A dotplot-sample generation utilities

This is a small script to generate:

• sample (randomly generated) dot-plot graphs in png format;
• corresponding history of events (inversions, insertions and deletions) -- also as a png-picture.

The result looks like this: (it is also possible to save just one state as a separate png-file -- see the notes below)

As a by-product the script offers several functions that might be useful in the teaching process;

The script relies on ggplot2, grid and gridExtra packages for visualization.

Getting started

No specific installation required:

• download two .R-files
• create samples and stepwise folders in the same directory

or just clone this repo.

Prerequesites

These are R scripts. If necessary packages are not installed, you might need to do that (from within R console):

install.packages(c("ggplot2","grid","gridExtra"));

Usage and fine-tuning

Main purpose -- dotplots generation

• review and refine samples generation parameters if necessary;

• run the dotplot-gen.R script (it will load dotplot-aux.R with function definitions automatically), e.g.

  source("./dotplot-gen.R")

  when in the directory with two R-scripts. (go to the corresponding directory if necessary e.g. with setwd("<directory>") from the R console).

• sample dotplots will appear in samples directory

• stepwise events are shown in combined plots with corresponding numbers in stepwise directory.

Key tuning parameters

The key parameters to be tuned are set in one place in dotplot-gen.R:

Parameter	Default value	Description
seqLength	100	length of the sequences to be generated
numExamples	10	number of samples (files) to generate
numEvents	5	number of events (ins, del or inv) to occur in each example
maxIns	0.5	maximum length of insertions (relative to the current seq length)
maxInv	0.5	--//--//-- inversion --//--//--
maxDel	0.5	--//--//-- deletion --//--//--

Event frequencies are defined as follows:

Parameter	Default value	Description
insertionProb	15%	probability to get an insertion
inversionProb	70%	-//--//- an inversion
deletionProb	15%	-//--//- a deletion

Note. Two technical parameters are also set up in dotplot-aux.R -- these are graphical parameters no_of_Xticks and no_of_Yticks corresponding to the number of ticks along the axes

• Position where ins, inv or del occurs, is drawn randomly, from the uniform distribution over the (current) length of the sequnece.
• Length of the distortion (ins, del or inv) is calculated randomly, from uniform distribution over a share of the current sequence length (see the rndDistLength variable in the inner loop, calculated using corresponding parameters above).

Useful functions

Notice that also it might be interesting to use functions from dotplot-aux.R separately, e.g. in the R console:

seq1 = paste("block",as.character(1:100),sep="_"); ## initialize the ``sequence''
seq1 # show the sequence as text
drawDotplot(seq1,seq1) ## obtain a standard ``y=x''-type graph

seqD = inversion(seq1, 20,20) ## add an inversion
seqD # show what's happened
drawDotplot(seqD,seq1) ## draw a corresponding dotplot

Available functions are (optional parameters are omitted -- see the source for comments):

• inversion(sequence, where, howLong) -- inverts howLong blocks of sequence starting from where;
• deletion(sequence, where, howLong) -- deletes howLong blocks of sequence starting from where;
• insertion(sequence, where, what)-- inserts what (atomic) vector into sequence after where position;

All the three functions return either a new sequence, or a list of two elements (depending on the optional argument; new-sequence only being the default):

• text description of the corresponding event (e.g. to be used in the plot title) -- can be accessed via [[1]];
• the new sequence -- can be accessed via [[2]];

The function drawDotplot(s_reference, s_query) takes two atomic vectors ("sequences") as input and return a ggplot-object (corresponding graph);

In order to visually track inversions with another color a separate logical vector color_change is used (see dotplot-gen.R).

Please refer to comments in the code for further details.

Feedback

If things do not function properly for some reason (e.g. a recent update have broken down something) or if you have a suggestion for improvement -- feel free to reach out to me by email or open an issue / feature request here in the repo.