Skip to content

ngs_BOWTIE

Stephen Fisher edited this page Feb 28, 2014 · 8 revisions

Module: BOWTIE

Align reads using Bowtie.

Usage:
	ngs.sh bowtie [-i inputDir] [-v mismatches] [-m maxMulti] [-minins minInsertSize] [-maxins maxInsertSize] -p numProc -s species [-se] sampleID
Input:
	sampleID/inputDir/unaligned_1.fq
	sampleID/inputDir/unaligned_2.fq (paired-end reads)
Output:
	sampleID/bowtie/sampleID.sorted.bam
	sampleID/bowtie/sampleID.suppressed.sorted.bam
	sampleID/bowtie/sampleID.stats.txt
Requires:
	bowtie ( http://bowtie-bio.sourceforge.net/index.shtml )
	samtools ( http://samtools.sourceforge.net/ )
Options:
	-i inputDir - directory with unaligned reads (default: trim)
	-v mismatches - maximum mismatches allowed per read length (default: 3)
	-m maxMulti - suppress all alignments if > m alignments (default: 1)
	-minins minInsertSize - minimum insert size for PE alignment (default: 100bp)
	-maxins maxInsertSize - maximum insert size for PE alignment (default: 450bp)
	-p numProc - number of cpu to use
	-s species - species from repository: /lab/repo/resources/bowtie
	-se - single-end reads (default: paired-end)

Run bowtie on the unaligned reads (ie sampleID/inputDir). The arguments used assume Bowtie version 1. Output is placed in the directory sampleID/bowtie. Multimapping reads that exceed the maxMulti count are output to the sampleID.suppressed.sorted.bam file (ie the Bowtie -max flag is used to direct the reads to this file). For paired-end samples, after the paired mapping is complete then all unmapped reads are aligned as if they were single-end with the alignments stored in the sampleID/bowtie/SE_mapping directory. The alignment stats for the single-end and paired-end mappings are reported in the stats file.

Clone this wiki locally