bwa perfect match only

[guil]

Hi everyone,

I'm trying to do an aling of short reads (putative micrornas) to putative precursors using bwa...

but I wanna only perfect matchs between mature and precursor....

in bwa manual I haven't found any parameter to map a read with no mismach.. only gap opening an extesion penalties...


Are there another way to do this???


tnks

[ttnguyen]

Have you tried this?
-k 0

[guil]

thanks ttnguyen,

but i tried the

-k 0 and it doesn't works....

I also tried to extend the mismatch penalty --> 10

and still results in mapping with mismatches...

any ohter idea???

best

[jiangyuehit]

what about filter out alignment that NM value is not 0?

[yifangt]

I am following the same question. There may be two options:
1) Redo the mapping with bowtie or bwa, what is the option to accomplish this? -k 0 does not work!

2) Filter out the SAM/BAM file. What is the option to filter no-mismatch!
Seems not enough information about the option "H:i:0"
I search over this forum and biostar, many discussion with the unique mapped reads, ot unmapped reads. No discussion about the perfectly mapped reads, i.e. mapped reads without any mismatch.
For column 6 of CIGAR string, I used

 grep "\s\d+M\s"

to filter out any matched with deletion and insertion, but the problem after filtering is M can stands for mismatch too, so that the reads still have some that contains mismatch!
Can anybody give some suggestion.
Thanks a lot!