output all possible match using bwa

[mingkunli]

I am moving from bowtie to bwa, as bwa considers gap while doing mapping.
But i don't know how to get all the possible hits for a single read on the reference genome(besides the perfect one), like using -a with bowtie. I tried -N, but it does't work. Although X1 in the sam file can give me the number of suboptimal hits, I need the details of the hit like the optimal one.
Anyone can give me some clue, if bwa can't do this, what else can?

ok, I got the result myself, after .sai, using bwa samse -n INT to display maximum n match for each read

[nilshomer]

I am moving from bowtie to bwa, as bwa considers gap while doing mapping.
But i don't know how to get all the possible hits for a single read on the reference genome(besides the perfect one), like using -a with bowtie. I tried -N, but it does't work. Although X1 in the sam file can give me the number of suboptimal hits, I need the details of the hit like the optimal one.
Anyone can give me some clue, if bwa can't do this, what else can?

ok, I got the result myself, after .sai, using bwa samse -n INT to display maximum n match for each read

I think you are looking for the -e option. Set this to the maximum length indel you want detect. My experience shows BWA has difficulty with longer indels (>10bp).

A solution to the indel problem as well as getting "all hits" is to use BFAST, admittedly my own software. I feel like I am doing a lot of self-promotion of late but I guess it is because I am passionate about my solution.

Nils

[mingkunli]

：）
Thanks for reminding me of the -e option. I will try your BFAST if longer than 10bp indels are expected in my reads.