Samtools

[mathew]

I ran bwasw and i noticed several reads which are aligning at more than one location nonspecifically. Which flag has to be used in samtools to get unique alignment if I am working with while mouse genome and single read alignments. Thanks.

[nilshomer]

I ran bwasw and i noticed several reads which are aligning at more than one location nonspecifically. Which flag has to be used in samtools to get unique alignment if I am working with while mouse genome and single read alignments. Thanks.

Could you post the given records? I would assume you want the "primary" alignment flag.

Also, could you please be more descriptive in your subject name. I would guess that is why you aren't getting any responses besides me!

[swbarnes2]

I'm pretty sure that bwa won't set the binary flag to indicate that. You can filter your .bam with samtools view for a minimum mapping quality, reads that map equally well to multiple places should have a mapping quality of 0.

[nilshomer]

I'm pretty sure that bwa won't set the binary flag to indicate that. You can filter your .bam with samtools view for a minimum mapping quality, reads that map equally well to multiple places should have a mapping quality of 0.

It does, download the latest code, and try "grep "0x100" *.c":

bntseq.c: m_seqs = m_holes = 8; m_pac = 0x10000;
bwt_gen.c:#define MIN_AVAILABLE_WORD 0x10000
bwtsw2_aux.c: p->flag |= 0x10000;
bwtsw2_aux.c: ksprintf(&str, "%s\t%d", ks->name, q->flag | (opt->multi_2nd && i? 0x100 : 0));

The last line is key.