mapping reads using BWA, no MM, no GAP

[nejma]

Dear All,

I'm trying to map reads on genome reference, with no mismatch and no gap, using bwa. Could you advise me how to control these parameters using the sam file?

thank you in advance!

[WhatsOEver]

Although it is not clear to me what the purpose of this could be, I could think of two solutions:

1) If you want to exclude imperfect mappings you could increase the mismatch ("-B") and gap open/extension ("-O" / "-E") costs to avoid introducing mismatches/gaps.

2) If you want to select perfect matches from a samfile after the mapping has been performed, you need to parse the CIGAR strings. If the samfile is in version 1.4 format, you would be interested only in those reads which contain only "=". If it is in version 1.3, the CIGAR string contains "M" which include mismatches and matches. In this case you need to have the optional "MD" field present, which shows the positions of mismatching positions in the read for proper filtering.