bagofmotifs

Description

Program searches for similar gene enhancers between query and target sequences using transcription factor binding site motif signatures.

Aligns divergent enhancers in organisms based on their collection of binding motifs and calculates a p-value based on randomizing the input data.

Usage

To run the program from the command line, enter:

./main.R <query file> <target file>

Relative path to query file (lasagna motif file)
Relative path to target file (lasagna motif file)

Command line arguments

The first two arguments must be the query and target motif files.

The following optional flags can be used to set parameters in the program

Flag	Short	Type	Constraints	Description	Default
-pval	-p	double	0 <= pval <= 1	Motif discovery threshold	0.001
-window	-w	integer	window >= 1	Window size for sliding window function	900
-seed	-s	integer	any integer	Using the same seed will ensure the same sample is taken when calculating emperical null distribution	NULL

Input file format

The input files should be in this format with a space seperated table with the following fields:

Chromosome	Start	End	Strand	Score	Pval	Motif

Input files can be generated using any method for identifying motifs and motif database.

We used Lasagna (PMID: 23522376) with a publicly available version of the TRANSFAC (in directory 'lasagna_scan').

Lasagna is first run on query and target FASTA sequences for each TRANSFAC motif using the following command:

python lasagna_scan/scan_UCSC_promoters.py --pvalue 0.01 --compute-pvalue motif_model sequence_file

We included code to wrap Lasagna's scan_UCSC_promoters.py to generate the input motif file to bagofmotifs:

Rscript lasagna_scan/lasagna.R sequence_file

Note: Input to lasagna.R (and scan_UCSC_promoters.py) must have the .fa suffix and a header line in the format '>sequenceID:start-end'.

Output format

By default output is directed to STDOUT and can be redirected into desired file. e.g.:

./main.R <query file> <target file> > output.txt

Motif order test

motifOrder.R contains code to test for conserved motif order using the Smith-Waterman alignment including randomization to test statistical significance.

XuanrZhang/bagofmotifs