges with more than two children. See,
for example, the polytomy in Figure 4.1.

Figure 4.1: Notung displays trees as cladograms. Polytomies are drawn as vertical edges with more than two children. This
tree contains only one polytomy, indicated by the arrow.

Interpreting disagreement between gene and species trees as evidence of gene duplication,
transfer, and loss is widely accepted when both trees are binary. Disagreement between
non-binary trees is less well-understood and there is no universally accepted approach to
non-binary reconciliation. In the next sections, we briefly review current theory regarding
non-binary nodes in gene and species trees and discuss how we apply these theories in Notung.
First, we discuss how Notung deals with incomplete lineage sorting when reconciling binary
gene trees with non-binary species trees. In the section following this, we discuss how Notung

29
 Chapter 4. Theory

considers the multiple, possible binary histories represented by a polytomy in a gene tree and
presents the most parsimonious set of events.

4.2.1 Fitting Binary Gene Trees to Non-Binary Species Trees

Non-binary species trees may be common; for example, 64% of branch points in the NCBI
Taxonomy Database [3] have three or more children.

Since a species tree represents the evolution of a population of organisms, a polytomy can
be considered from two perspectives: hard or soft polytomies. Hard polytomies can result
from several events, such as the isolation of subpopulations within a widespread species by
sudden meteorological or geological events, or from rapid expansion of the population into
open territory, resulting in reproductive isolation. Soft polytomies are frequently encountered
in species trees when the signal is weak or conflicting, resulting from insufficient evidence for
any particular binary branching pattern. Soft polytomies often occur if a sequence of binary
divisions proceeds rapidly and the time between these events is insufficient to accumulate
informative variation. These interpretations of polytomies are sometimes closely linked. The
branching order of species that arose through multiple speciations in rapid successions [8,23]
is often difficult to resolve.

Incomplete Lineage Sorting

Incomplete lineage sorting (ILS) refers to discordance between gene and species trees
resulting from allelic variation. Since a node in the species tree represents the evolution
of a population of organisms with genetic diversity, multiple alleles may be present at the
locus of interest. When lineages diverge, a different allele may fix in each lineage. The
resulting gene tree will be binary and will reflect the order in which new alleles arose in the
ancestral population. This pattern of divergence may not always be congruent to the pattern
of divergence in the species lineage. For example, Figure 4.2 shows three different binary
branching processes of a gene tree in the context of a species polytomy.

A true divergence between two genetic lineages corresponds to the point where allelic differ-
ences arose, not the time of speciation. Genetic divergence that greatly predates the time of
speciation is referred to as deep coalescence. In Figure 4.2, for example, the divergence
at x occurs much earlier than the separation of species A, B, and C, and represents deep
coalescence.

The probability of incomplete lineage sorting decreases as the time between speciation events
increases [14,18,22,27,28]. If branch lengths in the species tree are sufficiently long, the effect

30
Chapter 4. Theory

Figure 4.2: Three possible outcomes of the evolution of a single genetic locus in the context of a population. Different gene
families associated with the same species polytomy may have different binary branching patterns.

of incomplete lineage sorting on discordance between gene and species trees is negligible, and
does not need to be considered. However, when the species tree is non-binary, or has very
short branches, incomplete lineage sorting is a plausible explanation for gene and species tree
disagreement.

Inferring Events

Reconciliation focuses on incongruence that arises from processes that change the number of
loci in a gene family; i.e., duplication, loss, and transfer. It implicitly assumes that times
between speciation events are such that genetic drift and incomplete lineage sorting may
be safely excluded from consideration. This assumption breaks down when the species tree
contains polytomies or very short branches. In these situations, allelic variation can survive
multiple speciation events, leading to gene trees with branching patterns that differ from the
species tree. Such cases are increasingly common. Whole genome sequencing data of closely
related species is revealing an ever growing number of cases where ILS is active along with
duplication, loss, and transfer (e.g., [1, 24, 31]), leading to calls for algorithms that model
multiple evolutionary processes [5, 9]. Methods that do not consider ILS will incorrectly
interpret incongruence arising from ILS as evidence of duplication or transfer.

Notung algorithmically avoids this by distinguishing between regions of the species tree where
ILS is likely, and those where only gene duplication and transfer need be considered. These
differences are specified using a non-binary species tree. Notung assumes that the probability
of incomplete lineage sorting is negligible when a node in the species tree is binary. In this
case, ILS is considered to be so rare that disagreement between the trees is interpreted as
evidence only for gene duplication, transfer or loss. In contrast, if a binary gene tree node is
associated with a species polytomy, by definition the trees disagree at that point. A binary
divergence is not congruent with a non-binary divergence. How should events be inferred in
this case?

Notung treats a non-binary node in the species tree as a hard polytomy, where disagreement
can be due to the creation of a new locus, through gene duplication or transfer or the binary
divergence of a single locus, through a deep coalescence. Both of these cases are illustrated

31
 Chapter 4. Theory

(a) (b) (c)

(d) (e)

Figure 4.3: Black squares with a “D” indicate duplications. Losses are represented by dotted lines. (a) A marsupial species
tree with a polytomy. (b) The phylogeny of a hypothetical gene family sampled from the same marsupial species.
(c) Hypothesis 1: the disagreement between (a) and (b) can be explained by deep coalescence (node x), followed
by gene duplication (node y). (d) Hypothesis 2: the disagreement between (a) and (b) can also be explained
by duplication at x, followed by gene loss, followed by duplication at y. (e) The divergence at x is designated a
co-divergence (gray square) because it is not possible to determine whether the disagreement is due to duplication
or deep coalescence. The divergence at y is a duplication.

in Figure 4.3.

Our conservative approach is to assume that incongruence is due to ILS except when topolog-
ical disagreement cannot be explained by ILS. This means disagreement must have resulted
from the creation of a new locus through either duplication or transfer. A key aspect of our
work is that, sometimes, it is not possible for ILS to explain all disagreement between the
trees, even if the gene tree is uniquely labeled.

NOTE: This model can be invoked for both non-binary species trees and for
binary species trees with short branches where ILS is suspected: Even if the
binary branching pattern of the species tree is known, the user can collapse edges
in the species tree to indicate in which lineages ILS should be considered as an
alternate hypothesis (or where incongruence should not be interpreted only as
evidence of an event).

Our model allows a useful interpretation for soft polytomies. In this case, we can view the
polytomy as a set of hypotheses, namely the binary resolutions of the polytomy. Since the

32
Chapter 4. Theory

(a) (b) (c)

Figure 4.4: Losses associated with a polytomy in the species tree are ambiguous. (a) A species tree with a polytomy. (b) A
gene tree drawn from the species in (a), with a loss in species A. (c) This loss can be assigned to three possible
edges. Associating a loss with the green edge implies that g A diverged first and was then lost. the blue edge
implies that g A was lost after the divergence of g C ; the red edge implies that A was lost after g B diverged.

true branching pattern is unknown, it is possible that the binary branching pattern in the gene
tree in fact corresponds to the true (but unknown) branching order of the species. Notung
only infers an event when the gene tree does not conform to any one of the hypotheses, i.e.,
the binary resolutions. Otherwise, the gene tree is congruent with one of the hypotheses.
Notung infers that there could have been a co-divergence, and no penalty is incurred.

NOTE: Notung implements novel reconciliation algorithms [25,30] for non-binary

species trees that distinguishes among duplications, transfers, and co-divergences
and reports them separately.

Inferring Losses

Inferring loss events is also fundamentally different when the species tree is non-binary. When
both trees are binary, an inferred loss node can always be unambiguously assigned to a specific
edge in the gene tree, indicating when in the history of the gene family the loss occurred.
The node is labeled with the species in which the loss occurred. However, when a loss is
associated with a polytomy in the species tree, it is not generally possible to assign the loss
to a single edge in the gene tree. Rather, the loss can be associated with a set of candidate
edges, each of which corresponds to an alternate hypothesis regarding when the loss occurred.
The inferred loss must have occurred on one of the edges in this set, but it is not possible to
determine which one. Figure 4.4 shows an example of this ambiguity when assigning a gene
loss in species A. This loss could be associated with any of the three colored edges indicated
in Figure 4.4(b). The three hypotheses resulting from the three possible ways of assigning
the loss to an edge can be seen in Figure 4.4(c).

In a complex reconciliation with several losses, there may be many alternative hypotheses

33
 Chapter 4. Theory

(i.e., reconciliations with different loss histories) to consider. Notung uses parsimony to
reduce the number of candidate reconciliations. Specifically, Notung assigns each loss to a
specific edge within the set of candidates, with a goal to minimize the total number of losses.

The total number of losses depends on two factors. The first is the position of the loss relative
to other events in the gene tree. For example, assigning a loss to an edge above a duplication
implies that the loss occurred before the duplication, and only one loss is inferred. However,
assigning the loss to an edge below the duplication implies that the duplication occurred first.
Thus, two losses are inferred – one for each duplicated copy. Second, in some circumstances,
losses in sibling species can be more parsimoniously explained by a loss in their common
ancestor. The total number of losses may be reduced by assigning losses in such a way to
maximize the number of cases where multiple losses can be replaced by a single loss in an
ancestral species. These two factors are not independent of one another. While assigning
a loss lower in the tree will often increase the total number of losses, in some cases, these
extra losses may be combined with others assigned to the same edges, thus reducing the total
number of losses.

Two algorithms for inferring losses, one exact and the other a heuristic, have been imple-
mented in Notung. The exact algorithm infers a history with the fewest losses, taking both of
the above considerations into account. This algorithm is computationally intensive because
all possible combinations of loss assignments must be considered. Its worst-case running
time is an exponential function of the size of the largest polytomy in the species tree. In
practice, the exact algorithm performs efficiently on non-binary species trees with small poly-
tomies. However, users should be prepared for extended running times if the species tree has
a polytomy with more than 12 children.

NOTE: While the exact algorithm is guaranteed to return a reconciliation with

a minimum number of losses, there may be more than one such optimal reconcil-
iation; if so, Notung reports only one.

The heuristic runs significantly faster than the exact algorithm and yields the same results
in many, if not most, cases. It returns only one reconciliation, which is not guaranteed to
be optimal. However, in a comparison of the two methods on the 1,174 trees from TreeFam,
the heuristic found an optimal solution for more than 99% of the trees. Of the seven trees
where the heuristic did not find an optimal solution, in the worst case, the number of losses
was overestimated by four losses from a total of 249.

Both algorithms are integrated with the algorithm to identify duplications; however, only the
heuristic algorithm is used when inferring transfers. The interactive version of Notung uses
the heuristic to reconcile binary gene trees with non-binary species trees. The exact algorithm
is only available for duplication-loss parsimony and only in the command line version. See

34
Chapter 4. Theory

Chapter 12 - Command Line Options and Batch Processing on page 98 for information about
these options.

4.2.2 Fitting a Non-Binary Gene Tree to a Binary Species Tree

In a gene tree, each lineage represents a single gene and the result of any divergence is exactly
two descendant sequences. Thus, in contrast to species trees, the true branching pattern in a
gene tree is always binary [14], and all multifurcations are soft polytomies. For this reason,
non-binary gene trees are also referred to as unresolved trees. Some phylogeny reconstruction
programs output non-binary gene trees when the true binary branching process cannot be
resolved. Such uncertainty often arises if binary divisions occur too rapidly to accumulate
informative variation or if the data set is noisy.

Notung’s approach to reconciling non-binary gene trees rests on the assumption that the chil-
dren of a polytomy arose through an unknown series of binary divergences. Notung further
assumes that, in the absence of other information, the best hypothesis for the true evolution-
ary history of the children of the polytomy is the binary branching pattern that entails the
fewest duplications and losses; there may be more than one such binary resolution of a
polytomy. The problem of reconciling non-binary gene trees reduces to finding a binary tree
that agrees with the original tree everywhere except at the polytomies and has a minimal
DL Score.

The general approach is as follows: A non-binary gene tree is converted into a binary gene
tree by replacing each polytomy with a temporary binary resolution. This resolution is op-
timal under duplication-loss parsimony, when reconciled with the appropriate binary species
tree. The resolution is determined by using our rearrangement algorithm [7], which con-
structs an optimal duplication-loss parsimony tree in polynomial time per tree. Following
rearrangement, all nodes and edges not present in the original gene tree are then removed, to
obtain a reconciliation of the original non-binary gene tree. As nodes and edges are removed,
any duplications or losses assigned to them are reassigned to their associated polytomy.

This process is illustrated in Figure 4.5. The optimal resolution of the polytomy at node
z in the gene tree in (b) with the species tree in (a) is shown in the right subtree of (c).
This entails one duplication and one loss. This information is mapped onto the original gene
tree (b) to obtain the reconciled, non-binary gene tree in (d). The polytomy in the original
tree represents uncertainty, as reflected in the reconciliation. The reconciled polytomy in the
right subtree of (d) tells us that at least one duplication and one loss occurred in the subtree
rooted at z, but the exact order of these events is unknown.

Note that multiple duplications can be assigned to a polytomy in a reconciled non-binary tree.
If duplications are inferred on two or more temporary nodes in the optimal binary resolution

35
 Chapter 4. Theory

(a) (b)

(c) (d)

(e)

Figure 4.5: (a) A binary species tree. (b) A non-binary gene tree with genes sampled from (a). (c) Binary resolution of
gene tree (b), yielding a binary tree with three duplications and three losses. (d) Gene tree (b) reconciled with
species tree (a), yielding a non-binary tree with three duplications and four losses. (e) Gene tree (b) following
rearrangement. The polytomy has been resolved and the weak edge has been rearranged to eliminate a duplication.

36
Chapter 4. Theory

(a) (b)

(c)

Figure 4.6: (a) A non-binary gene tree. (b) An optimal, binary resolution of gene tree (a) reconciled with species tree in
Figure 4.5. (c) The reconciled non-binary, gene tree. The resulting tree has a polytomy with two duplications.

of a polytomy, the polytomy will be assigned multiple duplications when these nodes are
removed from the tree. For example, two duplications are assigned to the polytomy in the
reconciled, non-binary gene tree in Figure 4.6. This differs from standard reconciliation,
where every node has at most one duplication.

Rooting a non-binary gene tree

Notung can be used to infer the root of an unrooted tree by identifying the root that requires
the fewest duplications and losses. In Rooting mode, when the tree is binary, each edge is
assigned a root score; i.e., the DL Score of the tree when rooted on that edge. When the
gene tree is non-binary, it is also possible to root the tree on a polytomy, as shown in Figure
4.7. Placing a polytomy at the root of the tree implies that one of the edges in the true
binary resolution of the polytomy is the true root.

To calculate root scores, Notung roots the tree on each edge and polytomy in turn. For
each root, the rearrangement algorithm is applied to ensure that each polytomy is replaced
by an optimal binary resolution. The DL Score of the resulting tree is used as the root
score for that rooting. Note that it is necessary to optimize the binary resolutions separately
for each root because the DL Score depends on the location of the root. After all edges

37
 Chapter 4. Theory

(a) (b)

(c)

Figure 4.7: (a) An unrooted, non-binary gene tree. (b) The rooted, binary resolution of (a) with the lowest D/L Score.
Rooting the tree on any other edge would entail more duplications and losses. (c) When reconciled with species
tree in Figure 4.5(a), the polytomy in (a) is the root with minimum cost.

and polytomies have been scored, the original tree is reported to the user with edges and
polytomies annotated with root scores.

Note that in Reconciliation and Rooting modes, binary resolutions are used to infer du-
plications and losses, but the structure of the final, output tree is unchanged. In the Re-
arrangement and Resolve modes, Notung uses duplication-loss parsimony to transform the
non-binary input tree into a binary gene tree. Resolve mode is analogous to the reconcili-
ation method described here, with the exception that the final step of removing the added
nodes and edges is not performed. The result is a reconciled binary tree that is optimal with
respect to duplication-loss parsimony. For the example in Figure 4.5, the Resolve function
would return the tree in (c). As there may be more than one optimal resolution, Notung
presents the different histories that result in the optimal tree. See Chapter 8 - Resolve Mode
on page 81 for more information.

In Rearrangement mode, the rearrangement algorithm is applied not only to edges added
to the tree in the resolution of polytomies, but to all edges with an edge weight below the
edge weight threshold. The result is a reconciled, binary tree in which weak edges have
been rearranged to minimize the DL Score. Figure 4.5(e) shows the rearrangement of the
non-binary gene tree in (b), assuming an edge weight threshold of 90.

38
Chapter 4. Theory

4.3 Practical considerations

When the species tree is binary, our transfer algorithm has O(|VG ||VS |2 ) complexity when
finding a single optimal reconciliation. When the species tree is non-binary, the complexity is
O(|VG |(|VS |+nk 2k )2 (hS +k)), where k is the size of the largest polytomy in the species tree, nk
is the number of polytomies and hs is the height of the species tree. However, there are often
multiples optimal solutions, and sometimes many, each requiring an additional O(|VG |) for
binary species trees and O(|VG |(hS +k)) for non-binary species trees. Each candidate optimal
reconciliation must be checked for temporal feasibility, further increasing the computational
cost. The running time will increase with the number of optimal solutions. Typically, there
are more optimal solutions when a large number of transfers are inferred.

While Notung will run faster if you only look at a single solution, the user should be aware
that a single solution may be only one of several mathematically equivalent hypotheses. The
solution arbitrarily selected by Notung as the first one to present to you may not represent
the best hypothesis.

The complexity for algorithms with non-binary species trees will also increase with the size
of the largest polytomy and the number of polytomies. In practical data analyses, k is likely
to be small. Recent genome-scale analyses of ILS have focused on species trees with k = 3
(e.g., [8, 23]). In general, event inference will not yield informative results when the species
tree is highly unresolved.

Because of the increased computational complexity of reconciliation with transfers, Notung’s

default settings with the DTL model differ somewhat from the DL model defaults. In these
cases, the DTL default is to provide less comprehensive information in order to avoid very
long running times. A more comprehensive analysis can be obtained by changing the param-
eter settings. Some particularly computation intensive analyses can only be run from the
command line.

In the GUI, a warning message pops up in Reconcile mode if the number of candidate
optimal reconciliations for gene tree exceeds 10,000. The user can choose to proceed with
the reconciliation, which may require an hour or more to complete. Alternatively, the user
can choose to cancel the reconciliation in the GUI, and reconcile the tree from the command
line instead.

39
Chapter 5

Reconciliation Mode

In Reconciliation mode, Notung compares a gene tree with a species tree to infer gene du-
plications, transfers and/or losses. Notung will display a reconciled tree in the tree panel
with the inferred events indicated on the tree. The Event Score of a reconciled tree will be
displayed in the lower left corner of the screen (see Figure 5.1(b)).

Once a gene tree has been reconciled, Notung can detect homology relationships, described
in Section 5.6. Notung can also determine lower and upper bounds on the time of each
duplication and co-divergence, where bounds are represented in terms of internal nodes in
the species tree; i.e., relative to speciation events. The upper bound on the time of duplication
is the most recent species in which the duplication was not present. The lower bound is the
oldest species in which the duplication must have been present. This information, along with
statistics on transfers and losses, can be viewed in a pop-up window by selecting “Event
Summary” or “Parsable Statistics” from the “About This Tree” menu. Transfers,
duplications and bounds in this window are identified by internal node names. For losses,
each node in the species tree is listed, followed by the number of losses associated with that
taxon. A table of species is also included, where rows represent each node in the species tree,
and values in a row indicate the number of duplications, transfers from, transfers to, and
losses associated with that taxon.

Notung can infer reconciliations with or without transfer events (e.g., horizontal gene trans-
fer), along with duplications, losses, and co-divergences. When the Infer transfers option
in the Reconciliation panel is selected, Notung will allow the user to specify a transfer
cost when the Edit Values button on the bottom right corner is clicked. Then operations
including reconciliation and rooting analysis will be performed to minimize the event cost
using the DTL-model, in other words, with possible transfer events taken into consideration.
Please be aware that once the DTL-model is used, there are several aspects of the program
the user should pay attention to, namely, (1) temporal feasibility, (2) multiple optimal solu-

40
Chapter 5. Reconciliation Mode

tion, and (3) rearrange and resolve with transfers. These new features will be covered in the
following section. For theoretical insights into reconciliation with transfers, refer to Chapter
4.

5.1 Labeling and Pruning Gene and Species Trees

Notung’s reconciliation algorithms require that both trees be rooted and that each leaf in
the gene tree be associated with a leaf in the species tree. For this to work, we require a way
of representing the association between gene and species leaf nodes in the program’s input.
In Notung, the label of each leaf in the gene tree must include, as a substring, part or all
of the label of the associated leaf species. This substring is called the species tag. Notung
recognizes several different species tag formats. See Appendix A.4 - Specifying the Species
Associated with Each Gene on page 129 for further information on gene labels.

The reconciliation algorithm assumes that each gene is associated with a leaf in the species
tree. This is a many-to-one relationship; a species may be associated with one or more than
one gene. But, what about species that are not associated with any gene? There are two
strategies for dealing with this situation: These species can be removed, or pruned, from the
tree in a preprocessing step. In this case, no penalty is incurred for observing a species that
does not possess a member of the gene family. Alternatively, these species can be retained,
in which case losses will be incurred during the process of reconciliation.

Which strategy is more appropriate for a particular analysis depends on how the data were
sampled. If the sequences of interest were derived from species with good quality, whole
genome sequence, the failure to observe a gene family member in one of these species is
suggestive of a true gene loss. In this case, species that are not associated with any gene
should be retained, so that losses are incurred during reconciliation. However, if the species
represented in the gene tree do not have reliable whole genome sequences, then pruning the
species tree prior to reconciliation may be appropriate.

With this in mind, Notung offers two approaches to pruning the species tree. By default,
Notung finds the lowest common ancestor (LCA) of the species that are associated with
genes in the gene tree, and only prunes those species that diverged prior to the LCA. All
species within the clade rooted at the LCA are retained. This strategy is designed for high
quality data sets, where the absence of a gene is indicative of gene loss. Alternatively, the
user may opt to prune all species that are not associated with at least one gene. In the GUI,
this is achieved by selecting the “prune” check-box in the Reconciliation panel. Please see
Chapter 12.2 - Prune on page 104 for command-line options regarding species tree pruning.
This strategy is appropriate when the failure to observe a gene in a particular species may be
due to sparse data or experimental error. This approach is also appropriate in cases where

41
 Chapter 5. Reconciliation Mode

the species tree was constructed for some other purpose and contains species that are not of
interest in the current analysis.

Unlike species nodes, which may be associated with zero, one, or more than one gene, each
gene must be associated with exactly one species. If Notung cannot identify the species
associated with one or more leaves in the gene tree, it will print a diagnostic message and
the reconciliation process will be terminated. This could indicate a species tag formatting
problem. However, in some cases, the leaf set in the gene tree is a superset of the gene
family of interest in the current study. For example, the user may have started with a broad
analysis, but now wishes to restrict the study to a smaller set of species. In this case, it is
convenient to be able to prune from the gene tree, those leaves that are not associated with a
species in the restricted set. Notung’s command line interface offers a utility for this purpose
that can be used to obtain a pruned gene tree in a preprocessing step. See Section 12.2 -
Prune on page 104 for more information.

5.2 Reconciling Binary Trees with Transfers

In Notung, the user can reconcile gene trees with species trees to infer transfer events (e.g.
horizontal/lateral gene transfer), along with duplications, losses, and co-divergences. We
call the model with transfer events the DTL-model, or DTLI-model if the species tree is
non-binary. Currently, analyses performed on non-binary gene trees with the DTL model
are restricted to the command-line interface. See Section 12.8 - Non-binary gene trees with
the DTL model on page 117 for more information. Analyses with binary gene trees can be
performed either in the GUI or on the command-line. To enable the DTL-model or DTLI-
model, the user needs to select the “Infer Transfers” option in the Reconciliation panel. By
doing that, the user can specify a cost value for a transfer event. The event costs can be set
by clicking the “Edit Values” button on the bottom right corner. The default cost values
are 1.5 for Duplication, 3 for Transfer, and 1 for Loss. Once “Infer Transfers” is selected,
operations like reconciliation and rooting will be performed to minimize the events cost using
the DTL-model, in other words, with possible transfer events taken into consideration. For
a more detailed description of the DTL-model, please refer to Chapter 4.

After performing a reconciliation with the DTL-model, Notung will display a binary gene tree
with events labeled on it. Inferred Duplication and Co-Divergence events will be indicated
by red ‘D’ and ‘cD’ above corresponding internal nodes respectively, and gray external nodes
(leaf nodes) with ‘ LOSS’ in their name labels indicate Loss events. For an inferred transfer
event, the gene tree edge associated with that transfer (transfer edge) will be labeled in
yellow, and a yellow triangle, with the letter ‘T’, will be displayed on the middle of that
transfer edge (see Figure 5.3(a)). The yellow triangle on each transfer edge is clickable. When
clicked, the species donor and species recipient of that transfer will be displayed on top of that

42
Chapter 5. Reconciliation Mode

transfer edge, in the format of ‘from donor species to recipient species.’ Alternatively, the
user can also view the species information by selecting “Display Internal Node Species
Names” and “Display Leaf Node Species Names” under “Display Options” on the
menu bar. Information about all inferred transfers, along with other statistics related to
this reconciliation can be viewed in a pop-up window by selecting “Event Summary” or
“Parsable Statistics”from the “About This Tree” menu.

5.2.1 Temporal Feasibility

In reconciliation with the DTL-model, it is important to beware of the temporal feasibility

of a solution, as each inferred transfer will impose temporal constraints. For example, a
transfer implies both its donor and recipient species co-existed in a certain temporal interval.
When there are multiple transfer events inferred, it is possible that the multiple transfers
impose conflicting temporal constraints. We call such problematic reconciliation solutions
(temporally) infeasible solutions. Thus, it is crucial to check the temporal feasibility when
transfer is included in the reconciliation model. For theoretical background about temporal
feasibility, please see Chapter 4.1.1 - Temporal Feasibility on page 28.

By default, Notung checks the temporal feasibility whenever an event history is performed
with the DTL-model, to ensure that any solution that is presented to the user is temporally
feasible. If all the solutions with minimal Event Cost are temporally infeasible, a dialog box
will pop up (GUI mode) or a warning message will appear on the screen (command line
mode), warning the user of the issue. In that case, no reconciliation solution will be shown
in the GUI, and no reconciliation will be output in the command line mode. Please be aware
of the fact that all minimum-cost solutions are infeasible does not mean that a temporally
feasible solution with that particular set of event costs does not exist. A sub-minimum-cost
solution may be temporally feasible.

5.2.2 Multiple Solutions

When transfer is included in the reconciliation model, it is possible for a reconciliation to

have more than one minimum-cost solution with a given set of event costs. For theoretical
background about multiple solutions, please see Chapter 4.1.1 - Multiple Solutions on page
27. In the GUI, after a reconciliation is performed with the DTL or DTLI-model, the number
of feasible solutions will be displayed in the status bar on the bottom of the GUI, along with
the inferred number of events and total cost. The number of feasible solutions will also appear
in the pop-up window if the user selects “Event Summary” or “Parsable Statistics” from
the “About This Tree” menu. In the GUI, a green circle on a node in the gene tree panel
indicates the existence of multiple solutions within the subtree rooted at that particular node.

43
 Chapter 5. Reconciliation Mode

It is easy to browse through all feasible solutions by clicking the corresponding green circle(s)
repeatedly (see Figure 5.2).

5.2.3 Rearrange and Resolve with Transfers

Currently rearranging and resolving with transfers (DTL-model) are not supported in the No-
tung GUI. If the optional Infer Transfers from the Reconciliation panel is selected, clicking
Perform Rearrangement from the Rearrange panel, or clicking Resolve Polytomies
from the Resolve panel, will bring up a pop-up window with warning messages, reminding
the user that rearranging with transfer is only supported via the command-line.

5.3 Reconciling Non-Binary Trees

Notung can reconcile binary gene trees with non-binary species trees, as well as non-binary
gene trees with binary species trees. The differences between these functions and traditional
reconciliation of binary gene trees with binary species trees are summarized briefly here. For
a more detailed discussion of reconciliation with non-binary trees, see Chapter 4.2 - Non-
Binary Trees on page 29. Note that orthologs, paralogs, and xenologs can only be inferred
on binary gene trees reconciled with binary species trees.

Reconciling a binary gene tree with a non-binary species tree results in a binary
gene tree with duplications, transfers, and losses added. Notung distinguishes between cases
in which disagreement can only be explained by a gene duplication or transfers and cases in
which the disagreement may be due to co-divergence. When reconciling a gene tree with a
non-binary species tree, duplications appear in the tree as small red squares with red D’s,
and transfers appear as yellow edges with yellow triangles and T’s, while co-divergences,
other than binary speciation events, are small pink squares with pink cD’s (see Figure 5.2
and Figure 5.3).

If two or more orthologous genes are missing from species that are children of the same
polytomy, then it is more parsimonious to infer a loss of the common ancestor of those genes.
We refer to such losses as polytomy losses. For example, in Figure 5.1, members of the
hypothetical Y gene family are missing from two species, bandicoot and opossum. These
species are children of the same polytomy in the species tree in Figure 4.1. Notung infers
a single loss, labeled with the names of species from which the gene is absent, as well as
the label of the corresponding polytomy in the species tree. By default, polytomy losses are
labeled with the species that lack the gene. However, if a polytomy loss is associated with
many sibling species, the default display can produce very long labels. Users can instead opt

44
Chapter 5. Reconciliation Mode

to label polytomy losses with the number of species in which the loss occurred, as well as the
label and the total number of children of the polytomy, illustrated in Figure 5.3(b).

Reconciling a non-binary gene tree with a binary species tree results in a non-binary,
reconciled gene tree. A reconciled, binary gene tree can be obtained by using the Resolve
function (see Chapter 8 - Resolve Mode on page 81).

NOTE : In this case, transfers can currently be inferred only via the command-line
interface, as discussed in Section 12.8.

Reconciliation of a non-binary gene tree with a binary species tree differs from binary rec-
onciliation in two important ways. First, a polytomy in a non-binary gene tree may be
annotated with more than one event. For example, the reconciled non-binary gene tree in
Figure 5.4(a) has a polytomy annotated with two duplications and a loss. Recall that a
gene tree polytomy is an indication that although its children evolved by successive binary
divergences, the order in which the taxa diverged is unknown. Since this binary branching
pattern is unknown, the relative order of events with respect to those divergences cannot
not be determined, either. The polytomy in Figure 5.4(a) communicates that at least two
duplications and one loss occurred in the subtree descending from the polytomy, but the
exact timing of those events is unknown. See Chapter 4.2 - Non-Binary Trees on page 29 for
a detailed explanation of duplications and losses in reconciled non-binary gene trees.

Second, there may be several alternate hypotheses for the reconciliation of a non-binary gene
tree. Since the true binary branching pattern of a polytomy is unknown, Notung infers
events for all binary resolutions with the minimal Event Score. If there is more than one
optimal binary resolution, multiple reconciliations will result. Notung addresses this issue by
presenting all alternate event histories to the user. Each event history represents a different
combination of duplications and losses that could result in the same minimal DL Score.
Initially, Notung arbitrarily selects one event history to present in the tree panel. The other
optimal histories may be viewed using the drop-down menu labeled “Select an optimal
event history,” as shown in Figure 5.4. This menu gives a list of up to 50 optimal event
histories. If there are more than 50 optimal event histories, they can be generated using the
Command Line Interface (see Chapter 12 - Command Line Options and Batch Processing
on page 98). For a more detailed discussion of alternate event histories, see Chapter 7 -
Rearrange Mode on page 73.

5.4 Using Reconciliation Commands

To regulate reconciliation options and behaviors:

45
 Chapter 5. Reconciliation Mode

A number of different options affecting the reconciliation algorithm and results are now
available, including the ability to infer transfers.

• Click the appropriate checkbox in the Reconciliation task panel.

√
Infer Transfers (default: OFF) When this box is checked, transfers are included
in the optimization criterion. To infer transfers with duplications, check this box.
NOTE: Duplications are always included in the optimization. To consider
only transfer events as a source of tree disagreement, you may simply set
the cost of duplication to a very high value relative to the cost of transfers.
√
Prune (default: OFF) When this box is checked, the species tree is “pruned” to
remove taxa that are not represented in the gene tree.
NOTE: By default, Notung does not prune the species tree. This assumes
that genes missing in species are due to true loss events and not incomplete
taxon sampling. As a result, spurious losses may be inferred if the species
tree contains taxa that were insufficiently sampled for the genes under
consideration.

To change the parameter values:

1. Click the “Edit Values” button. A dialog box appears.

2. Enter the appropriate values in the text field, and then click “Apply Changes.”

NOTE: This will change the value settings only for the tree that is currently
selected. Each history state saves the parameter values used at that state; when
moving through the history, parameter values may change depending on the state
and tree viewed. For more information on history states, see Chapter 9 - History
on page 85.

NOTE: Transfer costs can only be edited if the “Infer Transfers” option is se-
lected.

To reconcile a gene tree with a species tree:

1. Click the Reconciliation tab to enter Reconciliation mode.

2. Click the “Reconcile/Rereconcile” button. A dialog box appears.

46
Chapter 5. Reconciliation Mode

3. In the dialog box, select the correct species tree in the drop-down menu.

4. Check that Notung correctly identified the species naming convention used in the gene
tree. The available settings are:

• Prefix of the gene label (i.e., SPECIESGENE)

• Postfix (underscore required) of the gene label (i.e., GENE SPECIES)
• NHX: species label is stored in the NHX comment field in the gene tree file. (i.e.,
GENENAME[&&NHX:S=SPECIES])

If the convention selected by Notung is not the naming convention used in

the gene tree, change it by selecting the appropriate radio button. See Ap-
pendix A.4 - Specifying the Species Associated with Each Gene on page 129
for details about species tag specifications.

NOTE: The Prefix and Postfix formats require species names to be embedded
in the gene names. NHX Species Tag format embeds the species information
in a Newick comment field. When this format is used, the information will
not appear on the screen unless the “Display Leaf Node Species Names”
option in the Display Options menu is selected (see Chapter 11.1 - Display
Options on page 93).

5. In the dialog box, click “Reconcile.”

The reconciled tree appears in the Tree Panel (see Figure 5.1(b)). Internal nodes of the tree
represent events that caused a bifurcation. In a reconciled gene tree, the inferred events
appear as:

Speciation Node is a small blue square with no text.

Duplication Node is a large red square, displayed with a red ’D’. In a non-binary gene tree,
the number of duplications associated with a polytomy will also be shown with a red
D (e.g., Figure 5.4(a)).

Co-divergence Node is a pink square with the letters “cD”; only inferred and shown if the
species tree is non-binary.

Transfer The edge is highlighted in yellow. A yellow triangle, with a yellow ’T’, appears
halfway down the edge. Transfer events are represented on tree branches to indicate
the bifurcation that resulted from the transfer, as well as the recipient of the transfer.
When the triangle is clicked, the donor and recipient taxa of that event are displayed.

47
 Chapter 5. Reconciliation Mode

Loss The leaf node appears in light gray type is labeled “<species>*LOST”, indicating the
species in which the loss occurred.

A status bar at the bottom of the program window (Figure 5.3) provides additional informa-
tion about the reconciliation. This includes the total cost of the reconciliation (Event Score),
as well as a summary of the number of inferred events, the total number of optimal, feasible
solutions, and the species tree used for the reconciliation. In addition to displaying events
visually, Notung-2.9can also generate a detailed report on event histories in a text format.
See Section 5.5 - Additional Reports on page 51 for more information.

The species associated with the ancestral nodes in the reconciled gene tree can also be
graphically displayed on the tree. The user can view this information by selecting “Display
Internal Node Species Names” and “Display Leaf Node Species Names” under
“Display Options” on the menu bar.

To hide losses/duplications/transfers/co-divergences:

The duplication, co-divergence, or transfer marks or loss nodes can be hidden to avoid a
cluttered image.

• Click the appropriate checkbox in the Reconciliation task panel.

√
Display Loss Nodes (default: ON) When this box is checked, the inferred lost
nodes appear in the tree. To hide inferred loss nodes, uncheck the box.
NOTE: When you uncheck “Display Loss Nodes,” Notung will reset the image
so that the whole tree fits in the tree panel.
√
Display Duplications (default: ON) When this box is checked, duplications are
indicated on internal nodes by red squares and D’s. To make the red D’s disappear,
uncheck the box.
√
Display Co-Divergences (default: ON) When this box is checked, co-divergences
are indicated on internal nodes by pink squares and cD’s. To make the pink cD’s
disappear, uncheck the box.
√
Display Transfers (default: ON) When this box is checked, transfers are indi-
cated on highlighted edges by yellow triangles and T’s. To make the yellow T’s
disappear, uncheck the box.

Options that are not currently available are displayed in gray type to indicate that they are
disabled. In particular, the above options will be grayed out if no reconciliation has been
performed. The “Display Co-Divergences” option will also be displayed in gray if the
gene tree was reconciled with a binary species tree, and the “Display Transfers” option
will be grayed out if the reconciliation did not consider transfers.

48
Chapter 5. Reconciliation Mode

To display the number of species in polytomy losses:

By default, polytomy losses are labeled with the names of the species from which they are
absent.

1. Go to the “Display Options menu.”

2. Click the “Use Species Names in Polytomy Losses” box.

This causes polytomy losses to be labeled with the number of children of the
polytomy lost, the total number of children of the polytomy, and the name of the
polytomy in which these losses occurred.

To display the donor and recipient of transfers:

The donor and recipient species of transfers in the gene tree can be displayed by clicking on
any transfer triangle in the gene tree. Clicking a triangle again will hide this information. If
the gene tree is non-binary, or if transfers were inferred, there may be more than one optimal
reconciliation.

To view alternate optimal event histories for non-binary gene trees:

If more than one optimal event history exists for a reconciled, non-binary gene tree, the drop
down menu, “Select an optimal event history,” will be enabled.

• From the drop-down menu, select an alternate event history.

The tree panel will now show a new tree corresponding to the selected alternate history.

If there is only one optimal history or if the tree has not been reconciled, the drop down
menu will be grayed out.

NOTE : When a user selects a different alternate event history from the “Select an
optimal event history” list, Notung rebuilds the tree from data saved at the time
of reconciliation. Any manual swaps made to a previously viewed event history
will be lost. Therefore, if you wish to save information after a manual swap, you
must save your tree. See Chapter 3.3 - Opening and Saving Trees on page 15 for
more information.

49
 Chapter 5. Reconciliation Mode

To view alternate optimal event histories when inferring transfers:

If more than one optimal event history exists for a tree reconciled with transfers, green circles
will appear around nodes in the gene tree.

• On the gene tree, click on the nodes circled in green. The tree will now show an
alternative optimal solution.

Clicking on a node will loop through the different alternate histories affecting areas below
the node. To loop through all optimal histories, click on the green circle(s) closest to the
root of the tree.

NOTE : When a user moves to a different mode, Notung rebuilds the tree from
data saved at the time of reconciliation. Any manual click to change the solution
will be lost. Therefore, if you wish to save information after a manual change,
you must save your tree. See Chapter 3.3 - Opening and Saving Trees on page 15
for more information.

To undo the reconciliation:

• Click the “Unreconcile” button. The most recent reconciliation will be undone. This
option is grayed out if the gene tree has not been reconciled.

To display ancestral associations:

1. Click “Display Options” from the menu bar.

2. Select “Display Internal Node Species Names”

To display the species tree:

1. Click the “Show pruned species tree” button. A dialog box appears.

2. Enter a title in the text field and click “OK.”

This option is grayed out if the gene tree has not been reconciled.

50
Chapter 5. Reconciliation Mode

5.5 Additional Reports

5.5.1 Event Summary

Notung can provide detailed information on the inferred event histories in text format. These
reports include information on the timing of duplications (upper and lower bounds on the
time of duplication), transfers (the species from which the gene originated and the which
received the gene), and losses (the species in which the loss occurred), as well as timing of
co-divergences with the gene tree. Transfers, insertions, duplications, co-divergences, and
bounds are identified by internal node names in these reports. For losses, each node in the
species tree is listed, followed by the number of losses associated with that species. Finally, a
table of species is included, where rows represent each node in the species tree, and values in
a row indicate the number of duplications, transfers from, transfers to, and losses associated
with that species. The files, and the information provided, are described in detail below.

When viewing this information, it may be helpful to display the names of internal
nodes in the gene tree, as well as the species associated with these nodes. To do so,
from the “Display Options” menu (see Chapter 11.1 - Display Options on page
93), turn on “Display Internal Node Names” and and “Display Internal
Node Species Names”.

The Event Score of the reconciled tree appears at the top of the window.

This is followed by a table on duplication bounds, described in three columns. The internal
node representing the duplication is listed in the first column. Duplication bounds are de-
scribed in the next two columns: lower and upper bounds, respectively, relative to the species
tree, expressed as node names in the species tree. The total number of duplications appears
below this table.

If the species tree is non-binary, co-divergence bounds, if any, are described in the three
columns below duplication bounds. The left column gives the internal node in the gene tree
where the co-divergence occurred. The next two columns provide the lower and upper bounds,
respectively, on the species tree node in which the event (speciation or allelic divergence) may
have occurred. The total number of co-divergences is listed below this table.

If included in the set of events, information on transfers will appear in a table below these
bounds. The first two columns list the nodes in the gene tree where the transfer occurred,
representing where the transfer originated and ended, respectively. The timing of the transfer
event is listed in the two subsequent columns, respectively: the species from which the transfer

51
 Chapter 5. Reconciliation Mode

came (donor), and the species which accepted the transfer (recipient). The total number of
insertions is listed below this table.

Information on losses follows. Rather than reporting each loss separately, losses are summa-
rized according to the taxa in which they occurred. This table has two columns. The first
column is a list of all the nodes in the species tree; the second column gives the number of
inferred gene losses that occurred in that species. Polytomy losses are assigned to the cor-
responding polytomy, rather than the individual species which lack the gene. For example,
the polytomy loss in Figure 5.3 is reported as a single loss in Metatheria. The total number
of losses is listed below this table.

Following information on each separate event type, there is a table providing total event
counts per species. The first column lists all the nodes in the species tree. Subsequent
columns give the number of times each species is associated with an events: duplication,
co-divergence, transfer donor, transfer recipient, and loss. The total number of each event is
listed, again, below this table.

To show information on events:

• Select the “Event Summary” menu item from the “About This Tree” menu. A
new window will appear.

This option is grayed out if the gene tree has not been reconciled.

Parsable Statistics

This file is designed to contain much the same information as the “Event Summary” file,
but in a format that is easily parsed by a computer. The first line is a tab-delimited set of
values, with the header on the line below.

This line includes: the number of duplications (nD), co-divergences (nCD), transfers (nT),
and losses (nL); the size of the gene tree without losses in terms of the number of leaves
(|L(G)|) and total nodes (|G|), as well as the size of the associated species tree in terms
of total nodes (|S|); the heights of the gene (h(G)) and species (h(S)) trees in terms of the
longest path from root to leaves; whether the root of the gene tree is mapped to the root
of the species tree (isRoot), the species to which the gene tree root is mapped (root(G)),
and the root of the species tree (root(S)); the range of edge weights (minEW, maxEW); the
number of optimal roots (Roots) and reconciliations, both candidate (Cand) and feasible
(Feas); the cost of duplication (cD), co-divergence (cCD), transfer (cT), and loss (cL); the
size of the largest polytomy (largest Poly) and the number of polytomies (numPoly).

52
Chapter 5. Reconciliation Mode

The remainder of the file provides the same information as “Event Summary.” The lines
are tab-delimited and start with #<x> to distinguish the type of information displayed. <x>
can be any of D, CD, T, L, S, or R. #D demarcates information on duplications, #CD
demarcates information on co-divergences if the species tree is non-binary, #T demarcates
information on transfer, and #L demarcates information on losses tallied per species. Lines
in the final summary table are demarcated with #S. The final information in this file begins
with #R. It provides the species edge, defined by the species node below the edge, on which
the gene tree is rooted.

To view tree statistics and event details in an easily parsable format:

• Select the “Parsable Statistics” menu item from the “About This Tree” menu. A
new window will appear.

This option is grayed out if the gene tree has not been reconciled.

5.6 Inferring Orthologs, Paralogs, and Xenologs

Many types of comparative analysis require the identification of different types of homologs.
Gene tree - species tree reconciliation is the most reliable and robust approach to distinguish-
ing orthologs from paralogs. Reconciliation is also increasingly used to infer horizontal gene
transfer events, and thus, can be used to identify xenologs.

Notung utilizes information from a reconciled gene tree to infer the homology relationships
between all pairs of leaves in the gene tree. Currently, homology inference in Notung is only
supported for the case where both the gene and species trees are binary. An extension to
non-binary trees will be included in a future version.

Notung’s homology classification functions can be accessed from the GUI or the command
line. The Notung GUI contains an interactive feature in the Reconciliation task panel that
allows the user to investigate homology through a point and click interface (Section 5.6.2).
In addition, Notung can output a table summarizing the homology relationships between all
pairs of leaves in a reconciled gene tree (Section 5.6.3). This homology table can be saved
from either the GUI or the command-line interface. The batch processing feature of the
command-line interface also allows for automated generation of homology tables for a large
number of gene trees. See Chapter 12 - Command Line Options and Batch Processing for
more information.

53
 Chapter 5. Reconciliation Mode

Homology Classification in Notung. Notung uses the cenancestor-based homology ter-

minology introduced by Fitch [12]:

Orthologs: “two homologous characters whose common ancestor lies in the cenancestor1
of the taxa from which the two sequences were obtained.”

Paralogs: “two homologous characters arising from a duplication of the gene for that
character.”

Xenologs: “two homologous characters whose history, since their common ancestor,
involves an interspecies (horizontal) transfer of the genetic material for at least one of
those characters.”

These definitions differ from the more commonly used event-based definitions, where a pair
of genes are:

Orthologs: two homologous characters that diverged via a speciation event.

Paralogs: two homologous characters that diverged by a duplication event.

However, under the duplication-loss model (Infer Transfers OFF), the cenancestor-based
definitions and the event-based definitions are identical.

In Section 5.6.1 - Homology Terminology, we discuss the definitions that Notung uses for ho-
mology classification, with particular attention to homology classification under the duplication-
transfer-loss model (Infer Transfers ON). If your goal is to distinguish between orthologs
and paralogs in gene families in which horizontal gene transfer does not occur, you may wish
to skip this section and go on to Section 5.6.2.

5.6.1 Homology Terminology

Notung employs the cenancestor-based definitions, stated above, for homology classification
because they present no ambiguity when distinguishing xenologs from orthologs. In a gene
family in which gene duplication is the only mechanism by which new copies of a gene family
arise, the commonly-used event-based definitions are unambiguous: Every pair of genes in a
reconciled tree can be assigned to exactly one of these two classes: orthologs or paralogs.
1
The cenancestor of a pair of genes is the lowest common ancestor (LCA) of the species containing that
pair.

54
Chapter 5. Reconciliation Mode

However, in a gene family with horizontal gene transfer, the event-based definitions of or-
tholog and paralog can overlap with the definition of xenolog. Consider, for example, a gene
family in which a speciation event gave rise to two gene lineages, one of which subsequently
experienced a horizontal transfer. Since their shared history includes a horizontal transfer,
they are xenologs. However, according to the event-based definition of orthologs, the result-
ing gene pair are also orthologs. For example, in Figure 5.5, the speciation event at node
2 gave rise to genes gbX and gZ . These genes are xenologs because their history, since their
LCA, involves a horizontal transfer. However, according to the event-based definition, they
are orthologs because the event that caused the divergence at node 2 was a speciation.

The cenancestor-based definitions remove this ambiguity. Under that definition, genes gbX
and gZ are not orthologs, because their LCA (node 2) is not in their cenancestor (species β).
Note that the cenancestor-based and event-based definitions are equivalent in a pair of genes
that evolved by duplication and loss only (either no transfers inferred or Infer Transfers
OFF). In the absence of transfers, the LCA of two genes lies in their cenancestor if and only
if there was a speciation event at the LCA.

As this example illustrates, Fitch’s cenancestor-based definitions result in a precise and non-
overlapping classification of xenologs and orthologs. However, a consequence of these defini-
tions is that xenologs encompass a broad set of gene pairs with highly diverse properties and
relationships.

First, a pair of xenologs may have arisen from any event — transfer, speciation, or duplication
— at the LCA of the gene pair. For example, in Figure 5.5, the three gene pairs (b gX , gY ),
0
gX , gY ), and (b
(b gX , gZ ) are each xenologous. The event at the LCA of genes gbX and gY is a
transfer (node 1). The event at the LCA of genes gbX and gZ , is a speciation (node 2). The
event at the LCA of genes gbX and gY0 is a duplication (node 4).

Second, xenologous pairs, even in the same species, may vary greatly in how closely they are
related; the species associated with a pair’s LCA may post-date, pre-date, or be the same
as their cenancestor. Figure 5.5 illustrates these three possibilities. Genes gbX and gY are
xenologs whose LCA (node 1) post-dates the cenancestor (species β). For xenologs gbX and
gW , their LCA (node 6) pre-dates the cenancestor (species γ). And in a third example, for
xenologs gbX and gV , their LCA (node 7) is in their cenancestor (species α).

Xenolog Classification. In order to achieve a more informative classification within the

broad class of xenologs, we have developed xenolog subtypes that capture distinctions in
how horizontally transferred genes are related [4]. Under the duplication-transfer-loss (DTL)
model (Infer Transfers ON), Notung infers orthologs, paralogs, and xenologs, and further
labels xenologs according to this subclassification. Here, we summarize these xenolog sub-
classes for a gene family in which a single horizontal transfer has occurred. The inference of

55
 Chapter 5. Reconciliation Mode

subclasses for a gene family history with an arbitrary number of transfers and duplications is
described in detail in Darby et al. [4]. The theoretical properties of the classification, pseu-
docode, and examples of its application to real data are also described in that publication.

Our classification groups xenologous gene pairs into five disjoint subclasses (Table 5.1). Recall
that a pair of genes are xenologous, if they are related through a transfer event at some point
in their shared history. In other words, the path between the two genes will pass through
the transfer. Let gb and g be genes such that there is a transfer on their path and gene gb is a
descendant of the transfer recipient. We say that gene g is a xenolog of gb and treat gb as the
reference.

P Paralogs
O Orthologs
PX Primary xenologs
SDX Sibling donor xenologs
Xenologs

Class SRX Sibling recipient xenologs

OX Outgroup xenologs
IX Incomparable xenologs
-X0 -autoxenolog
Suffix
-Xp -paraxenolog
Table 5.1: Abbreviations for different homology classes.

Gene g is a primary xenolog (PX) of gb if the divergence at their LCA in the gene tree was
caused by a transfer event. For example, gY is a primary xenolog of gbX (pair (b gX , gY ) is
designated PX) because their LCA (node 1) was the transfer donor. Otherwise, we classify
the xenologs based on the species containing g. If g is in a species that is more closely related
to the donor species than the recipient species, then g is a sibling donor xenolog (SDX) of gb.
Similarly, if g is in a species that is more closely related to the recipient species than the donor
species, then g is a sibling recipient xenolog (SRX) of gb. Otherwise, g is in a species that
diverged before the transfer cenancestor (the LCA of the donor and recipient species) and
is equally related to the donor and recipient species; in this case, g is an outgroup xenolog
(OX) of gb. In Figure 5.5, gZ is a sibling donor xenolog of gbX (pair (b gX , gZ ) is SDX); gW
is a sibling recipient xenolog of gbX (pair (b gX , gW ) is SRX); and gV is an outgroup xenolog
of gbX (pair (bgX , gV ) is OX). Sometimes, g will be in the same species as gbX . We add the
extra annotation of autoxenolog (X0 ) to note these special cases. For example, gX is a sibling
recipient autoxenolog of gbX (pair (b gX , gX ) is SRX0 ).

While Fitch’s cenancestor-based definitions distinguish between orthologs and xenologs, it is

still possible for a pair of genes to be both xenologs and paralogs, i.e., if g is a xenolog of gb
and the genes diverged at a duplication node. We introduce the term paraxenolog (XP ) to
account for such cases. A paraxenolog can also be a member of one of the standard classes
(above) and will be assigned that class. For example, in Figure 5.5, gene gY0 is a sibling donor

56
Chapter 5. Reconciliation Mode

gX , gY0 ) is SDXP ) since there was a duplication at their LCA (node

paraxenolog of gbX (pair (b
4).

Note that because of the directional nature of transfers, the classification of xenolog pairs is
not symmetric, and we say that g is a xenolog of gb. In a table with genes for columns and
rows, the entry for pair (b
g , g) will contain the annotation, and the entry for pair (g, gb) will
be ∗.

When there are multiple transfers on the path between genes gb and g, it is possible that gb
will be descended from one transfer and g will be descended from another transfer. In this
case, g is a xenolog of gb, and gb is a xenolog of g. We say that gb and g are incomparable
xenologs (IX). Both entries in the table — (b g , g) and (g, gb) — will contain the annotation.
Xenolog classification in the presence of multiple transfers is discussed in detail in [4].

5.6.2 Interactive Homology Mode

From the Notung GUI, users can investigate the homologs of any contemporary gene through
a point and click interface. The information displayed can then be saved as an image in PNG
format. We demonstrate how to get a table containing this information in Section 5.6.3 -
Homology Table.

Here, we demonstrate how Notung’s point and click interface can be used to investigate
homology relationships using two examples. One example shows an analysis with duplications
and losses. The second analysis demonstrates homology classification in a gene tree with both
horizontal transfers and duplications. Trees for both examples are included in the Notung
distribution and can be found in the sampleTrees directory.

Example 1: Investigating Orthologs and Paralogs. Functions for inferring homology

relationships when using the DL event model (Infer Transfers OFF) are illustrated here.

1. Select the Reconciliation task panel and reconcile the gene tree, if it has not yet been
reconciled.

2. Click on the “Show Homology” button to enter the interactive Homology mode. The
homology mode legend appears in the upper-left corner of the tree panel.

3. Click on or mouse over a leaf in the gene tree. It will be highlighted in light blue.
Orthologs of this gene will be highlighted in dark blue and paralogs in pink, as shown
in Figure 5.6.

57
 Chapter 5. Reconciliation Mode

4. To save an image displaying these relationships, select “File → Save Current View
as Image (PNG)” from the menu.

NOTE: The image will contain the Homology legend, as well as the orthologs
and paralogs of any currently selected gene. Currently, “File → Save Whole
Tree as Image (PNG)” does not show homology relationships.

5. To minimize the legend, click on “hide” in the legend. Click on the minimized legend
to show the full legend again.

6. The legend can be dismissed entirely by clicking “close.” If you re-enter Homology
mode, the legend will be visible again.

7. To exit Homology mode, click the “Show Homology” button.

Example 2: Investigating Orthologs, Paralogs, and Xenologs. Functions for infer-

ring homology relationships when using the DTL event model (Infer Transfers ON) are
illustrated here. This example reflects the scenario presented for xenolog classification in
Figure 5.5.

1. Select the Reconciliation task panel and reconcile the gene tree with Infer Transfers
ON, if it has not yet been reconciled.

2. Click on the “Show Homology” button to enter the interactive Homology mode. The
homology mode legend appears in the upper-left corner of the tree panel.

3. Click on or mouse over a leaf in the gene tree. It will be highlighted in light blue.
Orthologs of this gene will be highlighted in dark blue, paralogs in pink, and xenologs
in yellow, as shown in Figure 5.7.

NOTE: Specific colors are not associated with the different xenolog sub-
classes; however, this information is available in the homology table, as shown
in Table 5.3. See Section 5.6.3 - Homology Table on page 59 for instructions
on obtaining this table.

4. To save an image displaying these relationships, select “File → Save Current View
as Image (PNG)” from the menu.

NOTE: The image will contain the Homology legend, as well as the orthologs,
paralogs, and xenologs of any currently selected gene. Currently, “File →
Save Whole Tree as Image (PNG)” does not show homology relation-
ships.

58
Chapter 5. Reconciliation Mode

5. To minimize the legend, click on “hide” in the legend. Click on the minimized legend
to show the full legend again.

6. The legend can be dismissed entirely by clicking “close.” If you re-enter Homology
mode, the legend will be visible again.

7. To exit Homology mode, click off the “Show Homology” button.

5.6.3 Homology Table

Notung can also provide information on homology relationships between all pairs of leaves
in a gene tree in tabular format. This homology table is a matrix with a row (resp. column)
for every leaf in the gene tree. The [gi , gj ] entry in the table gives the homology relationship
of genes gi and gj . Orthologous and paralogous genes are indicated by an “O” and a “P” in
the table, respectively. Entries for xenologous pairs contain an “X”. Notung further classifies
xenologs into subtypes; information about each subtype, as described in Section 5.6.1, is
provided in the table as well. Abbreviations for homolog types are shown in Table 5.1.

NOTE: Xenolog pairs are not symmetric. Ordered pairs that are labeled with
a star will be classified in the other direction. For example, in Table 5.3, en-
try (gHat X,g Y) is PX, meaning g Y is a primary xenolog of gHat X. Entry
(g Y,gHat X) is *. For more information, refer to Section 5.6.1.

Notung can output the homology table in three table formats: tab-separated, comma-
separated (CSV), or HTML formats. HTML tables are coded to color cells according to
their assigned homology relationship. Cells representing orthologs have a blue background,
cells representing paralogs have a pink background, and cells representing xenologs have a
green background.

NOTE: Tab-delimited tables can usually be pasted directly into spreadsheet ap-
plications like Excel. CSV formatted tables can be opened by most spreadsheet
programs via the file menu. HTML format tables can be pasted directly into web
pages.

To view the Homology table:

1. Select the Reconciliation task panel and reconcile the gene tree, if it has not yet been
reconciled.

59
 Chapter 5. Reconciliation Mode

2. Go to the “About This Tree” menu.

3. Select the “Homology Table” option with the desired format (CSV, Tab delimited,
or HTML) in the pull-down menu. The selected table will be displayed in a pop-up
dialog box, for example as in Table 5.2 and Table 5.3.

NOTE: To save a copy of the homology table, in the pop-up box click “Copy
to clipboard,” then paste to a text buffer of your choice and save.

The homology table for the tree in Figure 5.6, reconciled under the DL event model (Infer
Transfers OFF), is illustrated in Table 5.2. The homology table for the tree in Figure 5.7,
reconciled under the DTL event model (Infer Transfers ON), is illustrated in Table 5.3.
This example reflects the scenario presented for xenolog classification in Figure 5.5.

Homology Table for: genetree SMALL

gB human gA human gA mouse g gorilla gB mouse gY cow gX cow

gB human . P P P P O O
gA human P . P P P O O
gA mouse P P . O P O O
g gorilla P P O . P O O
gB mouse P P P P . O O
gY cow O O O O O . P
gX cow O O O O O P .
Table 5.2: Homology table, showing orthologs and paralogs, for the example from Figure 5.6. In this example, genetree SMALL
has been reconciled with speciestree SMALL using the DL event model (Infer Transfers OFF). The abbreviation
key is provided in Table 5.1. Cells at the intersection of the column and row representing the same gene are labeled
with a dot.

60
Chapter 5. Reconciliation Mode

Homology Table for: xenolog gene

gY gHat X gZ gPrime Z gPrime Y gX gW gV

gY . * O P P O O O
gHat X PX . SDX SDXp SDXp SRX’ SRX OX
gZ O * . P P O O O
gPrime Z P * P . O O O O
gPrime Y P * P O . O O O
gX O * O O O . O O
gW O * O O O O . O
gV O * O O O O O .
Table 5.3: Homology table, showing orthologs, paralogs, and xenologs for the example from Figure 5.7. In this example,
xenolog gene has been reconciled with xenolog species using the DTL event model (Infer Transfers ON).
See Table 5.1 for a key to abbreviations. Notice that this table is not symmetric for xenologs. Xenolog pairs that
are not defined (i.e., (g, g
b)) are labeled with a star. Cells at the intersection of the column and row representing the
same gene are labeled with a dot.

61
 Chapter 5. Reconciliation Mode

(a)

(b)

Figure 5.1: Reconciliation of the binary gene tree with a binary species tree in Figure 3.1(b). (a) The dialog box allows you
to select which species tree to use, as well as the location of species labels in the gene tree. (b) The gene tree
decorated with duplications (red boxes and D’s) and losses (grey).

62
Chapter 5. Reconciliation Mode

(a)

(b)

Figure 5.2: A binary gene tree reconciled with a binary species tree using the DTL-model. There are two reconciliations with
the same, minimal Event Score. Both histories can be obtained by clicking on the green circle.

63
 Chapter 5. Reconciliation Mode

(a)

(b)

Figure 5.3: A binary gene tree reconciled with the non-binary species tree in Figure 4.1. Co-divergences are marked by pink
cD’s, while duplications are indicated with red D’s, and transfers are indicated by yellow T’s. Polytomy losses are
labeled with the name of the associated polytomy, as well as either (a), the names of the species from which they
are absent or (b) the number of species from which they are absent.

64
Chapter 5. Reconciliation Mode

(a)

(b)

Figure 5.4: Reconciliation of (a) a non-binary gene tree with the binary species tree in (b). More than one duplication may
be inferred at polytomies in the gene tree. In addition, it is possible to have more than one optimal event history,
as seen in the lower left-hand corner of the reconciliation panel in (a).

65
 Chapter 5. Reconciliation Mode

Figure 5.5: Xenolog classes of a gene family evolving through speciation, duplication, and horizontal gene transfer. The gene
tree (thin black lines) is shown embedded in the species tree. In the history of this hypothetical family, two events
occurred: a duplication (gene node 4) in species δ and a transfer (black arrow) from species Y to species X. The
cenancestor of the transfer is species β. Species that are more closely related to the donor and recipient species are
indicated by sets D and R, respectively. Species that are equally related to both the donor and recipient species
are indicated by the set O. Notung’s homology classification for this example is shown in Figure 5.7 and Table
5.3.

Figure 5.6: Notung’s point and click interface showing the homology relationships (orthologs and paralogs) to gene gA human
(highlighted in light blue). Orthologs are highlighted in dark blue, paralogs in pink. Gene tree genetree SMALL
has been reconciled with speciestree SMALL using the DL event model (Infer Transfers OFF).

66
Chapter 5. Reconciliation Mode

Figure 5.7: Notung’s point and click interface showing the homology relationships (orthologs, paralogs, and xenologs) in gene
tree xenolog gene, which has been reconciled with xenolog species using the DTL event model (Infer Transfers
ON). This reconciliation has two events: a duplication (node 4) and a transfer (node 1). Shown are the homology
relationships for genes (left) gHat X and (right) gPrime Z, both highlighted in light blue. Orthologs are highlighted
in dark blue, paralogs in pink, and xenologs in yellow. Since gene gHat X is a recipient of the transfer, all other genes
are xenologs to gHat X and are highlighted in yellow. The different xenolog subclassifications are not indicated.

67
Chapter 6

Rooting Mode

In Rooting mode, the event parsimony can be used to infer the root of a gene tree. Notung’s
Rooting Analysis calculates a root score for each edge in the tree, corresponding to the
DTL Score of the tree rooted on that edge. Note that the Rooting Analysis computes root
scores, but does not change the tree. The user must root the tree explicitly by clicking on an
edge in the tree panel. Rooting mode can also be used to root a tree manually by clicking
on any edge in the tree at any time, even if the Rooting Analysis has not been performed.

When the Rooting Analysis is complete, edges with the minimum root score are highlighted
in red. Notung also highlights edges with near-optimal scores in pink. Edges with scores that
are greater than the minimum by at most 5 percent of the difference between the maximum
and minimum score are highlighted in pink. Figure 6.1(a) shows the gene tree from Figure 5.1
after the Rooting Analysis has been applied. Note that optimal rooting edges are highlighted
in red, but the gene tree topology is unchanged from Figure 5.1. Figure 6.1(b) shows the
tree after it has been rooted by clicking in the tree panel.

When the gene tree is binary, applying Notung’s Rooting function with DL model results
in a binary gene tree with a root score on each edge. This score is a weighted sum of the
number of duplications, co-divergences, and losses. This score also represents the optimal
events cost when the binary gene tree is rooted at that particular branch. By default, the cost
of co-divergences is set to zero. Co-divergences will only influence the root score if this cost is
explicitly set to a positive number by the user. For more information on setting parameters,
see Chapter 3.5 - Parameter Values on page 22.

Rooting analysis with transfers proceeds somewhat differently from rooting with the DL-
model because temporal feasibility must be considered.

Rooting analysis when the gene tree is non-binary is currently only available for DL-model

68
Chapter 6. Rooting Mode

(a)

(b)

Figure 6.1: (a) The gene tree from Figure 5.1 after completing the Rooting Analysis. (b) The rerooted tree, after the user has
clicked on an edge to designate the root.

(without transfers). Rooting analysis with a non-binary gene tree differs from the binary case
in that root scores are assigned to polytomies, as well as edges. Edges and polytomies in the
original tree are assigned the DL Score associated with rooting on that edge or polytomy. If

69
 Chapter 6. Rooting Mode

rooting on a polytomy in a non-binary gene tree produces the minimum or near-minimum

score, that node will be circled and the vertical edge representing that polytomy will be
highlighted in the appropriate color (Figure 6.2).

Figure 6.2: Rooting analysis for a non-binary gene tree. The optimal root locations are colored in red. If an edge represented
by the polytomy can be selected as an optimal root, the polytomy will be circled and colored in red.

To re-root the tree, click on any edge or polytomy in the tree panel. You may root the
tree on any edge, not just the highlighted edges. Notung will root the tree on that edge
(or polytomy), and recalculate the reconciliation. The DL Score of the new, rooted tree is
displayed in the bottom-left corner of the screen.

Some tree reconstruction programs represent an unrooted tree as a rooted tree with a trifur-
cation (a polytomy with three children) at the root. Notung cannot distinguish between an
unrooted, binary gene tree and a rooted gene tree that has a single trifurcation. If such a
gene tree is opened and reconciled in Notung, a notification will appear to inform the user
that this tree may, in fact, be an unrooted, binary gene tree. Notung will assume that the
tree is rooted and non-binary, and will draw the tree and issue diagnostic messages accord-
ingly. If you want Notung to treat the tree as an unrooted binary tree, the trifurcation can
be removed by rooting the tree in the Rooting panel. Clicking on any edge will convert
the gene tree into a rooted binary tree. To select the edge with the optimal DL Score, run
the Rooting Analysis first.

NOTE : If the tree has not been reconciled before running a Rooting Analysis,
Notung will reconcile it automatically. You will be asked to select a species tree

70
Chapter 6. Rooting Mode

for reconciliation (see Chapter 5 - Reconciliation Mode on page 40).

6.0.1 Rooting with Transfers

In Rooting mode with transfers, Notung first visits each edge in the tree in turn, infers the
*candidate* optimal reconciliation(s) of the tree rooted on that edge and determines the
associated DTL score. This is the provisional root score for that edge, but is only a valid
root score if at least one of the candidate optimal reconciliations is also feasible. Once the
provisional root scores have been determined, the candidate reconciliations for the edge(s)
with the lowest provisional root score are tested for temporal feasibility until a feasible
solution is found. To reduce the running time, the remaining candidate reconciliations are
not tested, since the existence of one temporally feasible, optimal reconciliation is sufficient to
determine that the edge is a feasible root with minimum DTL score. If none of the candidate
reconciliations associated with an edge are feasible, then that edge is not a viable root.

In the GUI, Notung reports all edges with a minimum root score that have at least one feasible
solution, in contrast to the DL model, where Notung reports the root score for every edge
in the tree. With transfers, only edges corresponding to minimum cost roots are reported in
order to reduce the computational burden of testing (possibly many) candidate solutions for
feasibility. It is possible to infer the cost of sub-optimal roots from the command line. Please
refer to Section 12.7 - Rooting trees from the command line with the DTL model for details.

In the GUI, edges with the lowest root score are highlighted in red in the rooting panel, once
the rooting analysis is complete. To re-root the tree, click on one of the red edges. (You can
re-root the tree on any edge, but only red edges will give a minimum cost reconciliation.)
To see the reconciliation(s) associated with that root, go to the reconciliation panel and
re-reconcile the tree.

If no minimum cost edge has a feasible solution, then Notung does not report a root. Instead,
a pop-up warning will be displayed.

71
 Chapter 6. Rooting Mode

To find optimal root edges:

1. Click the Rooting tab to enter Rooting mode.

2. Click “Run Rooting Analysis.”

Good roots will be highlighted. If highlighted edges are small, they are circled in the appro-
priate color to help the user locate them visually. Use the Zoom feature (see Chapter 11.2 -
Zoom on page 95) to zoom in on these edges.

To show/hide Rooting Analysis results:

In Rooting Mode, the task panel contains several check boxes that allow the user to specify
what rooting related information should be displayed.

• Click the appropriate checkbox in the Rooting task panel.

√
Highlight good roots (default: ON) When the “Highlight good roots” box
is checked, the low-scoring root edges in the tree are colored (red for the min-
imum score, pink for scores within 5 percent of the minimum). To remove the
highlighting, uncheck the box.
√
Highlight polytomies (default: OFF) When the “Highlight polytomies” box
is checked, vertical edges representing polytomies in the gene tree, if any exist, are
highlighted and circled in cyan.
√
Circle small edges (default: ON) When the “Circle small edges” box is checked,
appropriately colored circles appear around small, low-scoring root edges in the
tree, if any exist, making it easier to locate them in large trees. To remove the
circles, uncheck the box.
√
Display root score (default: OFF) When the “Display root score” box is
checked, the associated D/L Score appears next to each edge and polytomy in
the tree. Rooting scores are shown in pink to distinguish them from edge weights,
which are displayed in green.

NOTE : For efficient run times, Notung only tests optimal roots for temporal
feasibility. Therefore, rooting with transfers cannot display root scores for each
edge, as they may not be feasible.

To re-root the tree:

• Click on any edge or polytomy of the tree in the tree panel.

72
Chapter 7

Rearrange Mode

Weakly-supported edges, as indicated by low edge weights, often imply that the inferred
history associated with those edges may not be accurate. Notung can rearrange weakly-
supported regions in a gene tree to produce alternate event histories with minimum DL
Score. When these edges or regions are rearranged, the structure of strongly-supported
edges or regions stays intact. Any edge that is added as a result of rearrangement will be not
be assigned an edge weight. Since support for edges is determined by edge weight, Notung’s
rearrangement function requires that the gene tree include edge weights which assess how
well each edge is supported by sequence data. These edge weights can be bootstrap values,
probabilities, or branch lengths.

NOTE : Currently, Notung can not perform rearrangement with non-binary species
trees.

Weak edges are defined as those edges with weights below the Edge Weight Threshold.
Selecting the “Highlight weak edges” checkbox in Rearrange mode will highlight all weak
edges in yellow, allowing the user to see which edges will be considered for rearrangement
(see Figure 7.1). This option is only available in Rearrange mode. The yellow highlighting
will disappear when another mode is selected. As a default, the Edge Weight Threshold is
90% of the maximum edge weight. While this is a good starting place for bootstrap values,
it may not be appropriate for probabilities or branch lengths. The threshold can be adjusted
by the user; see Chapter 3.5 - Parameter Values on page 22 for information on how to change
the Edge Weight Threshold. Notung also considers any edge without an assigned weight to
be a weak edge. If Notung’s rearrangement function is applied to a tree with no edge weights,
it will consider all edges to be weak, and will find all trees that are optimal when only events
are considered (i.e. those trees with a minimal Event Score).

73
 Chapter 7. Rearrange Mode

The Rearrangement function can be applied to a non-binary gene tree when the species
tree is binary (Figure 7.2). Notung will replace each polytomy with an arbitrary binary
resolution, inserting new nodes and edges. These new edges are treated as weak edges. The
standard rearrangement algorithm is then applied to the resulting binary tree to determine
the rearrangement that results in a minimal Event Score. Note that it is immaterial how the
polytomies are initially resolved, because subsequent rearrangement will result in a minimum
cost tree. Rearrangement cannot be performed when the species tree is non-binary.

7.1 Alternate Optimal Hypotheses

When rearranging a gene tree, there may be more than one tree that (1) agrees with the
original tree at strongly supported edges and (2) has minimal Event Score. If there are many
such trees, considering all of them may be a daunting task. Notung addresses this issue by
partitioning the set of all optimal trees into subsets in such a way that any tree in a given
subset can be generated from any other tree in the subset by a series of node interchanges.

All trees in any given subset are instances of the same event history. An event history
describes a series of events (duplications and losses) and the location in the species tree
where they occurred. “A duplication in the common tetrapod ancestor, a loss in the fish
lineage and three duplications in mouse” is an example of an event history. To see that
more than one tree can have the same event history, note that “three duplications in mouse”
corresponds to the subtree ((g1 mouse, g2 mouse), (g3 mouse, g4 mouse)), as well as
the subtree (((g1 mouse, g2 mouse), g3 mouse), g4 mouse).

If multiple minimum cost trees are found, Notung presents one tree from each subset (i.e.
one representative of each event history) to the user and provides a point and click interface
that allows the user to inspect any other tree in that subset. Initially, Notung arbitrarily
selects one event history to present in the tree panel. The other optimal histories may be
viewed using the drop-down menu labeled “Select an optimal event history,” which gives a
list of up to 50 optimal event histories. The user can perform Same Cost Swaps on a tree to
explore the space of all optimal trees corresponding to the current event history. Same Cost
Swaps are node interchanges that result in another tree with an optimal DL score. Clicking
the “Examine same-cost swaps” button will highlight all swappable nodes, nodes that
can be manually swapped without changing the DL Score.

If there are more than 50 optimal event histories, they can be generated using the Command
Line Interface (see Chapter 12 - Command Line Options and Batch Processing on page 98).
Note that both the drop down menu and command line options give distinct optimal event
histories, but for the DL event model, do not generate all optimal gene tree rearrangements.
It is only possible to view all trees by performing same cost swaps using the point and click

74
Chapter 7. Rearrange Mode

interface in the GUI.

For further details on Notung’s rearrangement algorithm see:

D. Durand, B. V. Halldorsson, B. Vernot. A Hybrid Micro-Macroevolutionary

Approach to Gene Tree Reconstruction. Journal of Computational Biology, 13(2):
320-335, 2006.
H. Lai, M. Stolzer, and D. Durand. Fast Heuristics for Resolving Weakly Sup-
ported Branches Using Duplication, Transfers, and Losses. Proceedings for RE-
COMB International Workshop on Comparative Genomics: 298-320, 2017.

75
 Chapter 7. Rearrange Mode

(a)

(b)

Figure 7.1: (a) The gene tree from Figure 5.1 with weak edges highlighted. (b) After clicking “Perform Rearrangement,” the
rearranged tree appears in the tree panel. Weak edges are still highlighted in yellow.

76
Chapter 7. Rearrange Mode

(a)

(b)

Figure 7.2: (a) When rearranging the non-binary gene tree, weak edges are highlighted in yellow. These edges, as well as the
polytomies, highlighted in cyan, will be rearranged to produce the binary tree with the minimal D/L Score. (b)
After the tree is rearranged, weak edges are highlighted in yellow. Notice that new edges have no edge weight and
are considered weak.

77
 Chapter 7. Rearrange Mode

7.2 Rearrangement Commands

NOTE : Currently, Notung can only perform rearrangement with transfers via the
command-line interface. See Section 12.8 - Non-binary gene trees with the DTL
model on page 117 for more information.

To rearrange the gene tree:

1. Click the Rearrange tab to enter Rearrange mode.

2. Click “Perform Rearrangement.”
A minimum cost rearrangement tree will appear in the tree panel as shown
in Figure 7.1(b). Note that weak edges, highlighted in yellow, will not have
edge weights. Some or all of these are edges that do not correspond to any
bipartition (split) represented in the original tree. The appropriate weights
for these edges are not known.

NOTE : If asked to rearrange a tree that has not been reconciled, Notung will
reconcile it automatically. In this case, the user is asked to select a species tree
for reconciliation.

To highlight all weak edges (default: OFF):

• Click the “Highlight weak edges” checkbox.

All weak edges in the tree will be highlighted in yellow.

To view alternate optimal event histories:

If more than one optimal event history exists for a rearranged tree, the drop down
menu “Select an optimal event history’’ will be enabled.

• From the drop-down menu, select an alternate event history.

The tree panel will now show a new tree corresponding to the selected alter-
nate history.

If there is only one optimal history or the tree has not yet been rearranged, the drop down
menu will be grayed out.

78
Chapter 7. Rearrange Mode

(a)

(b)

Figure 7.3: When examining same cost swaps, swappable nodes are marked with the enlarged square. (a) Same cost swaps
are highlighted. The selected node, shown in blue, can be swapped with the node highlighted in orange. Clicking
first on the blue node and then on the orange node results in the alternate optimal tree shown in (b).

79
 Chapter 7. Rearrange Mode

To swap individual nodes:

1. Click the “Examine same cost swaps” button in the right column on the Rearrange
task panel.

NOTE : If there are no swappable nodes in the tree or if the tree has not yet
been rearranged, this button will be grayed out.

Swappable nodes are marked with an enlarged blue and cyan square. As you pass
the mouse over a swappable node it will be highlighted with a blue triangle. Other
nodes that can be interchanged with it are temporarily highlighted with a light orange
triangle, as shown in Figure 7.3(a). If you have zoomed in, some swappable nodes may
be outside the boundaries of the tree panel. Swappable nodes that are not currently
visible are indicated by arrows in the tree panel, pointing in the direction of those
nodes. These can be seen by scrolling in the direction of the arrow.

2. Click a node to swap. The node you selected is highlighted with a blue triangle. Nodes
with which it can be swapped are now highlighted with red triangles.

3. Click a second node to complete the swap (see Figure 7.3(b)).

NOTE : When a user selects a different alternate event history from the “Select
an optimal event history” list, Notung rebuilds the tree from data saved at
the time of rearrangement. Any manual swaps made to a previously viewed event
history will be lost. Therefore, if you wish to save information after a manual
swap, you must save your tree. See Chapter 3.3 - Opening and Saving Trees on
page 15 for more information.

80
Chapter 8

Resolve Mode

Resolve mode is only applicable to non-binary gene trees. Its function is to resolve poly-
tomies in a non-binary gene tree by comparing it with a binary species tree, resulting in one
or more binary tree(s) with minimal DL Score.

NOTE : Notung currently cannot reconcile non-binary gene trees using the trans-
fer algorithm. Therefore, such trees cannot be resolved with transfers.

Specifically, the Resolve function removes all polytomies in the original gene tree, and uses
an algorithm similar to the rearrangement algorithm to replace them with new edges such
that: 1) the new tree is binary, and 2) the new tree has optimal DL Score. Note that each
edge in the original non-binary gene tree still exists in the resulting binary gene tree.

There may be more than one binary tree that agrees with the input tree at all edges except
polytomies and has minimal DL Score. In this case, the user can investigate these optimal
alternate hypotheses using a point and click interface as in Rearrange mode. See Section
7.1 - Alternate Optimal Hypotheses on page 74 for a more detailed explanation of alternate
hypotheses.

Selecting the “Highlight Polytomies” checkbox will highlight in cyan all vertical edges
representing polytomies in the tree, allowing the user to see which nodes will be resolved.
After running the resolve algorithm, the “Highlight New Edges” checkbox will be selected,
and will highlight in cyan all those edges in the gene tree that were previously represented
by the polytomy (see Figure 8.1). This option is only available in Resolve mode.

81
 Chapter 8. Resolve Mode

(a)

(b)

Figure 8.1: (a) Polytomies in the gene tree can be highlighted in cyan while in the Resolve task mode. (b) After the polytomies
are resolved, edges that were not present in the original tree are highlighted in cyan.

82
Chapter 8. Resolve Mode

To resolve the gene tree:

1. Click the Resolve tab to enter Resolve mode.

2. Click “Resolve Polytomies.”
A minimum cost binary resolution of all polytomies in the tree will appear in the tree
panel. Note that the new edges will not have edge weights.
If the gene tree is binary, the “Resolve Polytomies” button will be grayed out.

NOTE : If asked to resolve a tree that has not been reconciled, Notung will first invoke
the reconciliation algorithm. In this case, the user is asked to select a species tree for
reconciliation.

To highlight all polytomies (default: OFF):

• Click the “Highlight Polytomies” checkbox.

All vertical edges representing polytomies in the tree will be highlighted.

To highlight all new edges (default: ON, after resolving):

• Click the “Highlight New Edges” checkbox.

All edges that were represented by the polytomies in the original tree will be
highlighted.

To view alternate optimal event histories:

1. If more than one optimal event history exists for a resolved tree, the drop down menu
“Select an optimal event history” will be enabled.
2. From the drop-down menu, select an alternate event history.
The tree panel will now show a new tree corresponding to the selected alter-
nate history.

If there is only one optimal history or if the polytomies have not been resolved, the drop
down menu will be grayed out.

To swap individual nodes:

83
 Chapter 8. Resolve Mode

1. Click the “Examine same cost swaps” button on the Resolve task panel.

NOTE: If there are no swappable nodes in the tree or if the polytomies have
not been resolved, this button will be grayed out.

Swappable nodes are marked with an enlarged blue and cyan square. As you pass the
mouse over a swappable node, other nodes that can be interchanged with it with are
temporarily highlighted with a light orange triangle. Swappable nodes that are not
currently visible in the tree panel (for instance, if you have zoomed in), are indicated
by arrows in the tree panel pointing in the direction of those nodes.

2. Click a node to swap.

The node you selected is highlighted with a blue triangle. Nodes with which it can be
swapped are now highlighted with pink triangles.

3. Click a second node to complete the swap.

NOTE: When a different alternate event history is selected in the “Select an optimal
event history” list, Notung rebuilds the tree from data saved at the time of resolution.
Any manual swaps made to a previously viewed event history will be lost. Therefore,
if you wish to save information after a manual swap, you must save your tree. See
Chapter 3.3 - Opening and Saving Trees on page 15 for more information.

84
Chapter 9

History

The state of a gene tree changes each time a Notung operation, such as rooting, rearrange-
ment, reconciliation, or resolution, is performed on the tree. Notung maintains a history of
state changes for each gene tree. This history can be accessed via the History panel, allowing
the user to return to and operate on a previous state, or visually compare the state before
and after a task is performed.

Notung lists the states in the history panel by task name (see Figure 9.1). The first entry
in the list is always Start, which is the state of the tree when loaded; others entries may
include Changed Parameter Values, Reconciled, Rooting Analysis, Rooting on
X, Notung Rearrange, Notung Resolve Polytomies, Select Alternate Optimal
History, and Swapped Y and Z, where X is an edge and Y and Z are swapped nodes.
The list proceeds from top to bottom in the order tasks were performed, and includes the
Event Score for each state.

NOTE: Previous states in the History panel are not saved in a file. When the gene tree file
is closed, the history associated with the current tree is lost. To save trees associated with
intermediate states, select the state and click “File → Save As.”

NOTE: Parameter values are saved with each state in the history. For each state in the
history, the parameters will correspond to those values used at the time the operation was
performed. Any subsequent changes to parameter values will not be applied retroactively.

To view previous states of the gene tree:

1. Click the History tab to enter History mode.

2. Click on an item in the list.

85
 Chapter 9. History

Figure 9.1: The history of a gene tree that had been reconciled, rooted, and rearranged. Currently, the state of the tree after
reconciliation and prior to rooting is selected and displayed in the panel.

86
Chapter 10

Annotations

Notung can annotate the leaf nodes of both gene and species trees with colors specified by
the user. For example, the annotation function can be used to color all nodes associated with
a particular taxonomic group (e.g., plants) or a particular subfamily (e.g., HSP70). This
can help visually differentiate gene clusters in a large and complex tree, or highlight related
nodes that are distantly located in a tree.

The “New” button in the Annotations task panel opens the annotations dialog window
(see Figure 10.1), where the user can set the annotation parameters. Each annotation consists
of a title used to identify it, a color, and a specification of the nodes that are included in the
annotation. The title of an annotation is simply an alphanumeric string used to distinguish
it. You may use any string of characters as long as it is unique. The set of nodes associated
with a given annotation can be specified in two ways, by pattern matching or by selecting
them manually. In the first case, the user provides one or more alphanumeric strings, which
are compared with all leaf node names. Leaf nodes that contain one or more of the specified
strings as a substring are added to the annotation. Alternatively, nodes can be manually
added to the annotation by clicking on them.

All annotations for the currently selected tree are shown in the list box in the Annotations
task panel (see Figure 10.1(b)). After an annotation is created, individual nodes can be
added to it or removed from it, manually. Annotations can be edited to modify the list of
pattern matching terms or to change the color associated with the annotation. Annotations
can be shown or hidden at any time.

NOTE: A single node can match more than one annotation, but will only be colored by the
most recently created annotation.

NOTE: Annotations only apply to the tree that is currently selected, but can be exported and

87
 Chapter 10. Annotations

(a)

(b)

Figure 10.1: (a) The annotations dialog box. This figure shows the creation of an annotation, with the title “primates”,
associated with the color red, and the pattern matching terms “hu” and “go”. (b) A fully annotated gene tree.
In the Annotations task panel, the list box shows the annotations associated with the currently selected tree. A
check box indicates whether the annotation is hidden (unchecked) or showing (checked). The number next to
each annotation refers to the number of leaf nodes currently colored by that annotation. If the annotation is
hidden, this number is zero.

88
Chapter 10. Annotations

then imported into another tree. See subsections on importing and exporting annotations at
the end of this section.

To create an annotation using pattern matching (recommended):

1. Click the Annotations tab to enter Annotations mode.

2. Click the “New” button in the task panel. A dialog box appears.

3. Select a color in the color palette for the annotation. If you do not select a color,
Notung selects red by default.

4. Enter the title of the annotation in the text field in the center of the dialog box.

5. Select the radio button marked “Use this comma-delimited list to add nodes.”
(This button is selected by default.)

6. Enter the pattern matching term(s) in the text field, separated by commas. If no terms
are entered, Notung will use the title of the annotation as a pattern matching term.

For example, if you want to annotate all the node labels containing HU, en-
ter “HU” in the text field. Notung will annotate any node with a label that
contains “HU” as a substring, such as g1 human and g2 human. If you want
to annotate all node labels containing “HU” and “GO,” enter “Hu, Go” in
the text field. This will also annotate the node g1 gorilla, as seen in Figure
10.1.

NOTE : This process is not case sensitive.

7. Click “OK.”

Nodes with names that match a string in the comma-delimited list will change
color (e.g., Figure 10.1).

If a single node corresponds to more than one annotation, the node will be in the color
dictated by the most recently added annotation. The newer annotation will continue to take
precedence until the shared node is manually removed from that annotation, the annotation
is hidden, or a new, conflicting annotation is added. For example, adding an annotation in
yellow for “g1” would change the color of g1 human, g1 cow, g1 mouse, and g1 gorilla to
yellow.

To create an annotation with manually added nodes:

1. Click the Annotations tab to enter Annotations mode.

89
 Chapter 10. Annotations

2. Click the “New” button in the task panel. A dialog box appears.

3. Select a color in the color palette to use for the annotation (if you do not select a color,
Notung selects red by default.)

4. Enter the title of the annotation in the text field in the center of the dialog box.

5. Select the radio button marked “I want to manually select the nodes and sub-
groups to add.”

6. Click “OK.” This will create an empty annotation that does not color any nodes. To
color nodes, you must add them to the annotation as described below.

To add nodes to an annotation manually:

NOTE : This operation can only be performed if an annotation has already been created.

1. Click the “Add Nodes” button in the Annotations task panel

2. Select the desired annotation by clicking on it in the list box.

3. Click on nodes in the tree panel.

If a selected node is a leaf node, it will be highlighted with the color of

the annotation. If it is an internal node, all of leaf nodes below it will be
highlighted with the color of the annotation.

4. If you want to add nodes to another annotation, repeat steps 2 and 3.

5. When you are finished adding nodes to annotations, click the “Add Nodes” button
again to deselect it.

To remove nodes from an annotation manually:

NOTE: This operation can only be performed if an annotation has already been created and
nodes have been assigned to it.

1. Click the “Remove Nodes” button in the Annotations task panel, and then select
the desired annotation from the list of annotations.

2. Click on nodes in the tree panel.

90
Chapter 10. Annotations

If a selected node is a leaf node, it will be removed (i.e., disassociated) from

the annotation and the color of its label will revert to black or the color of
another annotation with which they are associated. Clicking on an internal
node removes all of the leaf nodes in the subtree rooted at that node.

3. When you are finished, click the “Remove Nodes” button again to deselect it.

To edit an annotation:

1. Select the annotation you want to edit in the list box.

2. Click the “Edit” button in the task panel. The annotation dialog box will reappear.
You can now change the color, title or pattern matching term(s) for this annotation.

3. Click “OK” when you are done making changes.

To hide/view an annotation:

1. Select the annotation in the list box.

2. Click the “Show/Hide” button.

If the annotation was displayed prior to clicking the “Show/Hide” button,

the nodes associated with the annotation will revert to black or the color of
another annotation with which they are associated. If the annotation was
hidden, the associated nodes will appear in color. A check mark next to the
annotation’s name denotes that it is visible (i.e., the current state is “Show”).
This is the default status.

To delete an annotation:

This function will remove an annotation from the list of annotations. All nodes associated
with it will revert to black or the color of another annotation with which they are associated.
Warning: this operation is not reversible.

1. Select the annotation in the list box.

2. Click the “Delete” button in the task panel.

3. A dialog box will appear to confirm that you really want to delete the annotation. Click
“Yes.”

91
 Chapter 10. Annotations

To export an annotation:

Annotations can be exported to a separate file for import into another tree.

1. Click “File → Export Annotations.”

2. A file dialog box will appear. Enter a name for the annotation, and click “Save.” The
annotations can now be imported from this file to another tree.

To import an annotation:

Annotations can be imported from any file that contains an annotation, including a Notung
format tree or an exported annotation file.

1. Click “File → Import Annotations.”

2. An open file dialog box will appear, select the annotation file or annotated tree, and
click “Open.” The annotations from the file will be added to the open tree. If the
open tree was previously annotated, those annotations will still be present.

NOTE: Imported annotations are added to the existing list of annotations.

If an imported annotation and a previously existing annotation correspond
to the same node, the imported annotation will take precedence.

92
Chapter 11

Changing the Appearance of the Tree

Panel

Notung offers the user a broad range of options for controlling the appearance of the tree
panel and the types of information that can be displayed. Visual presentation of trees in
Notung can be changed in two ways. One set of options is found in the Display Options,
Zoom, and Font menus in the upper left hand corner of the Notung window. These options,
which are described in this section, are relevant in all task modes.

In addition, certain visual features can be controlled from individual Task Panels. These
features are typically specific to that task and, in most cases, are only visible when the
relevant Task Panel is selected. These options, such as highlighting swappable nodes or weak
edges and displaying root scores, are described in the relevant mode sections.

11.1 Display Options

Notung allows users to show or hide node and edge labels using the Display Options menu.
Checkboxes next to each item in the menu show which display options are turned on. These
options are tree-specific: changing them in the currently selected tree will not change them
in other open trees.

To turn on/off a display option:

• Click Display Options, then click the desired option.

93
 Chapter 11. Changing the Appearance of the Tree Panel

√
Display Edge Weights (default : ON) When this option is turned on, the weight
of each edge is displayed in green type below the edge.
√
Display Loss Nodes (default: ON) When this option is turned on, the implied
lost nodes are displayed in the tree.
√
Display Duplications (default: ON) When this option is turned on, a red D
appears next to each duplication node.
√
Display Co-Divergences (default: ON) When this option is turned on, a pink
cD appears next to each co-divergence.
√
Display Transfers (default: ON) When this option is turned on, a yellow T ap-
pears next to each transfer edge.
√
Display Leaf Node Names (default: ON) When this option is turned on, leaf
node names are displayed.
√
Display Internal Node Names (default: OFF) When this option is turned on,
the name of each internal node appears in red type to the upper left of the node.
NOTE : If internal nodes are not labeled in the input gene tree file, Notung
assigns internal node names using a counting system. Node names derived
this way begin with n or r, followed by number signifying the order in
which the node was counted (e.g., n136 or r122). Node names beginning
with an r indicate that the node did not exist in the original tree; rather,
they were added during rearrangement of weak edges or resolution of a
polytomy. These node names may change if other tasks are performed
after the rearrangement or resolve task that produced the new internal
node(s). Counting may be different for each session, so node names may
vary from session to session, depending on how many trees have been
opened in Notung.
√
Display Leaf Node Species Names (default: OFF) When turned on, this op-
tion displays the name of the species associated with each leaf node. It appears
in black italic text next to the node.
NOTE: In previous versions of Notung, punctuation in species names
was used to indicate that Notung should search for a shorter species tag
in gene names, rather than searching for the entire species name. This
functionality has been removed in Notung 2.6 and beyond. For more
information, see Appendix A.5 - Punctuation in Species Names on page
130.
√
Display Internal Node Species Names (default: OFF) When this option is
turned on, the (ancestral) species associated with each internal node in the tree
is displayed. It appears in black italic text to the right of the node.
NOTE: If ancestral nodes are not labeled in the input species tree, internal
nodes are assigned numerical identifiers. Since the species name is derived

94
Chapter 11. Changing the Appearance of the Tree Panel

from the reconciliation, if the gene tree has not been reconciled, this
option has no effect.
√
Highlight Polytomies (default: OFF) When turned on, this option highlights
and circles in cyan each vertical edge representing a polytomy.
√
Use Species Names in Polytomy Losses (default: ON) When reconciling a gene
tree with a non-binary species tree, if orthologous genes are absent in two or more
children of a polytomy in the species tree, a single loss in an ancestral species is
inferred (called a polytomy loss). When this option is turned on, polytomy losses
are displayed with the names of the species in which the gene is absent, as well as
name of the polytomy (see Figure 5.3(a)). When turned off, combined losses are
labeled with the number of species in which the gene is absent, and the name and
the size of the polytomy (Figure 5.3(b).)

11.2 Zoom

Notung allows users to zoom in on the tree using either the Zoom menu or keypad controls.
Users can zoom in on the whole tree, maintaining the tree’s aspect ratio, or on the X or Y
axis independently, elongating the vertical or horizontal edges, respectively. These changes
apply only to the currently selected tree.

To zoom in on the whole tree:

• Click “Zoom → Zoom In.”

-or-
• Press “Ctrl/Cmd” and click on the region of interest in the tree.
NOTE: Ctrl/Cmd means to use the “Control” key for Windows or Linux,
and the “Command” (open apple) key for Mac. For operations which do not
involve clicking on the tree, the “Control” key is used for all three platforms.

To zoom out on the whole tree:

• Click “Zoom → Zoom Out.”

-or-
• Press “Shift” and click on the tree.

To zoom in on the X axis:

95
 Chapter 11. Changing the Appearance of the Tree Panel

• Click “Zoom → Zoom In on X axis.”

-or-
• Press “Ctrl - ]”

To zoom out on the X axis

• Click “Zoom → Zoom Out on X axis.”

-or-
• Press “Ctrl - [“

To zoom in on the Y axis:

• Click “Zoom → Zoom In on Y axis.”

-or-
• Press “Ctrl - Shift - ]”

To zoom out on the Y axis:

• Click “Zoom → Zoom Out on Y axis.”

-or-
• Press “Ctrl - Shift - [”

To fit the whole tree in the tree panel:

• Click “Zoom → Show Whole Tree.”

-or-
• Press “Ctrl - T”

11.3 Changing Font Size

Users can modify the font size of tree labels using the Fonts menu or keypad controls. Fonts
can be set to one of four sizes or changed incrementally.

To set a font size:

96
Chapter 11. Changing the Appearance of the Tree Panel

• Click “Fonts,” then click the desired size.

– Tiny fonts
– Small fonts
– Medium fonts (default)
– Large fonts

To increase font size incrementally:

• Click “Fonts → Increase Font Size.”

-or-

• Press “Ctrl - =”

To decrease font size incrementally:

• Click “Fonts → Decrease Font Size.”

-or-

• Press “Ctrl - Minus”

NOTE: Selecting “Large fonts” does not display the largest possible font; the
font can be made even larger by using the “Increase font size” option.

97
Chapter 12

Command Line Options and Batch

Processing

Notung offers a command line interface that can perform most operations from the command
line without launching the graphical user interface. The command line interface allows the
use of batch processing to apply Notung to many trees in a large-scale analysis without human
intervention. It can also be used to analyze a small number of trees without launching the
GUI, for example, by a user executing Notung on a remote computer over the network. The
GUI can also be launched from the command line, rather than by clicking on an icon, allowing
the user to initiate the GUI with parameter settings other than the default settings.

We follow the following stylistic conventions in this chapter.

• Text that should be typed verbatim on the command line, such as the name of the No-
tung executable and program options, is given in teletype font; (e.g. Notung-2.9.jar
--reconcile ....)

• Throughout this chapter, we refer to the Notung executable using the most recent major
version number (i.e., Notung-2.9.jar). We periodically release versions of Notung with
minor changes, such as bug fixes. The minor version number is increased with each
new minor release. You may have a more recent build with a minor version number,
(e.g., Notung-2.9.1.jar.) If so, when typing the commands given in the examples in
this chapter, you should replace Notung-2.9.jar with name of the Notung executable
in the distribution that is currently installed on your computer.

• Text in angle brackets indicates that the user must supply a parameter value. For ex-
ample, Notung-2.9.jar --reconcile -g <genetree>, indicates that Notung expects
a file name after -g.

98
Chapter 12. Command Line Options and Batch Processing

• Angle brackets are also used to describe fields that will be instantiated with a value
during Notung’s execution. For example, many output files include either reconcile,
rearrange, resolve, or root in their name, to indicate the task used to analyze the
original gene tree. Reconciling the gene tree stored in a file called mygenetreefile will
produce the file mygenetreefile.reconciled.ntg. In this chapter, we use <function>
to describe such file names, e.g., <genetree>.<function>.ntg.

• When Notung outputs more than one optimal tree, it generates multiple output files,
numbered starting at zero, yielding file names such as mygenetreefile.reconciled.0.ntg,
mygenetreefile.reconciled.1.ntg, etc. We use a ‘#’ sign to represent the number
in such file names; e.g., <genetree>.<function>.#.ntg.

12.1 Opening and Using a Command Window/Terminal

Prior to running Notung’s command line interface, you will need to open a command or
“terminal” window.

On Windows XP

Opening a command window

Click on the Start button, and select the “Run...” item. A dialog box will pop
up. Enter “cmd.exe” into the box, and click “OK.”

Navigating to the Notung directory

In the command window, type the following

cd <pathname>

where <pathname> is the path of the Notung directory. If the folder location has
any spaces in it, it must be enclosed in quotes. For example, if the following is
the location of the Notung folder:

C:\Documents and Settings\User\Desktop\Notung-2.9

99
 Chapter 12. Command Line Options and Batch Processing

Then you should use quotes so that it looks like this in the command window:

cd "C:\Documents and Settings\User\Desktop\Notung-2.9"

Hit Enter, and you will now be in the Notung Folder.

NOTE : To find the path of the Notung directory, select the Notung
folder in Explorer, and right click on it. This will pop up a menu - select
the Properties item. This will pop up a dialog listing the properties of
the Notung folder, including its location.

On Windows Vista or later

Opening a command window in the Notung directory

Select the Notung folder in Explorer, and SHIFT + right-click on it. This will
pop up a menu - select “Open command window here.”

On Mac

Opening a terminal

The Terminal application is located in the Applications folder in the Utilities

subfolder.

Navigating to the Notung directory

In the terminal window, type the following

cd <pathname>

where <pathname> is the path of the Notung directory. If the folder location has
any spaces in it, it must be enclosed in quotes. For example, if the following is
the location of the Notung folder

100
Chapter 12. Command Line Options and Batch Processing

/Users/user/Desktop/New Folder/Notung-2.9

Then it should look like this in the terminal window

cd "/Users/user/Desktop/New Folder/Notung-2.9"

Hit Enter, and you will now be in the Notung Folder.

NOTE : To find the path of the Notung directory, select the Notung
folder in the Finder, and select “Get Info” from the File menu. This
will pop up a dialog listing the properties of the Notung folder, including
its location. You could also drag and drop the Notung folder into the
Terminal window to paste the folder’s path into the window.

On Linux

Navigating to the Notung directory

In the terminal window, type the following

cd <pathname>

where <pathname> is the path of the Notung directory. If the folder location has
any spaces in it, it must be enclosed in quotes. For example, if the following is
the location of the Notung folder

/Users/user/Desktop/New Folder/Notung-2.9

Then it should look like this in the terminal window

cd "/Users/user/Desktop/New Folder/Notung-2.9"

Hit Enter, and you will now be in the Notung Folder.

101
 Chapter 12. Command Line Options and Batch Processing

12.2 Running Notung from the command line

Notung can carry out its four main tasks: reconcile, rearrange, root and resolve, from the
command line. In each case, Notung reads in gene and species trees (the input trees) and
executes the specified task, resulting in one or more modified trees (the output tree(s)).
This modified tree is written to a file. Notung can also generate images in PNG format
from the command line. This function can be carried out in conjunction with any of the four
main tasks, or independently to generate an image of an existing tree without performing any
analysis. The I/O requirements differ somewhat in the latter case; only one tree is required as
input and an image rather than a tree file is generated as output. In this section, we discuss
executing the four main tasks from the command line. Commands and options specific
to image generation are described in Section 12.5. In Section 12.3, automated execution
of Notung is described. Notung can gather useful phylogenomic analysis data from batch
reconciliations. These statistics are described in Section 12.4. Commands and options specific
to reconciliation with non-binary species trees are described in Section 12.6.

For the four major tasks, Notung is executed from the command line using the following
format:

java -jar Notung-2.9.jar [input tree(s)] [task] [options]

The four main tasks require both a gene tree and a species tree. These are usually supplied as
two separate input files. A single file containing a previously reconciled tree in Notung format
is also acceptable, since such files contain both a gene tree and species tree (see Appendix
A.3 - Notung File Format on page 128). If a gene tree file containing a reconciled tree in
Notung format and a species tree in a separate file are both given, the latter is used; the
species tree in the gene tree file is ignored. The task parameter must be one of --reconcile,
--rearrange, --root, or --resolve (the fifth task, --savepng, is discussed in Section 12.5
- Saving PNG Images of Trees.) Options are described at the end of this chapter.

NOTE :

• The input trees, tasks, and options may be given in any order.
• To launch the graphical interface from the command line, run Notung with-
out a task option and without a command to save an image.
java -jar Notung-2.9.jar

• Running Notung with the --help option causes it to print information re-
garding input, output, and other options.

102
Chapter 12. Command Line Options and Batch Processing

• You may also run Notung anywhere from the command line by giving the
path during launch:

java -jar <pathname>/Notung-2.9.jar [input tree(s)] [task] [options]

Alternatively, you can add the Notung directory to your CLASSPATH. For
example, if you run bash in Linux, you can do this by adding the following
command to your .bashrc file:
setenv CLASSPATH $CLASSPATH:<pathname>
See a java manual for more information about CLASSPATH settings.

The following list summarizes the types of information that Notung will generate, with a
brief description of each output file type. File formats for the information generated with the
--phylogenomics option are described in Section 12.4 - Running a phylogenomic analysis.
For more details on tree formats, including information on edge weights, species tags, and
output files, see Appendix A - File Formats on page 126.

Output options

Output Gene Tree(s) If one of the four main functions is given, the output gene tree will
be saved to a file called <genetree>.<function>.<extension>, where <function> is one of
the four major tasks and <extension> is ntg for Notung format, nhx for Newick Extended for-
mat, and nwk is Newick or New Hampshire format. If the analysis results in more than one op-
timal history, then the output files are numbered, (e.g. <genetree>.rearrange.0.<extension>,
<genetree>.rearrange.1.<extension>, etc.). By default only one tree is saved. To save
more than one history when inferring transfers, use --multsols. To save more than one his-
tory for all other situations (rearranging, rooting, or non-binary gene trees), use --maxtrees.

NOTE: For the remainder of the manual, we will use .ntg, the default, as the tree
file extension, but .nwk and .nhx are other options. To change the tree format,
use --treeoutput.

PNG Image of Tree (optional) If the --savepng or --saveweakedgespng option is

given, an image of the tree is saved in PNG format. For more information on saving PNG
images, see Section 12.5 - Saving PNG Images of Trees on page 111.

103
 Chapter 12. Command Line Options and Batch Processing

Pruned Species Tree (optional) In some cases, Notung prunes species that are not
represented in the gene tree from the species tree prior to reconciliation. There are two
possible pruning strategies, described in detail in Section 5.1. When the --stpruned option
is given, the pruned species tree is saved in the file <genetree>.<function>.species.ntg.

Pruned Gene Tree (optional) If the gene tree contains species taxa not represented in
the associated species tree, Notung will print an error message and terminate. If you wish
to ignore these taxa in the gene tree, you can use the --gtpruned option on the command
line. This will prune the gene tree to remove all taxa without a representative in the species
tree. Use caution when using this option, as taxa could be pruned due to incorrect labeling
rather than a true missing species.

Log (optional) When run on the command line, Notung outputs status information to
the terminal window. This information can be saved in a log file
<genetree>.<function>.ntglog by using the --log option. For a batch run, a log file is
not saved for each tree; rather, a single log file for the entire batch run is saved to the file
<batchfile>.<function>.ntglog.

General Tree Statistics (optional) General tree statistics can be saved in the file
<genetree>.<function>.stats.txt by giving the option --treestats. This file includes
information on both the gene tree and the associated species tree. For more information on
tree statistics, see Section 3.4 - General Tree Statistics on page 20.

Event Summary (optional) Information on each duplication, transfer, co-divergence,

and loss is saved in the file <genetree>.<function>.events.txt when the --events option
is used. For each duplication and co-divergence, upper and lower bounds on the time of the
event (represented as nodes from the species tree) are given. For transfers, the donor and
recipient species (represented as nodes in the species tree) are supplied. For losses, each node
in the species tree is listed with the number of losses associated with that taxon. At the end
of the file, a table is provided, where each row represents a node in the species tree and each
number in a row indicates the number of times a duplication, co-divergence, or loss occurred
in that species, as well as the number of times that species acted as a donor or recipient of a
transfer. For more information on the event summary, see Chapter 5.5.1 - Event Summary
on page 51.

Parsable Statistics (optional) When the --parsable option is used, information on in-
ferred events is saved to an easily parsable, tab-delimited file <genetree>.<function>.parsable.txt.

104
Chapter 12. Command Line Options and Batch Processing

To facilitate automated postprocessing by scripts, lines relating to bounds for duplications

and co-divergences are preceded by #D and #CD, respectively; while lines relating to transfer
donors and recipients are preceded by #T. Each line of the loss table, as in Event Sum-
mary, is preceded by a #L. The summary count table over all species and events is denoted
by #S. For more information on the parsable statistics file, see Chapter 5.5.1 - Parsable
Statistics on page 52.

Homology Tables (optional) Notung can output tables of orthologs, paralogs, and
xenologs for all pairs of leaf nodes in the reconciled tree. This table can be generated in
several formats: comma-separated values (CSV), tab-delimited values, or an html-formatted
table. Use options --homologtablecsv, --homologtabletabs or --homologtablehtml, re-
spectively. For more information on orthologs, paralogs, and xenologs, see Section 5.6 -
Inferring Orthologs, Paralogs, and Xenologs on page 53.

12.3 Running Notung from a Batch File

Batch processing allows the user to apply Notung to many trees in a large-scale, automated
analysis. The input trees are given in a batch file, which consists of a list of tree file names,
one per line. Blank lines and lines which start with # are ignored.

To create a batch file:

• The first line of the file should give the path of the species tree to be used.
• Each subsequent line should give the path of a gene tree.
NOTE: By default, Notung expects the tree file locations to be given rel-
ative to the location of the batch file. For example, if the batch file is in
/username/batchRun, Notung expects the gene trees to be in the batchRun
folder or in some subfolder of batchRun. Use the --absfilenames option to
indicate that file names are absolute path names.
NOTE: When using the --savepng option without any of the four main
functions (--reconcile, --root, --rearrange and --resolve), each tree
listed in the batch file is saved as an image. For more information, see
Section 12.5 - Saving PNG Images of Trees on page 111.

A sample batch file is provided with the Notung-2.9 distribution in the sampleTrees/batch
directory. This batch file includes all combinations of binary and non-binary gene and species

105
 Chapter 12. Command Line Options and Batch Processing

trees. Because not all of Notung’s task modes work for each of these combinations, you will
receive one or more warnings and errors when running this batch file. In addition, the batch
file lists a gene tree which does not exist, to give an example of the appropriate warning.

To run Notung from a batch file:

Use the -b <batchfile> option.

For example, from the Notung directory, enter the following on the command line:
java -jar Notung-2.9.jar -b sampleTrees/batch/batch.run --reconcile --speciestag
prefix
The --reconcile option tells Notung to reconcile all the gene trees listed in batch.run
with the species tree listed in batch.run. The --speciestag prefix option tells Notung how
species labels are specified in the gene tree files, and is required in batch mode. See Appendix
A.4 - Specifying the Species Associated with Each Gene on page 129 for more information
on species labels.

NOTE : All gene trees in the same batch file must use the same species tag format,
which is specified using the --speciestag option.

Required Options

In batch mode, the --speciestag option is always required. In addition, when using
--rearrange, --edgeweights and --threshold must be used to set the edge weight lo-
cations and threshold, respectively.

Batch Output

As Notung reads and processes each gene tree in the batch file, it prints diagnostic information
to the terminal. Notung will also print this information to a log file when the --log option
is given. Any errors that occur in the processing of a batch file are reported to the terminal
as they occur. The total number of errors is reported at the end of the batch run.

To print status information to a file:

Use the --log option from the command line. The information will then be written to the
file <batch_file_name>.ntglog.

106
Chapter 12. Command Line Options and Batch Processing

To save trees to a different directory:

By default, Notung saves each reconciled tree to the directory from which the program was
run.

• To save trees to a different directory, use the --outputdir <outputDir> option from
the command line. The information will then be written to the directory <outputDir>.

• Alternately, the --usegenedir option can be used to save output to the directory in
which each gene tree is located.

Progress Bar

For long runs, it may be convenient to use the options --silent and --progressbar together.
This will suppress all output to the terminal with the exception of a simple progress bar to
stderr. The option --log can still be used to save the (now suppressed) output to a file.

12.4 Running a phylogenomic analysis

The evolutionary history of a genome can be considered from the perspective of the combined
evolutionary histories of its constituent gene families. With the growing availability of whole
genome sequence, it is now possible to construct trees for a comprehensive set of gene families
drawn from a given set of genomes. Notung-2.9’s --phylogenomics option aggregates the
events and ancestral states inferred for individual trees to reveal genome scale trends, such
as

• an enrichment of duplications associated with a given ancestral species, suggestive of a

whole genome duplication;

• a wave of losses characteristic of genome reduction;

• co-evolution of gene families with related functions;

• bursts of horizontal gene transfer between particular pairs of species;

• stratigraphic analysis of gene age relative to the species tree.

Wagner parsimony, applied to the number of genes (or gene families) found in present day
genomes, is a common approach to asking such questions. However, Wagner parsimony will

107
 Chapter 12. Command Line Options and Batch Processing

underestimate events and ancestral gene content in families that have sustained parallel gains
or losses. Estimates obtained using reconciliation are more accurate because the inference is
guided by the topologies of the gene trees.

Given a batch file listing a species tree and a set of gene trees, Notung-2.9’s phylogenomic
function reconciles each gene tree and outputs tables summarizing ancestral gene content and
the combined set of duplications, losses, and transfers associated with each node or branch in
the species tree. The following command is the simplest example of a phylogenomics analysis:

java -jar Notung-2.9.jar -b batch.txt --reconcile --speciestag prefix

--phylogenomics

This command will reconcile each gene tree in the batch file and save the resulting reconciled
tree, as described in Section 12.2 - Running Notung from the command line above. In
addition, summary statistics aggregated over all reconciliations will be collected and output
in tab-delimited tables described below.

As with all analyses in --reconcile mode, the --speciestag must be used to specify the
location of the species name in the gene tree leaf labels. Additional command line options can
be used to change default parameter settings (e.g., the event costs) or to generate additional
output files (e.g., --events or --parsable).

Phylogenomic analysis with transfers requires the --infertransfers true option. The
phylogenomics function is only defined for batch analysis in --reconcile mode, executed
from the command line. It is not defined for --rearrange, --root, or --resolve mode. A
Notung-2.9 command that pairs one of these modes with --phylogenomics will generate an
error.

Phylogenomics output

The --phylogenomics command aggregates event statistics over all reconciliations and out-
puts tab-delimited tables with summary statistics. With the exception of the
<batchfile>.highway.txt file, described below, all tables have the same format: one row
for each gene tree in the batch file and one column for each node (including leaf nodes) in
the species tree. If reconciliation is performed with the --infertransfers option, only trees
that have at least one temporally feasible solution (see Chapter 5.2.1 - Temporal Feasibility
in Notung on page 43) are included in the table.

Gene family origins:

108
Chapter 12. Command Line Options and Batch Processing

The <batchfile>.<speciestree>.origin.txt file reports the origin of each gene family;

that is, the species tree branch in which the family was first observed. Stated technically,
the origin is defined to be the species node associated with the root of the reconciled gene
tree. The value of entry i, j in this table is 1 if family i arose in species j and 0 otherwise.

Ancestral gene content:

The <batchfile>.<speciestree>.geneCount.txt table reports the inferred ancestral gene

content in each species tree node. Entry i, j is a non-negative integer representing the number
of members of gene family i present in species j. The total gene content of each species
corresponds to the column sums of this table. The number of gene families that were present
in each species can be determined by counting the number of non-zero entries in each column.

Gain and loss of gene families:

The <batchfile>.<speciestree>.famGainLoss.txt file reports for each gene family, all

branches in the species tree where the family was gained or lost. The value of entry i, j in this
table is 1, if family i was present in species j, but absent in the parent of j; −1, if family i was
present in the parent of j, but absent in j; and 0 otherwise. This table includes the gain at the
origin of the gene family; that is, if entry i, j is 1 in <batchfile>.<speciestree>.origin.txt,
it will also be 1 in <batchfile>.<speciestree>.famGainLoss.txt. To obtain only those
cases where the family was lost and then regained, subtract the entries in
<batchfile>.<speciestree>.origin.txt from the entries in
<batchfile>.<speciestree>.famGainLoss.txt.

Net change in gene family size:

The <batchfile>.<speciestree>.geneGainLoss.txt file reports the change in gene copy

number in each family for every branch in the species tree. The value of entry i, j in this
table is n > 0, if the number of members of gene family i in species j has increased by n,
compared with the number of members in the parent of species j. If n is a negative number,
then the number of members of gene family i in species j has decreased by n, relative to the
count in the parent of j. If entry i, j is zero, then there was no net change.

Duplications:

The <batchfile>.<speciestree>.duplication.txt table reports the number of inferred

duplications in each species tree branch. Entry i, j is a non-negative integer representing the
number of duplications that occurred on the branch from the parent of species j to j.

Losses:

The <batchfile>.<speciestree>.loss.txt table reports the number of inferred losses in

109
 Chapter 12. Command Line Options and Batch Processing

each species tree branch. Entry i, j is a non-negative integer representing the number of
losses that occurred in family i on the branch from the parent of species j to j.

Transfers:

The <batchfile>.<speciestree>.transferFrom.txt table summarizes outgoing transfers.

Entry i, j is a non-negative integer representing the number of transfers in family i that
originated in species j, summed over all possible recipients.

The <batchfile>.<speciestree>.transferTo.txt table summarizes incoming transfers.

Entry i, j is a non-negative integer representing the total number of transfers in family i that
originated from any species and were received by species j.

These tables are only generated when --infertransfers is set to true.

Transfer highways:

The final table generated with the --phylogenomics option has one row for each species
and one column for each species. The <batchfile>.<speciestree>.highway.txt table
summarizes the total number of transfers, summed over all gene families, between pairs of
species. Entry i, j is a non-negative integer representing the number of transfers with donor
species i and recipient species j. This table is only generated with the --infertransfers
true option.

Phylogenomics with averaging

Recall that with transfers, a gene tree may have zero, one, or more than one optimal, tem-
porally feasible reconciliation (see Chapter 4.1.1 - Temporal Feasibility on page 28 for more
details). Only gene trees that have at least one temporally feasible reconciliation are included
in the tables generated by the --phylogenomics option. If a gene tree has more than one
optimal, temporally feasible reconciliation, one of those reconciliations is selected arbitrar-
ily for tabulation in the summary tables. The averaging feature allows the user to extract
information from all optimal, temporally feasible reconciliations.

This command exemplifies the minimal set of options required for a phylogenomics analysis
with averaging:

java -jar Notung-2.9.jar -b batch.txt --reconcile --speciestag prefix

--phylogenomics --avg --infertransfers true

110
Chapter 12. Command Line Options and Batch Processing

The --avg option can only be used with batch analysis in --reconcile mode with the
--phylogenomics and --infertransfers true options. Averaging is only defined when
transfers are included in the event model; with duplications and losses, alone, the optimal
reconciliation is unique.

This command generates the tables described above in Phylogenomics output, but infor-
mation from all optimal reconciliations of each gene tree is used to calculate the summary
statistics in those tables. Specifically, for a given gene tree, the values obtained by av-
eraging over all feasible, optimal reconciliations obtained with that gene tree. Thus, with
--phylogenomics --avg, the entries in these tables are real numbers, in contrast to analyses
with --phylogenomics alone, which generate tables with integer values.

In each table, except the Transfer highways table, entry i, j is the sum of the values
obtained with all optimal, temporally feasible reconciliations of gene family i, normalized
by the total number of such reconciliations. In <batchfile>.<speciestree>.highway.txt,
entry i, j is the average, over all gene families, of the mean number of transfers from species
i to species j, where the mean number of transfers experienced by a gene family is obtained
by averaging over all optimal reconciliations of the gene tree for that family. In other words,
although some gene trees may have more optimal reconciliations than others, each gene family
is weighted equally. As the number of optimal reconciliations increases, the contribution of
each reconciliation decreases accordingly.

Note that the command

java -jar Notung-2.9.jar -b batch.txt --reconcile --speciestag prefix

--phylogenomics --avg --infertransfers true

saves (at most) one reconciled tree for each gene tree in batch.txt, but enumerates all
optimal solutions during the averaging step. This will increasing the running time of the
analysis.

12.5 Saving PNG Images of Trees

The option --savepng saves a simple image representation of a tree in PNG format, while the
option --saveweakedgespng saves a simple PNG representation of a tree with weak edges
highlighted in yellow. A threshold defining weak edges must be given using --threshold
with the latter option; if not, a default of 90% will be used. These options can be used with
one of the four main tasks (--reconcile, --root, --rearrange and --resolve), in which
case an image of the final output tree is saved, in addition to the output tree file. To save

111
 Chapter 12. Command Line Options and Batch Processing

an image of the tree without losses, use the --nolosses command. Alternatively, they can
be used alone to save an image of a tree without performing any other tasks.

Using --savepng and --saveweakedgespng alone

When these options are used without one of the main four tasks, Notung reads in a tree and
generates and saves an image of that tree in PNG format. Unless a batch file is used, only
a single tree can be processed at a time (i.e., a gene tree and a species tree cannot both be
given). If the input tree is a previously reconciled tree in Notung format, the image will show
the appropriate events (to save an image without losses, use --nolosses). If the tree has
not been reconciled, the tree image will show only the structure of the tree and the names
of the leaves of the tree.

When using a batch file, each tree specified in the file is saved as an image. When generating
images without performing a major task, the batch file format differs slightly: Species trees
and gene trees can be listed in any order.

Output File Names

When --savepng or --saveweakedgespng is used alone, an image of the input tree is

saved in the file <treename>.png. When used with --reconcile, --root, --rearrange or
--resolve, an image of the output tree is saved in the file <genetreename>.<function>.png.
For analyses with more than one optimal history, an image file is saved for each history. The
number of files is limited by the parameters --maxtrees and --multsols.

Color Annotations

If a tree in Notung format contains color annotations, the leaves in the image of that tree
will be colored as specified by those annotations. An annotation file can be specified with
the option --annotationfile. For more information on color annotations, see Chapter 10 -
Annotations on page 87.

Making an Imagemap

Notung provides the option to produce an html imagemap for a tree image. If an imagemap
and image file are both included in a web page, each gene in the image will provide a link to
a specified web page. The format of these links is determined by the imagemap specification

112
Chapter 12. Command Line Options and Batch Processing

file given with --imagemapfile <imagemapfilename>, described below. The resulting im-
agemap is saved in the file <outputtreename>.png.html, where <outputtreename> is either
<genetree>.<function> or <treename>.

To include the image and imagemap in a web page, insert the entire contents of the saved im-
agemap file into the html of the web page. The saved image must be in the same directory as
the web page, unless you specify a different location for the image by changing <imagefile>
in the line:

<img border=0 src=’<imagefile>’ ...

Imagemap Specification

The specification file given by --imagemapfile <imagemapfilename> consists of a list of

gene-link pairs. Blank lines and lines that start with # are ignored. Entries creating the
imagemap link follow the format:

# Descriptive comment:
gene: <search string>(id)
link: <link start>(id)<link end>

The <search string> exactly matches strings in the gene tree. When a match is found, the
identifier (id) will capture all remaining text following the matched string. That text can
be used in the link, specified in the ‘link:’ line, thus creating a custom link for that gene.

For each gene in the gene tree, the first gene-link pair that matches will be used. If a gene
does not match any of the ‘gene:’ lines, a warning will be printed. To prevent this warning,
the user can add the following to the end of the imagemap specification file:

# generic imagemap - everything else links to google

gene: (id)
link: http://www.google.com/search?q=(id)

Any leaf in the gene tree that does not match a previous entry will match this entry, and
will link to a Google search for that leaf string.

An example gene tree and imagemap specification from the Princeton Protein Orthology
Database (ppod.princeton.edu/) are included in the Notung distribution. An sample spec-
ification file for this example follows:

113
 Chapter 12. Command Line Options and Batch Processing

# Danio rerio links:

gene: Danio_rerio|(id)
link: http://zfin.org/cgi-bin/ZFIN_jump?record=(id)

# generic imagemap - everything else links to google

gene: (id)
link: http://www.google.com/search?q=(id)

The gene Danio rerio|ZDB-GENE-031007-1 would match the first ‘gene:’ line, with (id)
equal to ZDB-GENE-031007-1 and with link
http://zfin.org/cgi-bin/ZFIN jump?record=ZDB-GENE-031007-1. The gene
Homo sapiens|gene1 would match the second pair, because ‘(id)’ will match any text
string. The resulting link would be http://www.google.com/search?q=Homo sapiens|gene1.

12.6 Inferring Losses when Reconciling with Non-Binary

Species Trees

When inferring losses during reconciliation with a non-binary species tree, it is not possible
to determine unambiguously the edge in the the gene tree to which a loss should be assigned.
Notung uses two different methods to deal with this problem. An exact algorithm finds
all possible assignments that minimize the total number of losses but has exponential time
complexity. A heuristic, which runs in polynomial time, is not guaranteed to find the optimal
assignment, but usually does in practice. These issues and algorithms are discussed in detail
in Section 4.2 - Non-Binary Trees.

Only the heuristic is implemented in the GUI and for the transfer algorithm. Either method
may be used when executing Notung from the command line without the --infertransfers
option. The CLI runs the heuristic by default. To use the exact algorithm, include the
--exact-losses option when running Notung from the command line with the --reconcile
or --root tasks.

The running time of the exact algorithm is exponential in the size of the largest polytomy.
Even when --exact-losses is used, Notung does not apply the exact algorithm to poly-
tomies with more than 12 children. Instead, the heuristic is applied to these polytomies.
To change the maximum polytomy size for which Notung uses the exact algorithm, use
the --polytomy-cutoff <maxPolytomySize> option when including the --exact-losses
option in the command line.

NOTE : Changing the polytomy cut-off to a larger value and using the exact

114
Chapter 12. Command Line Options and Batch Processing

algorithm on a species tree with a polytomy with more than 12 children may
greatly increase running time.

Command Line Options for Losses with Non-Binary Species Trees

--exact-losses
Computes the minimum number of losses when reconciling a binary gene tree with
a non-binary species tree. If this option is not included on the command line, the
heuristic is used. NOTE: In Notung 2.5, this option was named --combine-losses.

--polytomy-cutoff <maxPolytomySize>
Using this option with --exact-losses will change the default value for polytomy cut-
off. Only for losses associated with polytomies less than or equal to <maxPolytomySize>
will the exact algorithm be used. The default value is 12. If a polytomy greater than
<maxPolytomySize> is encountered, a warning will be printed to the terminal window
and/or log file.

--report-heuristic-losses
When run with --exact-losses, this option will report both the number of losses
obtained with the heuristic and with the exact algorithm. This is useful for determining
whether the heuristic is overestimating the number of losses and by how much. NOTE:
In Notung 2.5, this option was named --report-explicit-losses.

12.7 Rooting trees from the command line with the

DTL model

In Root mode, by default Notung finds the set of edges with optimal a root scores, when
--infertransfers is set to “true.” This stands in contrast to the rooting analysis with the
DL model, where the root score is calculated for every edge in the gene tree. With the DTL
model, the analysis is restricted to edges with optimal root scores because the transfer model
incurs the relatively high computational complexity of inferring candidate reconciliations and
testing temporal feasibility.

There may be more than one edge with the minimum root score and there may be more
than one optimal event history for each such edge. Under default settings, the command
line version outputs the minimum root score and saves a reconciled tree that is rooted on
one of the optimal edges, chosen arbitrarily. Various command line options allow the user

115
 Chapter 12. Command Line Options and Batch Processing

to save additional optimal rooted, reconciled trees. To see additional optimal roots, use the
--maxtrees and --allopt flags. The options

--maxtrees <n> --allopt

will save trees rooted on all edges with the optimal root score, or on <n> such edges, if there
are more than <n> optimal roots. If the --maxtrees flag is used without --allopt

--maxtrees <n>

then Notung will save <n> trees rooted on the <n> highest scoring edges with feasible recon-
ciliations. If the number of optimal roots is less than <n>, then trees rooted on edges with
suboptimal scores will be saved.

By default, the rooted trees saved during rooting analysis are reconciled trees in Notung
format. This means that they may be annotated with duplication and transfer events and
will have loss nodes. This is useful for further evaluation or visualization with Notung, but
may cause problems if you plan to further analyze these rooted trees with another application.
To save rooted trees without events and loss nodes, use

--treeoutput newick --nolosses

Because testing multiple candidate solutions for temporal feasibility can significantly increase
running time, the rooting analysis tests the feasibility of candidate optimal reconciliations
until a temporally feasible reconciliation is found. This is sufficient to determine that a tree
rooted on the current edge has at least one feasible reconciliation. As a result, when the
rooting analysis terminates, the optimal root score and edges with at least one minimum
cost, feasible solution are known, but not the number of optimal reconciliations for these
optimal roots.

To see all optimal reconciliations for each optimal root, use the

--maxtrees <n> --allopt

options with --root to save one rooted tree for each optimal root, as described above. Then
reconcile each saved rooted tree with Notung using the

--reconcile --multsols <n>

116
Chapter 12. Command Line Options and Batch Processing

flags to generate all optimal reconciliations for each root. In order to guarantee that you will
obtain all optimal roots (respectively, all optimal reconciliations), the value of <n> should be
greater than or equal to the number of edges in the gene tree. If your gene tree is large, this
may result in very long runtimes.

12.8 Analyzing non-binary gene trees with the DTL

model

Weakly-supported edges, as indicated by low edge weights, and non-binary nodes in the gene
tree often imply that the inferred history associated with those edges/nodes may not be
accurate. Notung can rearrange weakly-supported regions and resolve non-binary nodes in a
gene tree to produce alternate event histories with minimum Event Score. When reconciling
a non-binary gene tree, edges that were added to resolve the polytomy are removed before
saving the reconciled tree. Notung can perform these functions using the DTL event model
on the command-line. See Section 4.2.2 - Fitting a Non-Binary Gene Tree to a Binary Species
Tree on page 35 for more information.

Multiple Optimal Solutions There may be more than one rearrangement or resolution
with the minimal Event Score; in addition, under the DTL model, there may be more than
one optimal event history for a given gene tree rearrangement or resolution. By default,
Notung outputs only one minimal rearrangement/resolution and associated event history,
chosen arbitrarily. As with rooting (Section 12.7), command line options allow the user to
save additional optimal solutions. The option

--maxtrees <n>

will save all rearranged/resolved trees and their histories, or <n> trees and histories if there
are more than <n> optimal solutions.

NOTE : Currenlty, Notung does not distinguish betweeen different optimal topolo-
gies and different optimal event histories for the same topology.

To Rearrange with the DTL model Use the --reconcile command option with the
flag --infertransfers true. Since support for edges is determined by edge weight, No-
tung’s rearrangement function requires that the gene tree include edge weights, and that the
--threshold flag be set. If there are multiple values assigned to edges, you should use the
--edgeweights flag to set the proper edge weight locations.

117
 Chapter 12. Command Line Options and Batch Processing

To Resolve Non-binary Nodes with the DTL model Use the --resolve command
option with the flag --infertransfers true.

To Reconcile a Nonibinary Gene Tree using the DTL model Use the --reconcile
command option with the flag --infertransfers true.

To set the algorithm searching for the best binary representation With the DTL
model, finding the best rearrangement/resolution is NP-hard. Notung provides the user with
multiple options to search for the best topology.

12.9 Command line options

File Input
-g <genetree>
Load the file <genetree> as a gene tree. NOTE : The -g is optional.

-s <speciestree>
Load the file <speciestree> as a species tree. The -s is required.

-b <batchfile>
Load the trees listed in <batchfile>. Requires that the --speciestag option be set. If
rearranging, requires the --edgeweights and --threshold options. With this option,
-g <genetree> and -s <speciestree> should not be specified. See Section 12.3 -
Running Notung from a Batch File on page 105 for more information.

-absfilenames
Files listed in <batchfile> use absolute paths. See Chapter 12.3 - Running Notung
from a Batch File on page 105 for more information.

Tasks
--reconcile
Reconcile a gene tree with a species tree. For transfers, the top <multsols> best scoring
histories are saved in files named <genetree>.reconcile.#.ntg. For non-binary gene
trees, the <maxtrees> optimal histories are saved in files named

118
Chapter 12. Command Line Options and Batch Processing

<genetree>.reconcile.#.ntg. By default, <multsols> and <maxtrees> are set to 1.

In batch mode, --speciestag is required. For more information on reconciliation, see
Chapter 5 - Reconciliation Mode on page 40.
--rearrange
Rearrange the gene tree. The top <maxtrees> best scoring rearrangements are saved in
files named <genetree>.rearrange.#.ntg. By default, <maxtrees> is set to 1. The
option --threshold must be set. In batch mode, --speciestag and --edgeweights
are also required. For more information on rearranging gene trees, see Chapter 7 -
Rearrange Mode on page 73.
--resolve
This task, which removes polytomies from a non-binary gene tree, can only be carried
out if the gene tree is non-binary. The top <maxtrees> best scoring resolutions are
saved in files named <genetree>.resolve.#.ntg. By default, <maxtrees> is set to
1. In batch mode, --speciestag is required. For more information on resolving non-
binary nodes in a gene tree, see Chapter 8 - Resolve Mode on page 81.
--root
Root the gene tree. The top <maxtrees> best scoring rooted trees are saved in files
named <genetree>.rooting.#.ntg. By default, <maxtrees> is set to 1. In batch
mode, --speciestag is required. For more information on rooting gene trees, see
Chapter 6 - Rooting Mode on page 68.

Reconciliation Options

If these options are not given and the gene tree has previously been reconciled, the options
saved with that reconciliation will be used.

--infertransfers [true|false]
If set to true, Notung includes transfers in event inference and optimization. If set
to false, Notung uses the only the duplication-loss event model. This option can only
be set to true with --reconcile and --root, and only for binary gene trees. If this
option is not given and the gene tree has previously been reconciled, the option saved
with that reconciliation will be used. Otherwise, only duplications, co-divergences, and
losses will be inferred.
--ignorelosses
When this option is given, Notung does not consider losses in the event score dur-
ing event inference. Rather, events are inferred by minimizing only duplications, co-
divergences, and transfers. Losses are inferred post-hoc, after these events are assigned.

119
 Chapter 12. Command Line Options and Batch Processing

--prune
By default, Notung does not prune species unrepresented in the gene tree, from the
species tree. Use this option if you would like Notung to construct and use a species
tree that does not contain species that do not appear in the gene tree. If used in batch
mode, the species tree is pruned separately for each gene tree.

--gtpruned
This option will prune from the gene tree, all taxa with a label not found in the provided
species tree. This option should be done separately, not run with other action flags
(i.e., –reconcile, –root, or –rearrange).

--costdup <duplication cost>

Sets the cost of gene duplications. If not set, the cost is set to 1.5, by default.

--costcodiv <co-divergence cost>

Sets the cost of co-divergences. These only occur when reconciling a binary gene tree
with a non-binary species tree. If not set, the cost is set to zero by default. See Chapter
5 - Reconciliation Mode on page 40 for more information.

--costloss <lost gene cost>

Sets the cost of gene losses. If not set, the default cost of 1.0 is used.

--costtrans <transfer cost>

Sets the cost of transfers. These are only inferred if the --infertransfers option is
set to true. If not set, the default cost of 3.0 is used.

Input Data Options

--speciestag [prefix|postfix|nhx]
Indicates the format of species tags in the gene tree. If not set, Notung tries to guess
the correct format. See Appendix A.4 - Specifying the Species Associated with Each
Gene on page 129.

--threshold <threshold>|<percentage>%
Edges with weight higher than <threshold> are preserved during rearrangement. This
can be given as an absolute value or or as a percentage of the maximum value, using
<percentage>%; e.g. “--threshold 90%” sets the threshold at 90 percent of the high-
est edge weight in the tree. See Section 3.5 - Parameter Values on page 22 for more
information.

120
Chapter 12. Command Line Options and Batch Processing

--edgeweights [name|length|nhx]
Indicates where in the tree file the edge weights, if any, are specified. If this option
is not set, and the gene tree has values in more than one location, Notung will guess
the location of edge weights when using --rearrange. See Appendix A.6 - Location of
Edge Weight Values on page 131 for more information.

--bootstraps [name|length|nhx]
Same setting as --edgeweights. Kept for backwards compatibility.

--annotationfile <filename>
Attach the given annotation file to each input tree.

--imagemap <filename>
Used with --savepng. Notung uses the contents of <filename> to create an image
map file, which is saved in <outputtreename>.png.html. For more information, see
Section 12.5 - Saving PNG Images of Trees on page 111.

Output Options
--treeoutput [newick|notung|nhx]
Specify output tree file format. See Appendix A - File Formats on page 126 for more
information. Output trees will have the extension associated with the given format:
.ntg for Notung (default), .nwk for newick, and .nhx for newick extended.

--nolosses
Remove loss nodes from gene trees before they are saved. Useful when outputting tree
in Newick or NHX formats, which do not recognize loss nodes, or with --savepng to
output a tree image without loss nodes.

--maxtrees <maxtrees>
Maximum number of optimal trees to output during reconciliation with non-binary
gene trees, rearrangement, rooting, and resolving. Default is 1.

--allopt
In root mode, the flags --maxtrees <n> --allopt will find up to <n> optimal roots.
If there are more than <n> optimal roots, only <n> rooted trees will be saved. This flag
must be used with --maxtrees.

--multsols <multsols>
Maximum number of optimal trees to output during reconciliation with transfers. De-
fault is 1.

121
 Chapter 12. Command Line Options and Batch Processing

--outputdir <outputDir>
Save output files in the directory, <outputDir>. Default is the current working direc-
tory.
--usegenedir
Save output trees in the directory in which <genetree> is located.
--log
Writes diagnostic output to the file <genetree>.<function>.ntglog.txt, where
<function> is one of the four modes. For batch runs, the log file is saved in
<batchfile>.<function>.ntglog.txt.
--events
Save information on duplications, co-divergences, transfers, and losses in the file
<genetree>.<function>.events.txt. See Section 5.5.1 - Event Summary on page 51
for more information.
--parsable
Saves easily parsed information on duplications, co-divergences, transfers, losses, and
general tree statistics in the file <genetree>.<function>.parsable.txt. See Section
5.5.1 - Parsable Statistics on page 52 for more information.
--treestats
Save general statistics for a tree. Saved in <genetree>.<function>.stats.txt. Statis-
tics on the associated species tree will be included in this file. See Section 3.4 - General
Tree Statistics on page 20 for more information.
--stpruned
Save a version of the species tree that contains only the species found in the gene tree.
Saved in the file <genetree>.<function>.species.ntg.
--rootscores
Report a list of ordered root scores to standard output (only used with --root). This
option is useful for statistical examination of root scores for the gene tree. These scores
can be saved in a file with the --log option.
--silent
Suppresses reporting of diagnostic information to the terminal. This information may
still be printed to a file if the --log option is used.
--progressbar
In batch mode, print a simple progress bar to stderr for each tree analyzed. Useful
with --silent.

122
Chapter 12. Command Line Options and Batch Processing

--savepng
Save the tree as a PNG image. Unlike Notung’s other main functions, this function
does not require a species tree. For more information about --savepng, see Section
12.5 - Saving PNG Images of Trees on page 111.
--saveweakedgespng
Saves the tree with weak edges highlighted in yellow as a PNG. This requires the
--threshold option. If a threshold is not provided, Notung will use the default 90%.
Unlike Notung’s other main functions, this function does not require a species tree. For
more information about --saveweakedgespng, see Section 12.5 - Saving PNG Images
of Trees on page 111.

Homology Tables

For more information on orthologs, paralogs, and xenologs, see Section 5.6 - Inferring Or-
thologs, Paralogs, and Xenologs on page 53.

--homologtablecsv
Save a comma separated table of orthologs, paralogs, and xenologs to the file
<genetreename>.<function>.homologs.csv.
--homologtabletabs
Save a tab-delimited table of orthologs, paralogs, and xenologs to the file
<genetreename>.<function>.homologs.txt.
--homologtablehtml
Save a table of orthologs, paralogs, and xenologs in html format to the file
<genetreename>.<function>.homologs.html. This format can be included in a web
page.

Display Options
--show-species-tree
GUI only: if an input gene tree is reconciled, open the attached species tree in a separate
tab.
--homologgui
GUI only: if an input gene tree is reconciled, start Notung in the Reconciliation tab
with the Show Homology button selected.

123
 Chapter 12. Command Line Options and Batch Processing

Help Message
--help
Print information about these options.

Example commands

java -jar Notung-2.9.jar -g GENETREE.nwk -s SPECIESTREE.nwk

--reconcile --speciestag prefix --infertransfers true --costdup
4.0 --costtrans 6.0 --costloss 2.5 --parsable

This command reconciles the Newick file GENETREE.nwk (see Section A.1 - Newick File Format
on page 127 and see Section A.4 - Specifying the Species Associated with Each Gene on page
129) with the Newick file SPECIESTREE.nwk using custom costs and DTL reconciliation. The
reconciled gene tree in Notung format (the default) will be saved in the file GENETREE.nwk.
reconciled.0.ntg. The --speciestag prefix flag indicates that species names in the gene
tree are given in prefix format (Section A.1 - Newick File Format and Section A.4 - Specifying
the Species Associated with Each Gene). A summary of the inferred events will be saved in
an easily parsable format (5.3) in the file GENETREE.nwk.reconciled.0.parsable.txt (See
Chapter 5 - Reconciliation Mode on page 40).

java -jar Notung-2.9.jar -g "C:\Users\Notung\Desktop\Trees\GENETREE.nwk" -s

"C:\Users\Notung\Desktop\Trees\SPECIESTREE.nwk" --rearrange
--speciestag postfix --threshold 85% --edgeweights length
--treeoutput nhx --log

This command rearranges the gene tree in GENETREE.nwk, where the species are in post-
fix format, with the DL event model. The --edgeweights length flag indicates that edge
weights in the gene tree are stored in the branch length location in the Newick file (see
Section A.6 - Location of Edge Weight Values on page 131). The user specified rear-
rangement threshold is 85% of the maximum edgeweight in the tree. The output will be
saved to GENETREE.nwk.rearranged.0.nhx. Notung diagnostics will be saved to the file
GENETREE.nwk.rearranged.ntglog.txt in nhx format (see ch. 7).

java -jar Notung-2.9.jar -g GENETREE.nhx -s SPECIESTREE.nhx --root

--speciestag nhx --infertransfers true --treeoutput newick

124
Chapter 12. Command Line Options and Batch Processing

This command performs DTL rerooting on GENETREE.nwk informed by SPECIESTREE.nhx.

Species names in the gene tree are stored in the NHX species tag field in the gene tree
file. The top two best-scoring rooted trees will be saved in Newick format in the files
GENETREE.nhx.rooted.0.nwk and GENETREE.nhx.rooted.1.nwk (See Chapter 6 - Rooting
Mode on page 68).

java -jar Notung-2.9.jar -b "C:\Users\Notung\Documents\batch\batch.run"

--speciestag prefix --rearrange --edgeweights name --threshold 65 --log

This command rearranges the gene tree(s) specified in batch.run, which all have the species
names stored in prefix form and edgeweights in name form. The species tree is specified in
the first line of batch.run. The user specified rearrangement threshold is 65. Diagnostic
messages will be saved in the file batch.run.rearranged.ntglog (see Section 12.3 - Running
Notung from a Batch File on page 105).

NOTE : By default, file paths in the batch file will be interpreted as relative to the
location of the batch file; with the --absfilenames flag, file paths are interpreted
as absolute file names.

125
Appendix A

File Formats

Notung can save trees in three different file formats: Newick file format, NHX file
format, and Notung file format.

Newick file format specifies tree topology and node labels, but cannot be used to save rec-
onciliation information or information about the species tree with which the gene tree was
reconciled.

NHX and Notung file formats use the Newick comment field to store additional information
not captured in the standard Newick specification. A reconciliation involves a gene tree, a
species tree, the mapping from gene tree to species tree, and the inferred events. NHX format
can store a gene tree, with additional information to indicate which nodes are duplications.
Notung file format can store a gene tree, the species tree with which it was reconciled,
duplication and loss nodes, and transfer edges with donor and recipient species identified. If
you save a reconciled tree in Notung format, it will display all information stored with the
tree when it is next opened in Notung.

The Notung file format holds more information, but may not be compatible with other
software packages that use Newick format. The formal specification of Newick file format
allows bracket-delimited comments. Programs that follow the formal specification and ignore
information stored in comments will be able to read NHX or Notung format trees. However,
not all programs allow comments. If you plan to use a program that does not allow Newick
comments to further analyze trees saved by Notung, save your trees in standard Newick
format.

126
Appendix A. File Formats

A.1 Newick File Format

Newick is widely used by phylogeny programs. PHYLIP [11], PAUP* [26], and many other
programs will output trees in Newick.

The general Newick syntax looks like this:

treefile → subtree;
subtree → descendant list [internal node label] [:branch length]
descendant list → (subtree, subtree [, subtree]) | leaf node name

where descendant list is a string that specifies the organization of the subtree and
internal node label is the label of the root of a subtree. The optional branch length
field refers to the length of the edge from the root of the subtree to its parent. The
internal node label and branch length fields are optional. Some programs use these
fields to store other information. For example, Notung allows the user to use either of these
fields to store edge weight values.

Comments in Newick format are enclosed in square brackets and may appear anywhere
newlines are permitted. Some programs use the comment field to store additional information
that is not included in the Newick specification. By convention, this information is formatted
as follows:

[&&ApplicationID:Application_specific_comments]

where ApplicationID indicates a specific program or format.

For more information about Newick file format, go to:

http://evolution.genetics.washington.edu/phylip/newicktree.html.

or

http://geta.life.uiuc.edu/~gary/Newicks_845_Tree_Std.html.

127
 Appendix A. File Formats

A.2 NHX File Format - New Hampshire eXtended

NHX File Format is based on the Newick file format, but embeds additional information
about each node in the tree in the comment fields, as follows:

[&&NHX:TagID1=value1:TagID2=value2]

where TagID1 and TagID2 can specify bootstrap values, species labels, or event information.
This example has two tags, but NHX comments can have one or more tags. Trees saved in
NHX file format include information produced by a reconciliation, including duplications,
co-divergences, transfer, and species labels, but do not record any visual annotations made
in Notung. Nor do they record the species tree with which the gene tree was reconciled.

NOTE : The NHX format is case-sensitive.

More information about NHX format, including a complete list of tags used in comment
fields, can be obtained at:

http://www.phylosoft.org/NHX/.

A.3 Notung File Format

Notung File Format further extends the NHX format. Notung file format can record dupli-
cation marks, transfer edges, edge weights, and color annotations. A reconciled gene tree file
saved in Notung format will also have the associated species tree embedded in it. When the
reconciled gene tree is reopened in Notung, the species tree can be extracted and used in
the same way as any other species tree. A reconciled gene tree saved in Notung file format
also stores additional information on parameter values, including edge weight threshold, loss
cost, duplication cost, co-divergence cost, and transfer cost. In addition, a non-binary gene
tree reconciled with a binary species tree with more than one optimal history stores informa-
tion regarding which history was displayed when saved. When the gene tree is reopened in
Notung, the tree for that optimal history will be displayed.

To open an embedded species tree in a Notung format gene tree file:

1. Open the Notung format gene tree file.

128
Appendix A. File Formats

2. Click the Reconciliation tab to enter reconciliation mode.

3. Click the “Show Pruned Species Tree” button.

NOTE : None of the three file formats used in Notung embed alternate histories for gene
trees. Notung saves only the history that currently appears in the tree panel. To access the
other alternate histories when opening such a file, the tree must be analyzed again in Notung.

A.4 Specifying the Species Associated with Each Gene

In order to perform reconciliation, Notung must determine the species from which each leaf
taxon in the gene tree was derived. This is achieved by embedding the species name in the
gene leaf label or by using information embedded in the NHX comment field.

Notung offers three different conventions for specifying the gene to species mapping, described
below. Notung will attempt to guess the naming convention used; you can also specify this
in the reconciliation dialog (see Chapter 5 - Reconciliation Mode on page 40).

• Prefix: When using this format, Notung checks each gene name to see if it begins with
a label in the designated species tree. For example, the gene names HuGST002, Mm GST 5,
and Cow gene1 could correspond to species labels Hu, Mm, and Cow, respectively.

NOTE : When using this format, no species label should be a prefix of another
species label, such as with carp and carpinusBetulus. In this situation,
Notung may incorrectly identify the gene carpinusBetulus gene1 as a carp
gene, rather than a hornbeam gene.

• Postfix: When using this format, Notung looks for underscores ( ) in gene names, and
takes the species name as everything to the right of the last underscore. For example,
gene1 HUMAN and gene3 COW would be labeled as Human and Cow genes respectively.
Notung can distinguish between species names such as carp and carpinusBetulus in
this format.

NOTE : Postfix mode cannot be used if species names include underscores

( ); for example, Carpinus betulus cannot be used in Postfix mode.

• NHX: When using this format, Notung finds the species label in an NHX tag after
the gene name. For example, the gene gene2[&&NHX:S=human] would be labeled as a
human gene. In this case, the species name is not necessarily visible in the gene tree
leaf label, but can be viewed by selecting the “Display Leaf Node Species Name”

129
 Appendix A. File Formats

option in the Display Options menu. Gene trees output by Notung in NHX or Notung
file format are saved using the NHX species tag, even if the original tree used Prefix
or Postfix modes. Notung can distinguish between species names such as carp and
carpinusBetulus in this format.

A.5 Punctuation in Species Names

In previous versions of Notung, punctuation (-, /, _, ., \) in species names was used to

indicate that Notung should look for a shorter species tag in gene names, rather than looking
for the entire species name. For example, given the species name Hu.Homo_Sapiens, Notung
would look for the species label “Hu” in gene names.

Because many users found this confusing, this functionality has been removed since Notung
2.6. Notung now looks for entire species names during reconciliation, which also allows users
to use species names like Pan_troglodytes and Pan_paniscus in the same tree without
creating a conflict. Unfortunately, this means that some trees that were used in previous
versions of Notung will not work in the current version. This section explains how to change
these trees so that they can be used with Notung 2.7.

How do I tell if I need to convert my trees?

Any species tree with punctuation in the species names, where the full species names are not
present in either the gene tree names or in NHX style species tags, will need to be converted.
If your species names contain punctuation and you used them with older versions of Notung,
then your trees probably fit this description. If Notung 2.7 is used to open an older Notung
format tree that needs to be converted, a warning dialog will be shown.

Converting the trees

There are three ways to convert trees with punctuation in species names. The correct method
to use depends on your desired outcome.

Shorten species names This method requires changing only the species tree - gene trees
should not need to be modified. Remove any part of the species name after the first
punctuation, including the first punctuation. For example, if the leaf labels in the gene
tree are of the form “Hu-gene01”, change “Hu.Homo_sapiens” to “Hu” in the species

130
Appendix A. File Formats

tree. These shorter species names should now match the species labels in the gene
names.

Lengthen gene names This method requires changing the gene tree(s). Replace short
species labels in gene names with full species names. For example, change “Hu-gene01”
to “Hu.Homo_sapiens-gene01” in the gene tree. This solution will not work in Postfix
mode if your species names contain underscores ( ).

Add NHX style species tags This method requires changing the gene trees, but does not
change gene names. One benefit of this method is that switching from a very short
species label to a long species label will not affect the length of gene names.
If the gene tree is already in NHX or Notung format, modify the NHX comment after
each gene name. To modify an existing NHX comment, find the species tag and replace
the shorter species label with the full species name. For example, “[&&NHX:S=Hu]”
becomes “[&&NHX:S=Hu.Homo_sapiens]”.
If there are no comments in the file (i.e., the tree is in Newick format), add the follow-
ing after each gene name: “[&&NHX:S=<speciesname>]”, where <speciesname> is the
corresponding full species name from the species tree. For example, the gene tree:

(gene1_Hu,
(gene2_Hu, gene2_Mu));

would become:

(gene1_Hu[&&NHX:S=Hu.Homo_sapiens],
(gene2_Hu[&&NHX:S=Hu.Homo_sapiens], gene2_Mu[&&NHX:S=Mu.Mus_musculus]));

A.6 Location of Edge Weight Values

Notung uses edge weights to determine which edges are weakly supported and may be re-
arranged. These edge weights may correspond to bootstrap values, probabilities, branch
lengths, or any other numerical indication of support.

Edge weight values can be located in one of three places in a tree file, depending on how the
file was created. In Newick format, either the branch length field or the internal node name
may be used to specify edge weights. Many programs store bootstrap values in the Newick
node name field. In an NHX or Notung format file, edge weights can also be specified using
the NHX bootstrap tag in the comment field.

131
 Appendix A. File Formats

The example below shows a tree with a single edge weight in each of the three tree formats:

• Branch length field in Newick format:

(cow_gene1, (mouse_gene2, cow_gene2):100)

• Internal node name field in Newick format:

(cow_gene1, (mouse_gene2, cow_gene2)100)

• NHX bootstrap tag in the comment field:

(cow_gene1, (mouse_gene2, cow_gene2)[&&NHX:B=100])

Confusion can arise if an input tree has edge weights in more than one type of field. This could
occur, for example, in a tree that has both branch lengths and bootstrap values. Notung
tries to guess the type of edge weight specification in the file, but it is not always possible
for Notung to determine this unequivocally. You can specify the location explicitly using
command line options (see Chapter 12 - Command Line Options and Batch Processing on
page 98) or using the “Select Location of Edge Weights” dialog in the Display Options menu
(see Figure A.1).

To set the location of edge weights in Notung:

1. Click “Display Options → Select Location of Edge Weights.” A dialog box

appears.

2. Select one of the radio buttons (see Figure A.1).

The gene tree will immediately reflect the change, so you can check the tree panel to
verify that the choice you selected gives the desired values.

3. Click “Apply.”

132
Appendix A. File Formats

Figure A.1: The “Select Location of Edge Weights” dialog box.

133
Appendix B

Building a Species Tree

Most functions in Notung require a species tree. If you are familiar with the species in your
data set, you may already have an appropriate species tree. If you do not have one, you can
construct one using resources available on the web.

One such resource is the NCBI Taxonomy Browser, available at the NCBI website:

http://www.ncbi.nlm.nih.gov/Taxonomy/CommonTree/wwwcmt.cgi

The Taxonomy Browser contains a database of all organisms represented in the NCBI se-
quence database, and can automatically build a species tree using species selected by the
user. To create a tree in a format Notung can understand, add the species to be included
in the tree, and then use the Taxonomy Browser’s “Save As” option to save the tree as a
Phylip tree. The Phylip option causes the tree to be saved in a variant of Newick format.
The resulting tree can then be loaded into Notung as a species tree.

NOTE : The Taxonomy Browser does not recognize all common species names. Formal names
for species can be found at:

http://www.expasy.org/cgi-bin/speclist

To build a species tree using the NCBI Taxonomy Browser:

1. Go to: http://www.ncbi.nlm.nih.gov/Taxonomy/CommonTree/wwwcmt.cgi.
2. In the text field labeled “Enter name or id,” enter the Latin name or common name of
the species to add to the tree.

134
Appendix B. Building a Species Tree

3. Click “Add.” If the taxonomy browser did not recognize the species name, a pink error
bar which reads “Organism name ’name’ not found” will appear.

4. When you have finished adding species, find the pull-down menu that says “text tree.”
Drag down and select “phylip tree.”

5. Click “Save As,” and save the species tree.

Additional resources provide access to existing species trees built by other researchers. Tree-
BASE (http://www.treebase.org/treebase/search.html) allows users to search for species
trees from a large database of published papers. The Angiosperm Phylogeny Website and
the Phylomatic Project provide species trees for plant species.

http://www.mobot.org/MOBOT/research/APweb/welcome.html
http://www.phylodiversity.net/phylomatic/phylomatic.html

Other tree-building tools are listed on Felsenstein’s Phylogeny Programs website:

http://evolution.genetics.washington.edu/phylip/software.html.

NOTE :

• Species trees obtained from the Taxonomy Browser and other resources will not neces-
sarily be bifurcating. To correct this, you can either edit the tree file in a text editor
or tree editor or use the non-binary species tree with a binary gene tree. See Chapter
4.2 - Non-Binary Trees on page 29 for more information.

• The Phylip tree files generated by the Taxonomy Browser contain non-ASCII characters
and have branch lengths of four within the tree, but these characters have no effect on
the tree when it is opened in Notung.

135
Appendix C

Glossary

Binary Tree: A tree in which every internal node has degree three. If the tree is rooted,
every internal node has one parent and two children. Also known as a Bifurcating
Tree.

Bipartition: In the phylogenetic context, the separation of the leaf taxa into two sets.
Each edge in the tree specifies the taxon bipartition that would result if the edge were
removed. Also called a Split.

Combined Loss: The interpretation that the absence of a gene in two or more sibling
species is evidence of a single loss in their common ancestor.

Co-Divergence: In a gene tree reconciled with a non-binary species tree, a node whose in-
congruence with the species tree could be due to either deep coalescence or duplication.

Co-Divergence Cost: The weight cC of each co-divergence in the DTL Score. By default,
cC = 0.

Duplication/Transfer/Loss Score: The weighted sum, cL L + cD D + cC C + cT T , of losses

(L), duplications (D) co-divergences (C), and transfers (T) in a reconciled gene tree.
Also known as the DTL Score or the D/T/L cost. If transfers are not inferred, it is
the DuplicationLoss or DL Score.

Deep Coalescence: The divergence of genes when the time of separation of a gene lineage
predates the time of speciation.

Duplication Cost: The weight cD of each duplication in the DTL Score. By default, cD =
1.5

Edge Weight: A numerical value representing a quantitative assessment of the support in

the underlying data for the associated bipartition. Typically bootstrap values, branch

136
Appendix C. Glossary

lengths, or likelihood scores are also used as edge weights. If used, edge weights are
specified in the input gene tree file.

Edge Weight Threshold: Numerical value used to define strong edges. Edges with weights
below the edge weight threshold are considered unreliable or weak.

Event History: A set of event-edge pairs, where each event is a duplication, co-divergence,
transfer or loss and each edge is an edge in the species tree that specifies when the
event occurred. An event history specifies a set of trees in which any tree in the set
can be obtained from any other tree in the set by a series of Same Cost Swaps.

Hard Polytomy: A polytomy that represents the simultaneous divergence of the three or
more lineages. Hard polytomies are only found in species trees. See also Soft Poly-
tomy.

Height: The maximum path length between any leaf and the root of a tree.

Incomplete Lineage Sorting: Incongruence between a gene and species tree that occurs
when the lineages of a gene tree sort independently from the lineages of the associated
species tree.

Loss Cost: The weight cL of each loss in the DTL Score. By default, cL = 1.0

Non-binary Tree: A tree in which at least one node has degree greater than three. In a
rooted non-binary tree, at least one node has more than two children. Also known as
a Multifurcating Tree.

Orthologs: Homologous genes whose LCA is found in the cenancestor of the species from
which the genes were sequenced.

Paralogs: Homologous genes whose LCA is associated with a duplication event.

Paraxenologs: Homologous genes that are both paralogs and xenologs.

Polytomy: A node with degree greater than three. In a rooted tree, a node with more than
two children.

Polytomy Size: The number of children of a polytomy in a rooted tree.

Pruned Species Tree: A species tree containing only the species that appear in the gene
tree with which it was reconciled.

Rearranged Tree: A reconciled, binary gene tree with minimal D/T/L Score that agrees
with the original tree at all strongly supported edges.

Reconciled Tree: A gene tree that has been fit to a species tree, resulting in a mapping
between each node in the gene tree and a node in the species tree. From this mapping,
gene losses, duplications, co-divergences, and transfers are inferred.

137
 Appendix C. Glossary

Resolved Tree: A binary tree derived from a non-binary gene tree, in which each polytomy
has been removed and replaced with a set of binary divergences.

Same Cost Swap: An interchange of two nodes in the gene tree that does not change the
DTL Score of the tree, and does not break any Strong Edges.

Soft Polytomy: A polytomy that represents uncertainty in the true, binary branching pat-
tern of its descendant lineages. Soft polytomies can be found in either gene or species
trees.

Strong Edge: An edge with weight greater than or equal to the Edge Weight Threshold.
Any edge without a specified weight is assumed to be weak.

Transfer Cost: The weight cT of each transfer in the DTL score. By default, cT = 3.0.

Trifurcation: In a rooted tree, a node with exactly three children.

Weak Edge: An edge with weight lower than the Edge Weight Threshold. Any edge
without a specified weight is assumed to be weak.

Xenologs: Homologous genes who have undergone a horizontal gene transfer event since
their LCA.

138
Appendix D

Keystroke Shortcuts

Key Combination Action

Ctrl + O Open a gene tree
Ctrl + Shift + O Open a species tree
Ctrl + S Save the tree
Ctrl + P Print the current view
Ctrl + Shift + R Reload tree from file
Ctrl + W Close tree
Ctrl + = Increase font size (for all labels in the tree)
Ctrl + - Decrease font size (for all labels in the tree)
Ctrl + click on tree Zoom in on tree
Shift + click on tree Zoom out of tree
Ctrl + ] Zoom in on tree on the X-axis
Ctrl + [ Zoom out of tree on the X-axis
Ctrl + Shift + ] Zoom in on tree on the Y-axis
Ctrl + Shift + [ Zoom out of tree on the Y-axis
Ctrl + T Show whole tree
Ctrl + . Go to next tree
Ctrl + , Go to previous tree
Ctrl + Q Exit (end Notung)

NOTE : Ctrl indicates use of the control key. Ctrl + click on tree means that the user needs
to click on the tree while pressing the appropriate key. Mac users may have to use the
command, or open apple key to zoom in on the tree (i.e., command + click on tree), but
should use the control key for all other operations.

139
Appendix E

Worked Examples

The following exercises will help familiarize you with the basic tasks Notung can perform on
a gene tree. The tree files used in these exercises are included in the Notung distribution, in
the sampleTrees folder. If the program window becomes too cluttered, you may close trees
that are no longer being used by selecting the tree and clicking on “File → Close.”

E.1 Exercise 1 - Reconciling a gene tree with a species

tree

In this exercise, you will reconcile the gene tree genetree NOTCH with the species tree
speciestree mega. You will also generate a pruned species tree, and use Notung to de-
termine the upper and lower bounds on the time when a duplication occurred.

Open the tree files

1. Click “File → Open Gene Tree” and open genetree NOTCH.

The gene tree is located in the sampleTrees folder, which is included in the downloaded
zip file. Once loaded, the gene tree is displayed in the tree panel.
2. Click “File → Open Species Tree” and open speciestree mega.
The species tree is located in the sampleTrees folder. Once loaded, the species tree
appears in the tree panel. Because it is the most recent tree opened, it is now selected.

Note that the options that Notung offers differ depending on whether a species tree or a gene

140
Appendix E. Worked Examples

tree is selected. For example, because speciestree mega is now selected, the box showing
parameter values in the lower right corner has disappeared, and the task panel includes only
two task modes, History and Annotations.

Reconcile the gene tree with the species tree

1. Click on the genetree NOTCH tab to select the gene tree.

2. Click the “Reconciliation” tab.

The Reconciliation task panel opens below. From here you can reconcile a gene tree
with a species tree, display a pruned species tree, show duplication bounds, and hide
duplication marks and loss nodes.

3. Check the “Prune” box. The leaves of speciestree mega include more species than
are relevant to genetree NOTCH. These species are removed from the tree when this
checkbox is selected.

4. Click “Reconcile/Rereconcile.”
The Reconciliation dialog appears. In this dialog box, Notung asks you to specify
which species tree to use for the reconciliation and what naming convention is used in
the gene tree to specify the species associated with each gene.

5. Select speciestree mega in the drop-down menu labeled “Please select a species tree
to reconcile with.”
Currently, the only selection available is speciestree mega. However, if you have more
than one species tree open in Notung, you must specify here which species tree to use.

6. Under the section labeled “Specify Species Label” select “Prefix of the gene label.”
This section in the dialog box asks you to specify the naming convention used in the
gene tree to indicate from which species the genes originated. Notung tries to guess the
naming convention, but it does not always guess correctly. Notung should have guessed
correctly in this case. In general, remember to check the leaf node names in your gene
tree during this step to make sure that they agree with the naming convention you
choose.
For more details about the species label naming conventions, see Appendix A.4 - Spec-
ifying the Species Associated with Each Gene on page 129.

7. In the dialog box, click “Reconcile.”

141
 Appendix E. Worked Examples

Figure E.1: The gene tree should now look like this.

The reconciled gene tree now appears in the tree panel. The DL Score of the reconciled
tree, displayed in the bottom-left corner of the program window, is 20.5 - five duplica-
tions and thirteen losses. Five red D’s in the tree indicate the inferred duplications. At
the right end of the tree (at the leaves), thirteen loss nodes appear in light gray type.

Display the pruned species tree

After reconciliation, you can view the associated species tree. If the Prune checkbox is
checked, the species tree will be pruned of all species that are not represented by genes in
the gene tree.

1. Click the “Show Pruned Species Tree” button.

A dialog box appears asking you to give a title for the pruned species tree. The default
title is “Pruned Species Tree.”

2. In the dialog box text field, enter “Mega Pruned” (or any other name you like), then
click “OK.”
The pruned species tree appears in the tree panel. It contains only seven leaf nodes, all
of which are species represented in the reconciled gene tree. The pruned species tree
has a tab above the tree panel, labeled “Mega Pruned.” You can now select and use
this tree as you would any other species tree.

142
Appendix E. Worked Examples

Figure E.2: This is what you should see after reconciling and then clicking “show pruned species tree.”

Check the event details

The event summary provides information regarding when events occurred in the course of
species evolution.

1. Select genetree NOTCH.

2. Click “Display Options → Display Internal Node Names.”

Node name labels appear in red type next to each internal node. You can now identify
each duplication by name. If internal node names are not provided in the gene tree file,
Notung will assign the node an alphanumeric name (e.g. n132).

3. Select the Mega Pruned species tree.

4. Click “Display Options → Display Internal Node Names.”

Node name labels appear in red type next to each internal node.

5. Select genetree NOTCH.

6. Click the “Reconciliation” tab.

7. Click on “About This Tree → Event Summary” menu item.

143
 Appendix E. Worked Examples

A new window appears. Inferred duplications are listed in the left column, expressed
as node names in the gene tree. The lower and upper bounds are listed in the middle
and right columns, respectively, and are expressed as internal node names in the species
tree. Information on losses is displayed below duplication bounds. The left column lists
the species nodes in the species tree. The right column provides the number of losses
that occurred in each species.

8. Find the duplication node in the bottom-right area of the tree, from which the XENLAnotch1
gene extends. Find its node name and duplication bounds in the “Duplications and
Losses” window.
The node name may vary, depending on how many internal nodes Notung has counted
in your current session.

9. Close the window and select the pruned species tree Mega Pruned.
With Mega Pruned selected, you can see internal nodes representing euteleostomi and
coelom. The duplication occurred somewhere on the edge between those nodes.

E.2 Exercise 2 - Rooting an unrooted tree

The gene tree genetree ANK is unrooted. In this exercise, you will select a root based on
duplication loss parsimony.

Open the tree files

1. Click “File → Open Gene Tree” and open genetree ANK.

The gene tree is located in the sampleTrees folder.
Since this tree is unrooted, it has a trifurcation (a node with 3 children) at the top of
the tree, but is otherwise binary.

2. Click “File → Open Species Tree” and open speciestree mega. If speciestree mega
is already opened, you may skip this step.

3. Be sure that the genetree ANK tab is selected before proceeding.

Run the Rooting Analysis

1. Click the “Reconciliation” tab and be sure the ”Prune” option is checked. This will
ensure that our species tree is pruned before it is reconciled with this gene tree.

144
Appendix E. Worked Examples

2. Click the “Rooting” tab.

The Rooting task panel is displayed. Notung is now in Rooting mode.

3. Click “Run Rooting Analysis.”

You will be asked to reconcile the tree. Select speciestree mega and “Prefix,” click
“Reconcile”. The edge at the top of the tree panel, leading to caeel*unc-44, is
colored red. This means it has the minimum root score.

4. Optional: Click the “Display root score” checkbox.

Each edge is labeled with its root score. Notice that the red edge leading to caeel*unc-44
has a root score of 4.0. The next lowest score is 8.5.

Select a root

1. Click on the red edge in the tree panel.

The tree is now rooted on the edge leading to the caeel*unc-44 gene. The DL Score
of the tree is now 4.0, with two duplications and one loss.

Figure E.3: The gene tree should now look like this.

145
 Appendix E. Worked Examples

E.3 Exercise 3 - Rearranging a gene tree

In this exercise, you will reconcile the gene tree genetree SMALL with the species tree
speciestree small and use Notung’s rearrangement tasks to investigate alternate gene trees
with minimum DL Score. Both input trees are located in the sampleTrees folder.

Reconcile the gene tree with the species tree

1. Click “File → Open Species Tree” and open speciestree small.

2. Click “File → Open Gene Tree” and open genetree SMALL.
This is an artificial tree made up for this exercise. The edge weights in this tree
represent bootstrap values. Note that two internal edges have a bootstrap value of 100,
one has a bootstrap value of 73, and several have not been assigned a weight. (Note
that edges adjacent to leaves are usually not assigned bootstrap values since those edges
are present in all trees.) Notung sets the default edge weight threshold to 90% of the
maximum edge weight in the tree. Since the maximum edge weight in this tree is 100,
the edge weight threshold is set to 90.0.
3. Click the “Reconciliation” tab.
4. Click “Reconcile/Rereconcile.”
5. In the “Reconciliation Options” dialog box, select speciestree small and “Postfix”
and click “Reconcile.”
The reconciled tree appears in the tree panel. Note that it has a DL Score of 10.0, with
four duplications and four losses.

Rearrange the reconciled tree

1. Click the “Rearrange” tab.

The Rearrange task panel is now displayed.
2. Click the “Highlight weak edges” checkbox.
Several edges in the reconciled tree are highlighted in yellow. These are edges with
weights below the Edge Weight Threshold and are considered “weak.” Weak edges
may be rearranged to reduce the number of duplications and losses in the tree. Edges
with weights above the threshold will not be rearranged.
Note that in addition to the edge with weight 73.0, the internal edges with no edge
weight are also highlighted in yellow. Notung assumes that any internal edge that is
not explicitly assigned a weight is considered weak.

146
Appendix E. Worked Examples

Figure E.4: The gene tree with weak edges highlighted.

3. Click “Perform Rearrangement.”

The rearranged tree appears in the tree panel. It now has a D/L Score of 4.0, with two
duplications and only one loss.

147
 Appendix E. Worked Examples

Figure E.5: The gene tree should now look like this.

Change the parameter values and rearrange again

In the previous steps, we rearranged the tree using the default parameter values (cD = 1.5
and cL = 1.0). For the default values, there is only one minimum cost tree. We now explore
what happens when we rearrange the tree when duplications and losses are weighted equally.

1. Click the “Edit Values” button in the bottom-right corner of the program window.

2. In the dialog box, change the Duplication Cost to 1.0.

3. Click “Apply Changes.”

A message appears to warn us that although we have changed the parameter values,
this has had no effect on the tree. We must rearrange the tree again to see the effect
of rearrangement with this choice of parameter values. Click “OK.”
Duplications and losses are now weighted equally in Notung’s reconciliation algorithm.

4. Click the “Rearrange” tab, if it is not already selected.

5. Click “Perform Rearrangement.” The tree is rearranged with the new parameter
values. The newly rearranged tree appears in the tree panel. The DL Score of this tree
is 3.0, with three duplications and no losses.

148
Appendix E. Worked Examples

Figure E.6: The gene tree should now look like this.

View a different alternate event history

With the new parameter values, there is more than one alternate gene tree with minimal DL
Score. You are currently viewing history 0.

1. In the Rearrange task panel, click on the drop-down menu labeled “Select an optimal
event history.”
This opens a list of available alternate event histories. You should see history 0 and
history 1.

2. Select history 1.
A different tree appears. This tree also has a D/L Score of 3.0, but has two duplications
and one loss instead of three duplications and no losses.

Swap nodes in the rearranged tree

Note that this tree groups gB human with gA mouse and gA human with gB mouse. However,
the tree that groups gA human with gA mouse and gB human with gB mouse has the same
score.

149
 Appendix E. Worked Examples

Figure E.7: The gene tree should now look like this.

1. Click the “Examine same cost swaps” button.

Nodes that can be interchanged without changing the DL Score are marked with en-
larged light blue boxes.

2. Click the gB human node.

To select the node, you must click on the enlarged blue box. When you click and select
a node, a blue triangle will mark the node(s). Once selected, the node is marked with
a light blue triangle. Each node it can be swapped with is marked with a pink triangle.
In this case, there is just one: gA human.

3. Click the gA human node.

The nodes gB human and gA human are swapped. Once they have been swapped, they
are temporarily highlighted with yellow triangles, so that you can see the results of the
most recent action. Note that the gA genes are now grouped together, and the gB genes
are together in the same subtree, along with the g gorilla gene.
Try performing additional swaps to see how many alternate, minimum cost trees you
can find.

150
Appendix E. Worked Examples

E.4 Exercise 4 - Inferring duplications and losses in a

binary gene tree with a non-binary species tree

In this exercise, you will perform Notung’s main tasks to infer duplications and losses on the
gene tree exercise4 genetree with the non-binary species tree exercise4 speciestree.
You will reconcile and root the gene tree, and use Notung to determine the upper and lower
bounds on the time when a duplication occurred.

Open the tree files

1. Click “File → Open Gene Tree” and open exercise4 genetree.

This is an artificial tree made up for this exercise.

2. Click “File → Open Species Tree” and open exercise4 speciestree.

As you will notice, this is a non-binary species tree with a polytomy representing the
common ancestor of the marsupials.

Reconcile the gene tree with the species tree

1. Select exercise4 genetree.

2. Click the “Reconciliation” tab.

3. Click “Reconcile/Rereconcile.”

4. In the “Reconciliation Options” dialog box, select exercise4 speciestree and “Pre-
fix” and click “Reconcile.”

The reconciled tree appears in the tree panel. Note that it has a DL Score of 8.0, with
two duplications, one co-divergence, and five losses. Two red D’s in the tree mark the
duplications, while the one pink cD marks the co-divergence. At the leaves of the tree,
five loss nodes appear in light gray type.

151
 Appendix E. Worked Examples

Figure E.8: The gene tree should now look like this.

Check the event summary

The event summary provides information regarding when events occurred in the course of
species evolution.

1. Click “Display Options → Display Internal Node Names.”

2. Select exercise4 speciestree.

3. Click “Display Options → Display Internal Node Names.”

4. Select exercise4 genetree.

5. Click the “About This Tree → Event Summary” from the menu.
In the new window, duplications are described first. Co-divergences are described below
the duplications. For both types of events, the gene nodes are listed in the left column,
expressed as node names in the gene tree. The lower and upper bounds are listed in
the middle and right columns, respectively, and are expressed as internal node names
in the species tree. Information on losses is provided below the co-divergence bounds.

152
Appendix E. Worked Examples

6. Close the window by clicking the “Close this window” button.

Run the Rooting Analysis

1. Click the “Rooting” tab.

2. Click “Run Rooting Analysis.”

The edge leading to genes from placental mammals (cow, mouse, and human) is colored
red. This means it has the lowest root score.

3. Under the section labeled “Specify Species Label” select “Prefix of the gene label.”

4. In the dialog box, click “Reconcile.”

5. Optional: Deselect “Display Internal Node Names” (under “Display Options”) and
click the “Display root score” checkbox.

Notice that the red edge has a root score of 7.0. The next lowest root score is 8.0.

6. Click “Display Options → Display Internal Node Species Names.”

The name of the species to which the node is mapped appears in italics next to each
internal node.

7. Select the optimal root by clicking on the red edge in the tree panel.

The tree is rooted on the edge which splits the tree between placental mammals (Eu-
theria) and marsupials (Metatheria). The DL Score of the tree is now 7.0, with two
duplications, one co-divergence, and four losses.

Do not close these trees yet - they will be used in upcoming steps.

153
 Appendix E. Worked Examples

Figure E.9: The gene tree should now look like this.

Reconcile the Tree using the Combined Polytomy Losses algorithm

This step uses the command line interface and can be skipped, if desired. You will use the
command line interface to reconcile the gene tree exercise4 genetree with the species tree
exercise4 speciestree using the combined losses algorithm.

1. On the command line, navigate to the Notung directory.

For instructions on using Notung from the command line, see Chapter 12.2 - Running
Notung from the command line on page 102.

2. Type the following in the command window/terminal and hit enter:

java -jar Notung-2.7.jar sampleTrees/exercise4 genetree -s
sampleTrees/exercise4 speciestree --reconcile --exact-losses
--outputdir sampleTrees --report-heuristic-losses
Notung will print information to the screen as it reconciles the tree for both com-
bined and explicit losses. Notice that the first unrooted gene tree has a DL Score
of 8.0, with two duplications, one co-divergence and five heuristic losses as compared

154
Appendix E. Worked Examples

to the second unrooted gene tree, which has a DL Score of 7.0, with two duplica-
tions, one co-divergence, and four exact losses. The tree, reconciled and with ex-
act losses, will be saved to the sampleTrees folder (as specified by --outputdir) as
exercise4 genetree.reconciled.

Root the tree reconciled with the Combined Polytomy Losses algorithm

In the previous step, you reconciled the gene tree while using the combined polytomy losses
algorithm. In this step you are will find the optimal root for this gene tree. If you skipped
the previous step, you will need to use the gene tree exercise4 genetree-exactLosses.ntg
instead of exercise4 genetree.reconciled.

1. In Notung’s graphical user interface, click “File → Open Gene Tree” and open
exercise4 genetree.reconciled.

If you skipped the last step, use exercise4 genetree-exactLosses.ntg instead.

A warning will appear stating that the tree was reconciled using --exact-losses.
Click the “OK” button.

2. Click the “Reconciliation” tab.

3. Click the “Show pruned species tree” button.

A window will appear, and you can enter a title for the pruned species tree. Click the
“OK” button.

Switch back to the gene tree.

4. Click the “Rooting” tab.

5. Click “Run Rooting Analysis.”

A window will appear asking the position of species label. Choose the third option
“NHX species tag contains the species label”, and then click “OK”

6. Select the optimal root by clicking on the red edge in the tree panel.

The tree is rooted on the edge leading to placental mammals. The DL Score of the tree
is now 6.0, with two duplications, one co-divergence, and three losses.

155
 Appendix E. Worked Examples

Figure E.10: The gene tree should now look like this.

Compare this tree with the previously rooted gene tree (exercise4 genetree). Can
you find the difference between the trees? In exercise4 genetree, the loss node,
tasmanian devil*LOST, above the subtree containing genes gene3 and gene2, has been
moved below the duplication node and combined with opossum*LOST and bandicoot*LOST
in the gene3 and gene2 subtrees, respectively, in exercise4 genetree.reconciled.
This resulted in a reduction of the total number of losses.

View polytomy losses without species names included

There are two display options for polytomy losses. In this step, you will see the other way
to display these losses.

1. Select the exercise4 genetree.reconciled gene tree. (Use

exercise4 genetree-exactLosses.ntg if you skipped the step for reconciling the tree
using the exact losses algorithm.)

2. Deselect the “Display Options → Use Species Names in Polytomy Losses”

option. The gene tree no longer shows the species names for polytomy losses. For
example the loss that was previously displayed as “[tasmanian devil, bandicoot]
of Metatheria*LOST” is now displayed as “2/4 of Metatheria*LOST.”

156
Appendix E. Worked Examples

Figure E.11: The gene tree should now look like this.

E.5 Exercise 5 - Non-binary gene tree with a binary

species tree

In this exercise, you will perform Notung’s main tasks to infer duplications and losses on the
non-binary gene tree exercise5 genetree with the species tree exercise5 speciestree.
You will reconcile, root, resolve, and rearrange the gene tree, and use Notung to determine
some general statistics about the trees.

Open the tree files

1. Click “File → Open Gene Tree” and open exercise5 genetree.

This is an artificial tree made up for this exercise. Notice that this gene tree is non-
binary and contains multiple polytomies.

2. Click “File → Open Species Tree” and open exercise5 speciestree.

3. Select exercise5 genetree.

157
 Appendix E. Worked Examples

4. Click “Display Options → Highlight Polytomies.”

The polytomies in the gene tree are circled and highlighted in cyan.

Figure E.12: The gene tree with polytomies highlighted.

Reconcile the gene tree with the species tree

1. Click the “Reconciliation” tab.

2. Click “Reconcile/Rereconcile.”

3. In the “Reconciliation Options” dialog box, select exercise5 speciestree and “Pre-
fix” and click “Reconcile.”

The reconciled tree appears in the tree panel. Note that it has a DL Score of 20.0,
with ten duplications and five losses. Also note that some of the polytomies have more
than one duplication associated with the node (ex: the polytomy with eight children
has two duplications).

158
Appendix E. Worked Examples

Figure E.13: The gene tree should now look like this.

Get general tree statistics for the gene tree

In this step you will gather some general statistics about the reconciled gene tree and the
species tree.

1. Click on “About This Tree → General Tree Statistics”

The General Tree Statistics window appears. In this window is information on both
the gene tree, the reconciled gene tree, and the species tree. You may have to scroll
down to view all the information.

The General Tree Statistics Window should look like this.

159
 Appendix E. Worked Examples

Figure E.14: The General Tree Statistics Window should look like this.

2. After reviewing this information, click “Close this window” to close the window and
continue.
For more information on the data in the General Tree Statistics window, see Chapter
3.4 - General Tree Statistics on page 20.

Resolve the polytomies in the gene tree

In this step, you will resolve all the polytomies in the gene tree, thus creating a binary gene
tree.

1. Click the “Resolve” tab.

The Resolve task panel opens below.

2. Make sure that the “Highlight polytomies” checkbox is selected.

The polytomies in the gene tree are circled and highlighted in cyan.

3. Click “Resolve Polytomies.”

The resolved tree appears in the tree panel. Edges associated with the resolved poly-
tomies are now colored cyan. This is the same tree as before, only now the polytomies
have been resolved. The number of duplications and losses are identical to the recon-
ciled tree, and even the duplication bounds are the same.

160
Appendix E. Worked Examples

Figure E.15: The gene tree should now look like this.

Change the parameter values and view alternate event histories

In the previous steps, we reconciled and resolved the tree using the default parameter values
(CD=1.5 and CL=1.0). For the default values, there is only one minimum cost tree. We now
explore what happens when we reconcile the tree when duplications and losses are weighted
equally.

1. Click the “History” tab.

We must go back in the history before we change parameter values, as the tree has
already been resolved and the change in values might effect the current resolution of
the tree.

2. Go back to the pre-resolved step in the history by clicking “Reconciled with exer-
cise5 speciestree.”
The tree panel shows the state of the tree before the polytomies were resolved.

3. Click the “Edit Values” button in the bottom-right corner of the program.

4. In the dialog box, change the Duplication Cost to 1.0.

5. Click “Apply Changes.”

Duplications and losses are now weighted equally, and the gene tree is automatically
re-reconciled with the new parameter values.

161
 Appendix E. Worked Examples

6. Click the “Reconciliation” tab.

The reconciled tree appears in the tree panel. There is now more than one alternate
gene tree with the minimal DL Score. You are currently viewing history 0.
7. Click on the drop-down menu labeled “Select an optimal event history.”
8. Select history 1.
A different tree appears. This tree has a DL Score of 15.0, with ten duplications and
five losses. This tree has the same duplications and losses as the tree reconciled with a
duplication cost of 1.5 and a loss cost of 1.0 (see Figure E.14).
9. Click on the drop-down menu labeled “Select an optimal event history” and select
history 0.
A different tree appears. This tree also has a DL Score of 15.0, but has eleven duplica-
tions and four losses rather than the ten duplications and five losses in history 1. The
large polytomy with seven children now has three duplications and one loss, whereas
in history 1 it had two duplications and two losses.

Figure E.16: The gene tree should now look like this.

Run the Rooting Analysis

1. Click the “Rooting” tab.

2. Click “Run Rooting Analysis.”
Many edges and one polytomy are colored red, which indicates that all of these com-
ponents of the tree have the lowest root score.

162
Appendix E. Worked Examples

Notice that the large polytomy is circled in red. Placing a root at a polytomy indicates
that at least one edge in the binary resolution of the polytomy has the lowest root
score.

Figure E.17: The gene tree should now look like this.

3. Optional: Click the “Display root score” checkbox.

Each edge and polytomy is labeled with its root score.

4. Select an optimal root by clicking on the polytomy with the red circle.

The tree is rooted on the polytomy and the DL Score of the tree is still 15.0, with
eleven duplications and four losses.

163
 Appendix E. Worked Examples

Figure E.18: The gene tree should now look like this.

Resolve the polytomies in the gene tree

In this step, you will resolve all the polytomies in the gene tree, thus creating a binary gene
tree.

1. Click the “Resolve” tab.

The Resolve task panel opens below.

2. Make sure that the “Highlight polytomies” checkbox is selected.

The polytomies in the gene tree are circled and highlighted in cyan.

3. Click “Resolve Polytomies.”

The resolved tree appears in the tree panel. Edges associated with the resolved poly-
tomies are now colored cyan.

164
Appendix E. Worked Examples

Figure E.19: The gene tree should now look like this.

View a different alternate event history

With these parameter values, there is more than one alternate gene tree with minimal DL
Score. You are currently viewing history 0.

1. Click on the drop-down menu labeled “Select an optimal event history.”

This displays a list of available alternate event histories. You should see history 0
and history 1.

2. Select history 1.
A different tree appears. This tree also has a DL Score of 15.0, but has ten duplications
and five losses instead of eleven duplications and four losses.
Note that these alternate histories correspond to the same alternate histories that were
presented after reconciliation.

Swap nodes in the resolved tree

Note that this tree groups human-gene-BB1 with mac-gene-BB2 and human-gene-BB2 with
mac-geneBB1. However, the tree that groups human-gene-BB1 with mac-geneBB1 and human-gene-BB2
with mac-gene-BB2 has the same score.

165
 Appendix E. Worked Examples

1. Click the “Examine same cost swaps” button.

Nodes that can be interchanged without changing the DL Score or history implied by
the polytomies are marked with enlarged light blue boxes.

2. Click the node for mac-gene-BB1.

The node is now marked with a light blue triangle. Each node it can be swapped with
is marked with a pink triangle. In this case, there is just one: the node leading to
mac-gene-BB2.

3. Click the node for mac-gene-BB2.

The nodes mac-gene-BB1 and mac-gene-BB2 are swapped. Once they have been
swapped, they are temporarily highlighted with yellow triangles, so that you can see the
results of the most recent action. Note that the BB1 genes are now grouped together,
and the BB2 genes are together in the same subtree.

Figure E.20: The gene tree should now look like this.

Annotate the Gene Tree

This step will introduce you to Notung’s annotations capabilities.

1. Click the “Annotations” tab.

The Annotations task panel is displayed.

166
Appendix E. Worked Examples

2. Click on the “New” button to add a new annotation.

A box will appear to edit the new annotation.

3. In the space labeled “Please enter a title for the annotation”, type in “-A” Select a
color from the palate and click “OK”.
This will automatically annotate all the leaves that contain the string “-A” with the
color you selected.

4. Click on the “New” button. In the space labeled “Please enter a title for the annota-
tion,” type in “-BA” and select a color from the palate and click “OK.”
This will automatically annotate all the leaves that contain the string “-BA” with the
color you selected.

5. Click on the “New” button. In the space labeled ‘Please enter a title for the annota-
tion,” type in “BB1” and select a color from the palate and click “OK”.
This will automatically annotate all the leaves that contain the string “BB1” with the
color you selected.

6. Click on the “New” button. In the space labeled “Please enter a title for the annota-
tion,” type in “BB2” and select a different color from the palate and click “OK”.
This will automatically annotate all the leaves that contain the string “BB2” with the
color you selected.

7. Click on the “New” button. In the space labeled “Please enter a title for the anno-
tation,” type in “BA 4, 5, 6” and select a different color from the palate and select
the button labeled “I want to manually select the nodes and subgroups to add.” Click
“OK”.
This option lets you select the nodes to add to the annotation without searching for a
substring.

8. Click on the node leading to the subtree with genes BA4, 5, and 6 in humans.
Notice that these leaves were previously in the color selected in step 3. The leaves are
a new color now because the newer annotation takes precedence.

9. Click on the “New” button. In the space labeled “Please enter a title for the anno-
tation,” type in “BA 1, 2, 3” and select a different color from the palate and select
the button labeled “I want to manually select the nodes and subgroups to add.” Click
“OK.”

10. Click on the node leading to the subtree with genes BA1, 2, and 3 in humans.

167
 Appendix E. Worked Examples

Figure E.21: The gene tree should now look something like this.

Rearrange the resolved tree

In this step, you will rearrange the gene tree to obtain the minimal DL Score. In this exercise,
you have resolved the polytomies in the gene tree before rearranging the weak areas of the
tree. However, it is possible to do both task at the same time while in the rearrangement
mode. Both Resolve and Rearrangement are available because these two functions have
different purposes. If you want to obtain a hypothesis of the binary gene tree, but wish
to retain all the information in the gene tree, use the Resolve task mode. However, if you
wish to consider edges with an edge weight below a certain value as uninformative, use the
Rearrangement task mode.

1. Click the “Rearrange” tab.

2. Click the “Highlight weak edges” checkbox.

Several edges in the reconciled tree are highlighted in yellow. These are edges with
weights below the Edge Weight Threshold and are considered “weak.” Weak edges
may be rearranged to reduce the number of duplications and losses in the tree. Edges
with weights above the threshold will not be rearranged.

168
Appendix E. Worked Examples

Figure E.22: The gene tree with weak edges highlighted.

3. Click “Perform Rearrangement.”

The rearranged tree appears in the tree panel. It has a DL Score of 15.0, with twelve
duplications and only three losses. Note that the score did not change; the rearranged
tree is not necessarily “better” than the original tree.

169
 Appendix E. Worked Examples

Figure E.23: The gene tree should now look like this.

View a different alternate event history

You are currently viewing history 0.

1. Click on the drop-down menu labeled “Select an optimal event history.”

This opens a list of available alternate event histories. You should see history 0,
history 1, and history 2.

2. Select another history and examine the same cost swaps by clicking the “Examine
same cost swaps” button.
Nodes that can be interchanged without changing the DL Score are marked with en-
larged light blue boxes. Try performing additional swaps to see how many alternate,
minimum cost trees you can find.

3. See if you can find the original tree by changing the histories and examining same cost
swaps.
HINT 1: Select the history with ten duplications and five losses.
HINT 2: Swap the subtree of BA1 and BA2 genes in “pan” with the LOST “pan” gene
in the BA subtree.
HINT 3: Swap the subtree of BA4, BA5, and BA6 in human with the node for BA3 in
human.

170
Appendix E. Worked Examples

E.6 Exercise 6 - Inferring transfers in a binary gene

tree with a non-binary species tree

In this exercise, you will use Notung to infer transfers with duplications, losses, and ILS in
the gene tree, genetree-bacteria with the non-binary species tree speciestree-bacteria.

Open the tree files

1. Click “File → Open Gene Tree” and open genetree-bacteria.

This is an artificial tree made up for this exercise. Notice that this gene tree is non-
binary and contains multiple polytomies.

2. Click “File → Open Species Tree” and open speciestree-bacteria.

Reconcile the gene tree with the species tree

1. Select genetree-bacteria.

2. Click “Reconciliation” tab.

3. Click “Reconcile/Rereconcile.”

4. In the “Reconciliation Options” dialog box, select speciestree-bacteria and “Pre-

fix” and click “Reconcile.”

This tree has been reconciled with the duplication-loss algorithm. Note that it has a
DL score of 23with 4 duplications, 17 losses, and 1 co-divergence.

5. Check the “Infer Transfers” box.

Now when you reconcile the tree you will include transfer inference.

6. Click “Reconcile/Rereconcile.”

7. A message will appear saying “No temporal feasible solutions found.”

171
 Appendix E. Worked Examples

Figure E.24: A warning will pop up when there’s no temporal feasible solution

Click “OK.” This means that with these costs and this root, all optimal solutions were
temporally infeasible. For an explanation of this error, see Chapter 5.2.1 - Temporal
Feasibility in Notung on page 43.

Run the Rooting Analysis

Let’s see if another root will produce a temporally feasible solution.

1. Click “Rooting.”

2. Click “Run Rooting Analysis.”

3. In the “Reconciliation Options” dialog box, select speciestree-bacteria and “Pre-

fix” and click “Reconcile.”

4. Click the “Display root score” checkbox.

172
Appendix E. Worked Examples

Figure E.25: Root scores will be displayed on the gene tree branches

Optimal edges that have been checked for temporal feasibility will be labeled with a
roots score. Notice the current root edge has a score of inf. This means there are no
temporally feasible solutions for this root, and the root score is infinity. see Chapter 6
- Rooting Mode on page 68 for more information. The edge(s) with the lowest score
and at least one feasible solution are highlighted in red.

5. Notung is suggesting that the gene tree be rerooted on the red branches. Click on the
red branch that is closer to the root (left most).

173
 Appendix E. Worked Examples

Figure E.26: What the gene tree will be once rooted at one of the red branches

The tree is now rooted and has a DTL score of 12with 0 duplications, 3 transfers, 3
losses and 1 co-divergence.

View alternate event histories

1. Click on the “Reconciliation” tab.

Notice that some nodes are circled and colored green. This indicates areas that will be
affected by alternative optimal histories. see Chapter 5 - Reconciliation Mode on page
40 for more information.

2. Click on any circled node.

174
Appendix E. Worked Examples

Figure E.27: Green circles on the nodes indicate alternative event history, and are clickable

This will cycle through all other histories below that node. Clicking on the circled
node(s) closest to the root will cycle through all the histories. The total number of
alternative optimal solutions is indicated in the status bar.

Get event details for the history

1. Click on “About This Tree → Event Summary” menu item.

175
 Appendix E. Worked Examples

Figure E.28: Events Summary window

The “Event Summary” window appears. You will see the DTL score for the current
history. Below that, you will see a table listing gene duplication events. The third
table lists all the transfers that Notung inferred. The first column lists the name of
the ancestral gene which was transferred. The “From” column lists the donor species
that is the source of the transferred gene, while the “To” column lists the name of the
ancestral recipient species. The next table lists the number of times each species lost a
gene. The final table provides a summary of all events involving each species.

2. Close the window by clicking the “Close this window” button.

176
Appendix F

Troubleshooting

Problem Possible Causes Solutions

When I tried to reconcile • The species labels in the • Check the species labels in
the trees, I received this er- gene tree leaf node names the gene tree to make sure
ror message: “None of the are not compatible with they match the species la-
species labels in this tree the species labels in the bels in the species tree.
can be found in the species species tree.
tree. Try checking your rec- • In the Reconciliation Op-
onciliation settings.” • The “Specify Species La- tions dialog, make sure
bel” setting in the Rec- you select the appropri-
onciliation Options dialog ate naming convention for
box has been set incor- species labels.
rectly.
• In the Reconciliation Op-
• The incorrect species tree tions dialog, make sure
has been selected for rec- you select the appropriate
onciliation. species tree for reconcilia-
tion.

The edge weights on the Notung has mistaken the First, open the gene tree file
gene tree are not what you branch length values in the in a text editor to deter-
expected. Newick file for edge weight mine the location of edge
values. See Appendix A.6 weight values. Then, click
- Location of Edge Weight “Display Options → Se-
Values on page 131 lect Location of Edge
Weights” and set the loca-
tion of Edge Weights appro-
priately.

177
 Appendix F. Troubleshooting

Problem Possible Causes Solutions

My gene tree should have The gene tree file is sup- Open the original tree file
edge weights, but when posed to be in Newick, NHX in a tree editing program
I load the tree, weights or Notung format, but con- or text editor and cor-
are not displayed on some tains a typo or format- rect any formatting errors.
branches. ting error, affecting the edge NOTE : Some formats are
weight location. case-sensitive.
When I tried to recon- You have opened a species Reopen the desired species
cile the gene tree with the tree as a gene tree. tree as a species tree us-
species tree, I received this ing “File → Open Species
message: “There are no Tree” or “Ctrl-Shift-O”.
species trees to reconcile
with.” -or- My species tree
is not listed in the drop
down menu in the Reconcil-
iation Options dialog box.
After reconciliation, I found The gene was lost in an an- Use “Display Options →
lost genes in unrecognizable cestral species that was not Display Internal Node
species, such as “n101.” given a label in the original Names” to examine inter-
species tree file. When in- nal species names in the
ternal node names are not species tree. If you pre-
specified in the input file, fer taxonomic names, use a
Notung generates them us- tree editing program or text
ing an arbitrary counting editor to add real species
system (ex: n101). names to internal nodes in
the species tree.
When I tried to open a tree, • The gene tree file is sup- • Open the original tree file
I received this message: posed to be in Newick, in a tree editing program
“An error occurred while NHX or Notung format, or text editor and correct
opening your file. Please but contains a typo or for- any formatting errors.
check the format.” Or “An matting error.
error occurred while open- • Convert the file to Newick
ing your file. Node had mal- • The gene tree file is in a or NHX file format. See
formed information.” format Notung does not Appendix A - File For-
accept, (ex: Nexus). mats on page 126 for more
information about file for-
mats.

178
Appendix F. Troubleshooting

Problem Possible Causes Solutions

Notung reports that you do You have multiple versions • On Windows, bring up the
not have a recent enough of Java installed. properties window for the
version of Java, but you Notung-2.7 jar file. Check
have the latest version in- the “Opens With” field - if
stalled. the wrong version of java
is listed, change it so that
the right version of java is
being used.

• On Linux or Mac, type

java -version - this will
tell you which version of
Java is being used. If
it is incorrect, alter your
path environment variable
to include the proper ver-
sion of Java.

The species tree file I cre- As part of its file construc- These characters are ig-
ated using the NCBI Tax- tion, the NCBI Taxonomy nored by Notung, but you
onomy Browser contained Browser includes some non- can open the tree file in a
non-ASCII characters. ASCII characters. text editor and delete the
non-ASCII characters.
The species tree file I cre- As part of its file construc- These branch lengths are ig-
ated using the NCBI Taxon- tion, the NCBI Taxonomy nored by Notung, but you
omy Browser contained 4’s. Browser includes a branch can open the tree file in a
length of 4 for every edge tree editing program or text
in the species trees it pro- editor and delete the branch
duces. lengths.

179
 Appendix F. Troubleshooting

Problem Possible Causes Solutions

The names of internal nodes • The gene tree file does • If you want the internal
in my gene tree change over not specify internal node node names to be the
time. names and has been same every time the tree
reloaded. When inter- is opened, use a tree edit-
nal node names are not ing program or text editor
specified in the input to add names to internal
file, Notung generates nodes in the gene tree.
them using an arbitrary
counting system (ex: • Notung cannot track in-
n101). ternal nodes that are tem-
porary or not present in
• Node names were given the original file. If you
in the original tree file, need permanent names for
but additional nodes have these nodes, save the file
been added, because ei- and use a tree editing pro-
ther rearrangement or re- gram or text editor to
solve has been performed. specify names for these
Added nodes are assigned nodes.
names that begin with an
‘r’ and are followed by
numbers (ex: r245).

I use the <Tab> key to nav- This is a problem with Use the mouse to select but-
igate to a different button some versions of Java. The tons in the windows
in a pop-up box, but when <Tab> key option to nav-
I hit the <Enter> key, the igate to different buttons
selected button is not en- does not select the “high-
gaged. lighted” button. When the
<Enter> key is pressed, the
originally selected button is
used.
I have added a node to an There are conflicting anno- Manually remove the node
annotation, but the node tations - the node corre- from the other annotations,
does not appear in the cor- sponds to more than one or check the other annota-
rect color. annotation and is currently tions and remove any search
being described by another strings that identify the
annotation. Annotations node of interest. See Chap-
have precedence - those an- ter 10 - Annotations on page
notations added later will 87 for more information.
always take precedence over
earlier annotations.

180
Bibliography

[1] J. Andersson. Horizontal gene transfer between microbial eukaryotes. Methods Mol Biol,
532:473–487, 2009.

[2] K. Chen, D. Durand, and M. Farach-Colton. Notung: A program for dating gene
duplications and optimizing gene family trees. J Comput Biol, 7(3/4):429–447, 2000.

[3] R. R. Copley, I. Letunic, and P. Bork. Genome and protein evolution in eukaryotes.
Curr. Opin. Chem. Biol., 6(1):39–45, Feb 2002.

[4] Charlotte A. Darby, Maureen Stolzer, Patrick J. Ropp, Daniel Barker, and Dannie
Durand. Xenolog classification. Bioinformatics, 33(5):640–649, Mar 2017.

[5] J. Degnan and N. Rosenberg. Gene tree discordance, phylogenetic inference and the
multispecies coalescent. Trends Ecol Evol, 24:332–340, Jun 2009.

[6] J. Doyon, V. Ranwez, V. Daubin, and V. Berry. Models, algorithms and programs for
phylogeny reconciliation. Brief Bioinform, 12:392–400, Sep 2011.

[7] D. Durand, B. Halldorsson, and B. Vernot. A hybrid micro-macroevolutionary approach

to gene tree reconstruction. J Comput Biol, 13(2):320–335, 2006.

[8] I. Ebersberger, P. Galgoczy, S. Taudien, S. Taenzer, M. Platzer, et al. Mapping human

genetic ancestry. Mol Biol Evol, 24:2266–2276, Oct 2007.

[9] S. Edwards. Is a new and general theory of molecular systematics emerging? Evolution,
63:1–19, Jan 2009.

[10] J. Felsenstein. Inferring Phylogenies. Sinauer Associates, Sunderland, MA, 2003.

[11] J. Felsenstein. PHYLIP (Phylogeny Inference Package) version 3.6. Department of

Genome Sciences, University of Washington, Seattle, 2005.

[12] W. M. Fitch. Homology: a personal view on some of the problems. Trends Genet,
16(5):227–231, May 2000.

181
BIBLIOGRAPHY BIBLIOGRAPHY

[13] M. Hallett, J. Lagergren, and A. Tofigh. Simultaneous identification of duplications and

lateral transfers. In RECOMB 2004: Proceedings of the Eigth International Conference
on Research in Computational Biology, ACM Press, pages 347–356, New York, NY,
USA, 2004. ACM.

[14] R. Hudson. Gene genealogies and the coalescent process. In Oxford surveys in evolu-
tionary biology, volume 7, pages 1–44. Oxford University Press, 1990.

[15] D. H. Huson and C. Scornavacca. A survey of combinatorial methods for phylogenetic

networks. Genome Biol Evol, 3:23–35, Nov 2011.

[16] H. Lai, M. Stolzer, and D. Durand. Fast heuristics for resolving weakly supported
branches using duplication, transfers, and losses. In RECOMB International Workshop
on Comparative Genomics, volume 10562, pages 298–320. Springer, Springer, 2017.

[17] W. Maddison. Reconstructing character evolution on polytomous cladograms. Cladistics,

5:365–377, 1989.

[18] W. Maddison. Gene trees in species trees. Syst. Biol., 46(3):523–536, 1997.

[19] L. Nakhleh. Evolutionary phylogenetic networks: models and issues. In L Heath and
N Ramakrishnan, editors, The Problem Solving Handbook for Computational Biology
and Bioinformatics, pages 125–158. Springer, 2010.

[20] L. Nakhleh, D. Ruths, and H. Innan. Gene trees, species trees, and species networks.
In R. Guerra and D. Goldstein, editors, Meta-analysis and Combining Information in
Genetics and Genomics, pages 275–293. CRC Press, 2009.

[21] R.D.M. Page and E.C. Holmes. Molecular Evolution: A phylogenetic approach. Blackwell
Science, Malden, MA, 1998.

[22] P. Pamilo and M. Nei. Relationships between gene trees and species trees. Mol Biol
Evol, 5(5):568–583, 1988.

[23] D. Pollard, V. Iyer, A. Moses, and M. Eisen. Widespread discordance of gene trees
with species tree in Drosophila: evidence for incomplete lineage sorting. PLoS Genet,
2(10):e173, Oct 2006.

[24] M.H. Serres, A.R.W. Kerr, T.J. McCormack, and M. Riley. Evolution by leaps: gene
duplication in bacteria. Biol Direct, 4:46, Nov 2009.

[25] M. Stolzer, H. Lai, M. Xu, D. Sathaye, B. Vernot, and D. Durand. Inferring dupli-
cations, losses, transfers, and incomplete lineage sorting with non-binary species trees.
Bioinformatics, 28(18):i409–i415, 2012.

[26] D. L. Swofford. PAUP*: Phylogenetic Analysis Using Parsimony (*and Other Methods).
Version 4. Sinauer Associates, Sunderland, MA, 2002.

182
BIBLIOGRAPHY BIBLIOGRAPHY

[27] F. Tajima. Evolutionary relationship of dna sequences in finite populations. Genetics,

105(2):437–460, Oct 1983.

[28] N. Takahata and M. Nei. Gene genealogy and variance of interpopulational nucleotide
differences. Genetics, 110(2):325–344, Jun 1985.

[29] A. Tofigh, M. Hallett, and J. Lagergren. Simultaneous identification of duplications and

lateral gene transfers. TCBB, 8:517–535, Mar/Apr 2011.

[30] B. Vernot, M. Stolzer, A. Goldman, and D. Durand. Reconciliation with non-binary

species trees. J Comput Biol, 15(8):981–1006, 2008.

[31] O. Zhaxybayeva and W. Doolittle. Lateral gene transfer. Curr Biol, 21:R242–R246, Apr
2011.

[32] C. M. Zmasek and S. R. Eddy. ATV: display and manipulation of annotated phylogenetic
trees. Bioinformatics, 17(4):383–4, Apr 2001.

183