SelectionMap: a tool for detection and visualisation of natural patterns within codon alignments

I. DescriptionSelectionMap is a free Windows based program designed for the purpose of detection

comparison natural selection patterns within homologous genes of related virus species. The

program detects selection using the FUBAR[1] method implemented in HYPHY[2] and then

produces selection map illustrating the type and the degree of selection at each individual

codon sites. The program also takes as input FUBAR and/or MEME [3] out files containing

selection data in coma separated value (csv) format and corresponding codon alignments

and plot the selection map. Generally, the produced map is a multi-coloured figure showing

sites that are evolving (1) under negative selection for the same amino acid, (2) negative

selection for different amino acids (3) sites under negative selection in only one species, (4)

sites under positive selection, (5) under episodic positive selection.

II. Download and installationThe program can be downloaded from http://web.cbio.uct.ac.za/~brejnev/downloads/ComputationalTools/

1. Extract the SelectionMap-v1.0.zip file into a temporary folder.

2. In the temporary folder double-click on the file “SETUP.EXE and follow the instructions of

the setup program. Please use the default installation directory “C:\ SelectionMap-v1.0”

for installation.

3. Inside the installation folder, run the “HyPhy-CLI2.2.6.exe” file to install HYPHY

III. Running the program1. Starting the programTo plot a selection map you simply have to double-click on the SelectionMap-v1.0.exe file, when the program has launched you can either choose to load codon alignment(s)

and detect selection with FUBAR before plotting the selection map or load FUBAR and/or

MEME csv output file(s) and their corresponding codon alignment(s) and then plot the

map. The Start-up screen (Fig 1.) displays the two option of loading the input files.

Fig 1. Start-up screen

2. Loading the input codon alignment (s)Once the user clicks the “Load codon alignment(s) in FASTA format” the “Loading codon

alignment(s) screen” (Fig 2.) will appear and to load the input the user will either click on the

frame in the middle or will drag-and-drop the codon alignment file(s) to the frame.

Once the alignments are loaded, the user will click on the command button to detect

selection.

NB: If two or three alignments are loaded they must be profile aligned before the can be

used.

Fig 2. Loading codon alignment(s) screen

3. Loading the input data file(s)

There are three types of analyses: (1) FUBAR-MEME that uses selection data from FUBAR

and MEME, (2) FUBAR involving selection data from only FUBAR and (3) MEME involving

selection data from only MEME. The analysis can be performed on one virus species, or

multiple virus species in which case direct comparison of selection pattern are made.

4. Loading the input data file(s)To load the first input files click on “Load data file 1”, then “Load data file 2” to load the

second and “Load data file 3” to load the third file. Alternatively drag and drop the input file(s)

to the program picture fram.

5. FUBAR&MEME analysisInput file for this analysis (Fig 1.) consist of a text file in comma- or tab-separate format

whose first to tenth columns contain the FUBAR output, eleventh to twentieth columns

contain the MEME output and the twenty-first column contains the consensus (most

frequent) amino acid at each codon site.

Fig 1. Input format for FUBAR-MEME analysis

NB. In case two or three input files are to be compared, the alignments must be profile-

aligned and separated before running FUBAR and MEME, to insure that homologous sites

are aligned.

6. FUBAR analysis

Input file for this analysis (Fig 2.) consist of a text file in comma- or tab-separate format

whose first to tenth columns contain the FUBAR output and the eleventh column contains

the consensus (most frequent) amino acid at each codon site.

Fig 2. Input format for FUBAR analysis

NB. In case two or three input files are to be compared, the codon alignments must be

profile-aligned and separated before running FUBAR, to insure that homologous sites are

aligned.

7. MEME analysis

Input file for this analysis (Fig 3.) consist of a text file in comma- or tab-separate format

whose first to tenth columns contain the MEME output.

Fig 3. Input format for MEME analysis

NB. In case two or three input files are to be compared, the codon alignments must be

profile-aligned and separated before running MEME, to insure that homologous sites are

aligned.

IV. Program features

Fig 4. Start-up screen

Fig 4. SelectionMap main interface

1. Menusa. The File menu gives the option to exit the program

b. The Save menu allows saving on the disk the selection map in either EMF or PNG

format.

c. The Copy menu allows copying and pasting the map into another program such us

MS PowerPoint or MS Word. An alternative option is to write-click on the image and

select copy.

2. Command buttonsa. The Plot map command button allows plotting the map once the input files have been

loaded.

b. The Zoom in and Zoom out command buttons allow zooming the map in and out

repsectively.

c. The Set method allows selecting the method to use. The methods supported are:

FUBAR-MEME, FUBAR and MEME.

3. The Input data file buttons (see three labels bellow the “FUBAR-MEME” caption), allow

loading the input files, once they have been loaded their name are displayed on the

labels. An alternative and easy option is to drag and drop the input data files on the

picture frame.

4. The Posterior probability and P-value text boxes allow setting the statistical significance

thresholds for selection inferred by FUBAR and MEME respectively.

5. The Delete file button allows removing the input files.

6. The Colour boxes allow setting the colour for different of types of selection detected.

V. ReferencesMurrell B, Moola S, Mabona A, et al. (2013) FUBAR: a fast, unconstrained bayesian

approximation for inferring selection. Mol Biol Evol 30:1196–205. doi:

10.1093/molbev/mst030

Murrell B, Werthim J, Moola S, et al. (2012) Detecting individual sites subject to episodic

diversifying selection. PloS Genetics. DOI: 10.1371/journal.pgen.1002764

Authors: 1Brejnev Muhire, 2Arvind Varsani and 1Darren Martin 1Institute of Infectious Diseases and Molecular Medicine, Computational Biology Group, University of

Cape Town, South Africa2School of Biological Sciences, University of Canterbury, Private Bag 4800, Christchurch, 8140, New Zealand

Bug reporting: mhrbre001@myuct.ac.za or mubrejnev@gmail.com