MM filters out long deletions?

[noranekonobokkusu]

Hi,
I have a question about the algorithm. I noticed NM tag includes the length of a deletion, even though it was a single event. Do I understand correctly that the default MM value of 4 will exclude any deletions in my data longer than 3 bp? Or am I getting the algorithm wrong?

Thank you!

[RoelKluin]

Dear Ksenia,

I wrote a draft reply, but then got distracted and never sent this, sorry.

The mismatch threshold can be increased to keep a variant like this, and I'd suggest you do if you care. I'd suggest to increase it beyond 4 even, because you will still lose reads that have a 1bp mismatch elsewhere, in particular if this allele has two events within the read length. If you increase the mismatch threshold, there is a chance that more reads are included that are not from the graft, but the location that those reads will align to are likely multi-mapper, repetitive regions. Those regions are generally less interesting anyway.

Best Regards,
Roel