Structured Association Mapping using STRUCTURE and

TASSEL

K K Vinod
Indian Agricultural Research Institute

With advances in genotyping technology, including rapid increases in the number of genetic
markers available for QTL studies, association analysis is now a viable approach for the
dissection of complex genetic traits (Churchill et al. 2004). Association mapping involves
assessment of population structure and using this population information and kinship
information among individuals to assess marker – trait association. Two common software
packages widely used today for association mapping are STRUCTURE (Pritchard et al.
2010) and TASSEL (Buckler et al. 2009). STRUCTURE implements a model-based
clustering method for inferring population structure using genotype data consisting of
unlinked markers. This program can demonstrate the presence of population structure,
identify distinct genetic populations, assign individuals to populations, and identify migrants
and admixed individuals. Trait Analysis by Association, Evolution and Linkage, or TASSEL,
makes use of the most advanced statistical methods to maximize statistical power for finding
QTL. Both a structured association approach (Pritchard et al. 2000; Thornsberry et al. 2001)
and a unified mixed model method have been implemented to minimize the risk of false
positives by integrating population structure and family relatedness within populations (Yu et
al. 2006).
I. Determining population structure using Structure 3.2.2
A. Preparation of marker genotype data
Prepare a matrix of marker genotype data in Excel as given below, for microsatellite data:
 A B C D E F G H I J K
1  SSR1 SSR2 SSR3 SSR4 SSR5 SSR6 SSR7 SSR8 SSR9 SSR10
2 GENO1 110 330 190 140 220 140 240 160 200 180
3 GENO1 110 330 190 140 220 140 240 160 200 180
4 GENO2 110 330 190 140 230 140 240 160 190 180
5 GENO2 110 330 190 140 230 140 240 160 190 180
6 GENO3 110 320 190 140 220 140 240 160 200 180
7 GENO3 110 320 190 140 220 140 240 160 200 180
8 GENO4 110 320 999 140 220 140 240 160 200 180
9 GENO4 110 320 999 140 220 140 240 160 200 180
10 GENO5 110 330 180 140 220 140 240 160 200 180
11 GENO5 110 330 180 140 220 140 240 160 200 180

SSR is the code for markers; GENO is for genotype; -999: missing data value
x Save the data file in Text (tab delimited) type with a suitable filename <genodata.txt>.
B. Download and install STRUCTURE. Latest version of STRUCTURE is available for
download at http://pritch.bsd.uchicago.edu/structure.html.

Advanced faculty training on "Impact of genomics in crop improvement: Perceived and achieved", Jan
20 - Feb 9, 2011, Centre for Advanced Faculty Training in Genetics and Plant Breeding, Tamil Nadu
1
Agricultural University, Coimbatore.
Structured Association Mapping using STRUCTURE and TASSEL

x Run the Structure 3.2.2 software, by double clicking the icon at desktop.

DataTreePane ResultPane

AnalysisPane

(1) Building a project

x Click on File > New Project

x Fill in these boxes: Name of project, Select directory and Choose data file
x Select the file saved in step A and Click Next

Advanced faculty training on "Impact of genomics in crop improvement: Perceived and achieved", Jan 20 -
2 Feb 9, 2011, Centre for Advanced Faculty Training in Genetics and Plant Breeding, Tamil Nadu
Agricultural University, Coimbatore.
 K K Vinod

x Fill in these boxes: number of individuals, ploidy of data ('2' for diploid), number of
loci, and missing data value ('-999'). Click [Next]

x Since our data contains marker and genotype labels, check 'Row and marker names'.
Click [Next]

x Since the data file contains genotypes labels, check Individual ID for each
Individual. Click [Finish]
(2) When project is done, a parameter set needs to be configured. For this, in the
STRUCTURE Main window, click on Parameter Set > New

Advanced faculty training on "Impact of genomics in crop improvement: Perceived and achieved", Jan
20 - Feb 9, 2011, Centre for Advanced Faculty Training in Genetics and Plant Breeding, Tamil Nadu
3
Agricultural University, Coimbatore.
Structured Association Mapping using STRUCTURE and TASSEL

x Fill in these boxes: Length of Burnin Period: (100000), and Number of Markov chain
Monte Carlo (MCMC) Reps (simulations) after Burnin: (100000).
This number should be high, preferably more than 100000 to get reliable
convergence.
x Click [OK] button
x In the Ancestry Model tab, select Use Admixture Model (This is the default). Click
[OK] button.

x In the Allele Frequency Model tab, select Alleles Frequencies Correlated

x Click [OK] button.

x Name the newly created parameter set in the input dialogue (e.g. test1)
x Click [OK].
(3) Running Simulations
[If optimum population structure (K) is already known by some other means, then
skip this step and go to step (5)]

Advanced faculty training on "Impact of genomics in crop improvement: Perceived and achieved", Jan 20 -
4 Feb 9, 2011, Centre for Advanced Faculty Training in Genetics and Plant Breeding, Tamil Nadu
Agricultural University, Coimbatore.
 K K Vinod

x In the STRUCTURE main window, click on Project > Start a job

x In the Scheduler dialogue, select the Parameter set you want to run (e.g. test1);
x Set K between a range say 1 to 10;
x Set the number of replications (Iterations) to run (e.g. 5)
x Click [Start]

x The program will run for a very long time (>72 hours) depending on the speed of the
computer, size of the data, and number of iterations and the replications defined in the
parameter set.

x STRUCTURE displays Job is Completed! dialogue after successful analysis.

Advanced faculty training on "Impact of genomics in crop improvement: Perceived and achieved", Jan
20 - Feb 9, 2011, Centre for Advanced Faculty Training in Genetics and Plant Breeding, Tamil Nadu
5
Agricultural University, Coimbatore.
Structured Association Mapping using STRUCTURE and TASSEL

(4) Determining the optimum population structure

x To determine optimum value for K, Click on the Simulation Summary in the Tree
Pane of STRUCTURE window

x In the Result pane, click File on the left hand top corner, to save the simulation
summary in a text file.
x Copy the values of K and Ln P(D) into a convenient data editor (Excel is the best
option) and calculate average of Ln P(D) against each K, across replications. The K at
which Ln P(D) plateaus is to be taken as optimum K.

K=

(5) Once optimum population structure (K) is known, estimate Inferred ancestry (Q matrix)
of individuals,
x In the STRUCTURE main window, Select Parameter Set >Run

x Enter the value of K in the box and click [OK].

Advanced faculty training on "Impact of genomics in crop improvement: Perceived and achieved", Jan 20 -
6 Feb 9, 2011, Centre for Advanced Faculty Training in Genetics and Plant Breeding, Tamil Nadu
Agricultural University, Coimbatore.
 K K Vinod

x When the job is complete, go to the result pane and select the latest run from the
results folder
x From the results on the right pane, select inferred ancestry of individuals
x Copy and paste it in notepad.
Alternately,
x Go to the directory in where you save the project, open the folder with project name,
and open the result folder. There will be several files with a "f" suffix.
x Open the file with later run number in Notepad. In this output file, copy the values of
"Inferred ancestry of individuals"
x Inferred ancestry of individuals (Q matrix) is used as covariate in TASSEL.
x A typical Q matrix will look like as follows:
1 GENO1 0: 0.212 0.020 0.131 0.043 0.593
2 GENO2 0: 0.173 0.201 0.538 0.062 0.027
3 GENO3 0: 0.200 0.270 0.131 0.189 0.211
4 GENO4 0: 0.092 0.506 0.155 0.142 0.105
5 GENO5 0: 0.124 0.329 0.046 0.427 0.074
6 GENO6 0: 0.339 0.053 0.096 0.450 0.062
7 GENO7 0: 0.343 0.039 0.246 0.120 0.251
8 GENO8 0: 0.376 0.208 0.201 0.059 0.155
9 GENO9 0: 0.172 0.044 0.590 0.137 0.058
10 GENO10 0: 0.172 0.163 0.131 0.445 0.089
11 GENO11 0: 0.093 0.470 0.101 0.165 0.171
12 GENO12 0: 0.313 0.156 0.237 0.108 0.187
13 GENO13 0: 0.184 0.371 0.299 0.030 0.117
14 GENO14 0: 0.078 0.159 0.036 0.675 0.052
15 GENO15 0: 0.705 0.076 0.065 0.078 0.077
x Save the Q matrix. This need to be formatted to be read in TASSEL
II. Association Analysis using TASSEL
Association mapping can produce spurious association between marker and phenotype;
therefore, the population structure is an important component in estimating marker – trait
associations. This is done by incorporating the Q matrix of inferred ancestry coefficients of
the individuals across the sub-populations as covariate in the association mapping analysis.
To refine the results, the kinship coefficients are also used in association analysis. Kinship
matrix (K matrix) can be estimated using software such as SPAGeDi (Hardy and Vekemans,
2002) or can be estimated within TASSEL itself.
Unlike that of STRUCTURE which is a complete program by itself, TASSEL stand-alone
version runs only under Java runtime environment (JRE) version 1.5 and above. JRE is freely
downloadable software from Sun Microsystems, http://java.sun.com/. Alternatively, online
versions of TASSEL are also available.
Note: Latest version of TASSEL 3.0 does not support microsatellite data anymore. So SSR
data analysis can be done only using TASSEL version 2.1.
Both TASSEL 2.1 and 3.0 are available for free download at the following website:

Advanced faculty training on "Impact of genomics in crop improvement: Perceived and achieved", Jan
20 - Feb 9, 2011, Centre for Advanced Faculty Training in Genetics and Plant Breeding, Tamil Nadu
7
Agricultural University, Coimbatore.
Structured Association Mapping using STRUCTURE and TASSEL

http://www.maizegenetics.net/index.php?option=com_content&task=view&id=89&Itemid=1
19.
x Once software platforms are ready, double clicking on the file sTASSEL.jar will run
TASSEL 2.1.
A. Preparation of data
TASSEL requires three types of data primarily for the analysis. (i) Marker segregation data
(ii) Phenotype data and (iii) Ancestry coefficient data (Q matrix)
x Prepare these data in Excel, and save as Text Tab delimited (*.txt) files.
(a) Genotype data
Genotype data using microsatellites uses the following format:
 A B C D E F G H I J
1 40 96:2        
2  SSR1 SSR2 SSR3 SSR4 SSR5 SSR6 SSR7 SSR8 SSR9
3 GENO1 110:110 330:330 190:190 140:140 220:220 140:140 240:240 160:160 200:200
4 GENO2 110:110 330:330 190:190 140:140 230:230 140:140 240:240 160:160 190:190
5 GENO3 110:110 320:320 190:190 140:140 220:220 140:140 240:240 160:160 200:200
6 GENO4 110:110 320:320 190:190 140:140 220:220 140:140 240:240 160:160 200:200
7 GENO5 110:110 330:330 180:180 140:140 220:220 140:140 240:240 160:160 200:200
8 GENO6 110:110 ?:? 180:180 140:140 230:230 140:140 240:240 160:160 190:190
9 GENO7 110:110 330:330 190:190 140:140 220:220 140:140 240:240 160:160 200:200
10 GENO8 110:110 320:320 180:180 140:140 220:220 140:140 240:240 160:160 200:200
11 GENO9 110:110 330:330 190:190 140:140 220:220 140:140 250:250 160:160 200:200
12 GENO10 120:120 320:320 180:180 140:140 220:220 140:140 240:240 160:160 200:200

Note: The number in the first row tell TASSEL, the number of individuals, followed by
number of markers, and (:2) indicate diploid nature of the individuals. ? is commonly used
for missing data. Don't put individual with missing data in the first row. Instead, move it into
another row.
Genotype data using SNP uses the following format:
 A B C D E F G H I J
1 40 96:2        
2  SNP1 SNP2 SNP3 SNP4 SNP5 SNP6 SNP7 SNP8 SNP9
3 GENO1 C:C G:G T:T G:G C:C G:G G:G A:A T:T
4 GENO2 C:C G:G T:T G:G A:A G:G G:G A:A T:T
5 GENO3 C:C G:G T:T G:G C:C G:G G:G A:A T:T
6 GENO4 C:C G:G T:T G:G C:C G:G G:G A:A T:T
7 GENO5 C:C G:G A:A G:G C:C ?:? G:G A:A A:T
8 GENO6 C:C G:G A:A G:G A:A G:G G:G A:A T:T
9 GENO7 C:C G:G T:T G:G C:C G:G G:G A:A T:T
10 GENO8 C:C G:G A:A G:G C:C G:G G:G A:A T:T
11 GENO9 C:C G:G T:T G:G C:C G:G T:T A:A T:T
12 GENO10 T:T G:G A:A G:G C:C G:G G:G A:A T:T

Advanced faculty training on "Impact of genomics in crop improvement: Perceived and achieved", Jan 20 -
8 Feb 9, 2011, Centre for Advanced Faculty Training in Genetics and Plant Breeding, Tamil Nadu
Agricultural University, Coimbatore.
 K K Vinod

Note: The number in the first row tell TASSEL, the number of individuals, followed by
number of markers, and (:2) indicate diploid nature of the individuals. ? is commonly used
for missing data. Don't put individual with missing data in the first row. Instead, move it into
another row.
x Save the data matrix in Text (Tab delimited) type with a suitable filename,
<Markername.txt>.
(b) Phenotype data
Phenotype data uses following format:
 A B C D E F G
1 40 6 1    
2  PHE1 PHE2 PHE3 PHE4 PHE5 PHE6
3 GENO1 12.72 21.64 121.23 88.30 2.40 12.72
4 GENO2 11.32 25.16 129.20 95.30 2.46 11.32
5 GENO3 12.38 25.32 139.10 92.35 2.72 12.38
6 GENO4 13.00 25.19 123.60 104.80 2.26 13.00
7 GENO5 12.67 24.19 129.70 97.50 2.95 12.67
8 GENO6 10.80 24.24 118.10 86.10 2.57 10.80
9 GENO7 9.62 27.92 129.60 94.85 1.94 9.62
10 GENO8 9.35 25.30 114.20 96.70 2.28 9.35
11 GENO9 9.68 25.41 83.70 99.70 1.65 9.68
12 GENO10 9.16 26.44 94.50 91.00 2.10 9.16
Note: The number in the first row tell TASSEL, the number of individuals, followed by
number of traits, and 1 indicate number of header rows. -999 is commonly used for missing
data.
x Save the phenotype data matrix in Text (Tab delimited) type with a suitable filename,
<traitname.txt>.
(c) Population structure data
Population structure data (Q matrix) uses following format:
 A B C D E F
1 40 5 1   
2  Q1 Q2 Q3 Q4 Q5
3 GENO1 0.000 0.003 0.037 0.003 0.956
4 GENO2 0.000 0.003 0.006 0.016 0.975
5 GENO3 0.000 0.001 0.001 0.001 0.996
6 GENO4 0.000 0.004 0.005 0.002 0.989
7 GENO5 0.000 0.001 0.001 0.001 0.996
8 GENO6 0.000 0.001 0.002 0.001 0.996
9 GENO7 0.001 0.003 0.629 0.004 0.362
10 GENO8 0.000 0.002 0.001 0.001 0.995
11 GENO9 0.000 0.021 0.004 0.125 0.850
12 GENO10 0.001 0.002 0.002 0.002 0.993
Note: The number in the first row tell TASSEL, the number of individuals, followed by
number of sub-populations (K=5), and 1 indicate number of header rows.

Advanced faculty training on "Impact of genomics in crop improvement: Perceived and achieved", Jan
20 - Feb 9, 2011, Centre for Advanced Faculty Training in Genetics and Plant Breeding, Tamil Nadu
9
Agricultural University, Coimbatore.
Structured Association Mapping using STRUCTURE and TASSEL

x Save the Q matrix in Text (Tab delimited) type with a suitable filename, <Q_matrix
name.txt>.
(d) Kinship data (K matrix)
Kinship data is an optional requirement for Association mapping. Structured association
analysis is done using a general linear model (GLM) algorithm, which does not require K
matrix. K matrix is however, essential for mixed linear model (MLM) analysis.
If the kinship output from SPAGeDi is used, it should be formatted to read in TASSEL as
given below. For this, (i) add a value of "2" for relative kinship between same individuals and
(ii) change the all negative values of relative kinship into "0".
 A B C D E F
1 40     
2 GENO1 2.000 0.595 1.688 1.688 0.506
3 GENO2 0.595 2.000 0.572 0.550 1.286
4 GENO3 1.688 0.572 2.000 1.465 0.483
5 GENO4 1.688 0.550 1.465 2.000 0.416
6 GENO5 0.506 1.286 0.483 0.416 2.000

Note: The number in the first row tell TASSEL, the number of individuals. No missing data
are permitted in K matrix.
x Save the K matrix in Text (Tab delimited) type with a suitable filename <kinship.txt>
B. Running Structured Association Mapping
(i) Loading data
x Double click on the file sTASSEL.jar in the TASSEL 2-1 directory. Following
window opens.

OptionsPanel

DatatreePanel

MainPanel

ReportPanel

x Click on the [Data] button from Options panel.

Advanced faculty training on "Impact of genomics in crop improvement: Perceived and achieved", Jan 20 -
10 Feb 9, 2011, Centre for Advanced Faculty Training in Genetics and Plant Breeding, Tamil Nadu
Agricultural University, Coimbatore.
 K K Vinod

x From the buttons below click on [File] button

x Select the data type to load, if not sure about data, it is better to choose “I will make
my best guess and try”, this will allow TASSEL to select the data type by itself.
x Click [OK] and select the file containing marker name <markername.txt>.
x Repeat the step, and load phenotype data <traitname.txt> and Q-matrix
<Q_matrixname.txt> one after the other.
x Once the data are loaded, they appear in the Data tree panel.
(ii) Geneotype data processing
When diploid microsatellite data is used, convert the raw format to genetype state, to do this,
x Click on the button [A:a Genotype] to start Genetype Converter

x Select Create alignment based on genotypic state (eg. A:a > Aa) and click [OK]
x This will add another dataset named “GenoStates” in the Data tree panel
(iii) Joining marker, phenotype and population structure data
x In the Data tree Panel, select data "GenoStates", "<Traitname>" and
"<Q_Matrixname>" by clicking on them, while <Ctrl> key is pressed

Advanced faculty training on "Impact of genomics in crop improvement: Perceived and achieved", Jan
20 - Feb 9, 2011, Centre for Advanced Faculty Training in Genetics and Plant Breeding, Tamil Nadu
11
Agricultural University, Coimbatore.
Structured Association Mapping using STRUCTURE and TASSEL

x Click on the button [U Join] in the Options Panel

x A new Data set named “GenoStates+<Traitname>+<Q_matrixname>” appear on the
data tree panel
(iv) Loading relative kinship data (for MLM analysis only)
x If kinship information is available, load it by clicking [File] button on the Options
Panel.
x Select either "Load square numerical matrix (eg. kinship) (phylip)" or “I will make
my best guess and try” and click [OK]
x Select Kinship file and Open
x The kinship data appear under Matrix in the Data tree panel
x Alternately Kinship can be calculated within TASSEL, by selecting the “GenoStates”
and clicking on [Analysis] and then [Kinship].
Note: This is a simple kinship matrix generated from the distance matrix. In order to
use more robust Kinship estimates it is recommended to use SPAGeDi or SAS.
(v) Structured association analysis using least squares GLM
x Select “GenoStates+<Traitname>+<Q_matrixname>” from the Data tree panel, by
clicking on it while holding the <Ctrl> key pressed
x Click on the [Analysis] button and then click on [GLM]

x Select all phenotype as data, and Sub-population data as covariate, Exclude the last
sub-population
x Check “Analyse Each Data Column Separately”
x Click [OK]

Advanced faculty training on "Impact of genomics in crop improvement: Perceived and achieved", Jan 20 -
12 Feb 9, 2011, Centre for Advanced Faculty Training in Genetics and Plant Breeding, Tamil Nadu
Agricultural University, Coimbatore.
 K K Vinod

x Click on [Define Output]

x In Define Ftests, we can set number of permutations, say 1000. Click [Run]
x A new data set appear under Result>Association in the Data tree Panel named
“GLM_ GenoStates+<Traitname>+<Q_matrixname>”
(vi) Viewing and saving results
x Click on the button [Results] from Options Panel
x Select the result data from Data tree Panel, by holding the <Ctrl> key pressed and
clicking on “GLM_ GenoStates+<Traitname>+<Q_matrixname>”
x Click [Table] button from the Options Panel

Advanced faculty training on "Impact of genomics in crop improvement: Perceived and achieved", Jan
20 - Feb 9, 2011, Centre for Advanced Faculty Training in Genetics and Plant Breeding, Tamil Nadu
13
Agricultural University, Coimbatore.
Structured Association Mapping using STRUCTURE and TASSEL

x By clicking on the [Print] results can now be printed, or exported to Tab delimited
text file or Comma separated values (CSV) text file by clicking on buttons [Export
(CSV)] and [Export (Tab)] respectively.
(vii) Understanding the result file
Chr_ df_ F_ p_ #perm_ pperm_ padj_ df_ df_ MS_ Rsq_ Rsq_
Trait Locus Site Chr pos Marker Marker Marker Marker Marker Marker Model Error Error Model Marker

SPY RM1 0 0 0 1 0.0187 0.8921 3000 0.8957 1 5 33 7.6772 0.1557 4.78E04
SPY RM101 0 0 0 1 0.0612 0.8061 3000 0.8137 1 5 33 7.6674 0.1568 0.0016
SPY RM107 0 0 0 1 8.1549 0.0074 3000 0.0117 0.0643 5 33 6.1595 0.3226 0.1674
SPY RM11 0 0 0 1 0.2542 0.6175 3000 0.6148 1 5 33 7.6229 0.1617 0.0065
SPY RM127 0 0 0 2 0.5874 0.5616 3000 0.5648 1 6 32 7.6411 0.1851 0.0299
SPY RM13 0 0 0 1 2.0689 0.1597 3000 0.1799 1 5 33 7.2284 0.2051 0.0498
SPY RM144 0 0 0 1 0.0401 0.8425 3000 0.8414 1 5 33 7.6723 0.1563 0.001
SPY RM152 0 0 0 1 1.2263 0.2761 3000 0.2832 1 5 33 7.4064 0.1855 0.0303
SPY RM153 0 0 0 2 0.2129 0.8094 3000 0.8177 1 6 32 7.8176 0.1663 0.0111
SPY RM154 0 0 0 2 1.3971 0.262 3000 0.2602 1 6 32 7.2855 0.2231 0.0678
SPY RM16 0 0 0 2 0.2159 0.807 3000 0.8117 1 6 32 7.8162 0.1665 0.0112

The result file, in addition to displaying the F-statistics and p-values for the requested F-tests,
also contains information about degrees of freedom, the error mean square for the model, R-
square of the model, and Rsquare for the marker. The model R-square is the portion of total
variation explained by the full model. The marker R-square is the portion of total variation
explained by the marker but not by the other terms in the model. When permutations are
requested, #perm_Marker is the number of permutations run, pperm_ Marker is a test of
individual markers, and p-adj_Marker is the marker p-value adjusted for multiple tests. The
p-adj_Marker value is a permutation test derived using a step-down MinP procedure (Ge et
al. 2003) and controls the family-wise error rate (FWER). For example, if only markers with
p-adj values of .05 or less are accepted as significant, then the probability of rejecting a single
true null hypothesis across the entire set of hypotheses is held to .05 or less. This test takes
dependence between hypotheses into account and does not assume that hypotheses are
independent as do other multiple test correction procedures.
Note:
Both STRUCTURE and TASSEL comes with well written tutorials. This document is no
substitution for those. For any clarification and in depth information please read these
tutorials carefully. Besides, there are online discussion forums available for these software

Advanced faculty training on "Impact of genomics in crop improvement: Perceived and achieved", Jan 20 -
14 Feb 9, 2011, Centre for Advanced Faculty Training in Genetics and Plant Breeding, Tamil Nadu
Agricultural University, Coimbatore.
 K K Vinod

packages, in which users post their doubts and suggestions. These discussions are watched by
the developers of these software and they incorporate modifications/ fix bugs as and when
required.
To join these forums visit following sites,
STRUCTURE: https://groups.google.com/d/forum/structure-software
TASSEL : http://groups.google.com/d/forum/tassel
Major References :
Buckler E, Casstevens T, Bradbury P, Zhang Z (2009) Trait Analysis by aSSociation, Evolution and
Linkage (TASSEL): User Manual. Cornell University
http://www.maizegenetics.net/tassel/docs/TASSEL_help.pdf
Pritchard JK, Wena X, Falush D (2010) Documentation for structure software: Version 2.3.
Department of Human Genetics, University of Chicago.
http://pritch.bsd.uchicago.edu/structure_software/release_versions/v2.3.3/structure_doc.pdf
Other references:
Bradbury PJ, Zhang Z , Kroon DE , Casstevens TM, Ramdoss Y, Buckler ES (2007) TASSEL:
software for association mapping of complex traits in diverse samples. Bioinformatics 23:
2633-2635
Churchill G, Airey DC, Allayee H, Angel JM, Attie AD et al. (2004) The Collaborative Cross, a
community resource for the genetic analysis of complex traits. Nature Genet 36: 1133-1137
Hardy OJ, Vekemans X (2002) SPAGeDi: a versatile computer program to analyse spatial genetic
structure at the individual or population levels. Mol Ecol Notes 2: 618-620
Pritchard JK, Stephens M, Rosenberg NA, Donnelly P (2000) Association mapping in structured
populations. Am J Human Genet 67: 170-181.
Pritchard, J. K., Stephens, M., and Donnelly, P. (2000) Inference of population structure using
multilocus genotype data. Genetics, 155:945–959
Thornsberry JM, Goodman MM, Doebley J, Kresovich S, Nielsen D et al. (2001) Dwarf8
polymorphisms associate with variation in flowering time. Nature Genet 28: 286-289.
Yu JM, Pressoir G, Briggs WH, Bi IV, Yamasaki M et al. (2006) A unified mixed-model method for
association mapping that accounts for multiple levels of relatedness. Nature Genet 38:203-
208

Advanced faculty training on "Impact of genomics in crop improvement: Perceived and achieved", Jan
20 - Feb 9, 2011, Centre for Advanced Faculty Training in Genetics and Plant Breeding, Tamil Nadu
15
Agricultural University, Coimbatore.