Hey all,
I am mapping Solexa reads against a reference genome using Bowtie and BWA.
With Bowtie I am allowing for 2 mismatches and obtained 1.5M hits.
Using BWA (aln/samse) with default settings resulted in 1.4M hits.
I was expecting BWA to produce more hits than Bowtie since BWA allows indels as well as mismatches. Now, of cause that comes down to the parameters used, and unfortunately I fail to see how BWA treat mismatches. I am also unsure of how BWA reports mismatches - the CIGAR strings only indicate full matches or matches with indels.
So, what is the deal with BWA and mismatches?
Martin
I am mapping Solexa reads against a reference genome using Bowtie and BWA.
With Bowtie I am allowing for 2 mismatches and obtained 1.5M hits.
Using BWA (aln/samse) with default settings resulted in 1.4M hits.
I was expecting BWA to produce more hits than Bowtie since BWA allows indels as well as mismatches. Now, of cause that comes down to the parameters used, and unfortunately I fail to see how BWA treat mismatches. I am also unsure of how BWA reports mismatches - the CIGAR strings only indicate full matches or matches with indels.
So, what is the deal with BWA and mismatches?
Martin
Comment