DMDtoolkit manual
Sep, 2016
The functions of DMDtoolkit include:
1) assisted diagnosis for DMD / BMD using genetic testing;
2) drawing the mutated protein sequence and motifs;
3) drawing pedigree of DMD family;
4) smartly screening the data to maximize the use of existing data;
5) performing statistics for the DMD population and visualizing the results.
Note: Please install R (https://cran.r-project.org/) and Perl (https://www.perl.org/)
before running the following commands.
1 assisted diagnosis for DMD / BMD using genetic testing
For Windows users, open the dos/cmd window, and move to the working directory, e.g.
D:/DMDtoolkit by typing “D:” and “cd DMDtoolkit”. Then use the command: DMDtoolkit.pl
DMDsamples.txt. For Linux/Unix users, open the terminal window and move to the working
directory, then use the same command.
After several seconds, you will get six output files:
“DMDsamples.Dp427m.*”(rdata/pros/stats/diag/diag2/diag3) and open the diag3 file by
Excel/WPS to see the diagnosis.
You can create your own input file according to the format of DMDsamples.txt. Five
columns are required: Subject ID, Gender, Age, Diagnosis and Mutation, separated by a tab.
Missing data is allow.
Note: Please ensure that the following files --- “codon list.txt”, “DMD gene.fa”, “Dp427m
CDs.fa”, “Dp427m protein.fa”, “Dp427m CDs.txt”, “Dp427m Domains.txt” and “ESE
matrices.txt” --- are in the same working directory.
2 drawing the mutated protein sequence and motifs
Use the commands under the R console: setwd(“the/working/directory”) to move to the
working directory; input_file<-“file name” to read the input file; source(“DMDtoolkit.R”) to
perform the graphing.
This command will call the files
“DMDsamples.Dp427m.*”(rdata/pros/stats/diag/diag2/diag3) and create 64 graph files in pdf
format automatically. One of them is as follows:
3 drawing pedigree of DMD family
Use the commands under the R console: setwd(“the/working/directory”) to move to the
working directory; source(“DMDtoolkit.R”) to call the program; plot.ped("file name") to draw the
pedigrees in pdf format automatically. Take “pedigree.txt” as an example. Seven columns are
required: famid --- family ID, id --- individual ID, fid --- father ID, mid --- mother ID, sex, aff ---
affected or not (1 no / 2 yes), and mutation. Missing data is allow.
One of the pedigrees is as follows:
4 smartly screening the data to maximize the use of existing data
Optional: You can use the SmartScreen.R script to perform imputation which is based on
random forests method, and then obtain the weights via linear regression with key indicator
(“Mutation” in default) against independent variables. Use the commands under the R console.
The following figure is an example.
The commands will create two files: “testing data_imputed.txt” and “weights_estimated.txt”.
You can use TextPad or any other text editor to open them. The testing data before and after
imputation are shown as follows:
Before imputation
After imputation
The weights are shown as follows.
Use the command under the Windows dos/cmd window or Linux/Unix terminal window:
SmartScreen.pl “file name” column_No weights. For example: SmartScreen.pl “testing data” 5
1,1,1,1,2,2,1,1 will create a filtered file named “testing data_LVEDD.rdata”. 5 means column 5
(“LVEDD”) is required (5,6 means columns 5 (“LVEDD”) and 6 (“SNIP”) are required);
1,1,1,1,2,2,1,1,1 means the 5th and 6th columns weight 2 and other columns weight 1.
The output file “testing data_LVEDD.rdata” is shown as follows:
You can also use the weights calculated by the SmartScreen.R. The weight of “Mutation” can
be any number (one by default) which will not affect the screening result since each record
contains it.
5 performing statistics for the DMD population and visualizing the results
Use the commands under the R console: setwd(“the/working/directory”) to move to the
working directory; file_name<-"testing data_LVEDD" to set the input file; indicator_threshold<-#
(e.g. 40) to set the threshold of subgroups for t-test; source(“Stats.R”) to perform the statistics.
After the above commands, four output files will be created: “testing data_LVEDD.*”
(sum/cor/reg) and “testing data_LVEDD 40.t-test”, indicating summary, correlation, regression
and t-test results.
Use the command source(“Graph.R”), and then call plot.freq(type, num) to draw the mutation
frequency histogram. type: “del”, “dup” and “all”; num: Arabic numbers.
Call plot.trend(clmn_1, clmn_2) to draw the scatter plot and trend line of column 2 against
column 1. Take plot.trend(3, 6) as an example, 3 means column 3 (“Age”), and 6 means column 6
(“SNIP”).
Call plot.stem(clmn_no) to draw the stem and leaf plot. For example, plot.stem(3) will draw
the stem and leaf plot of column 3 (“Age”), plot.stem(c(3,6)) will draw the plots of columns 3 &
6, and plot.stem(c(3:6)) will draw the plots of columns 3 to 6.
Call plot.clust(clmn_no, cex_no) to draw the cluster dendrogram. clmn_no means column
number; cex_no is a numerical value giving the amount by which plotting text and symbols should
be magnified relative to the default 1. plot.clust(1:6,0.1) will create the following pdf file.
Thank you for using DMDtoolkit. Any questions, please don't hesitate to contact
[email protected].
