Closed lt11 closed 1 year ago
No, it shouldn't, I don't see why, insertions and deletions can be of any length. The program classifies records as follows
ACGT A .. deletion
A ACGT .. insertion
A <INS> .. other
A <DEL> .. other
It is possible to filter by length for example like this
bcftools view -i 'abs(ILEN)<=50'
Hi,
Thanks for the input. When insertions and deletions are larger than 50 bp they are usually classified as structural variants. The term "indel" is associated, in the literature, to events ≤ than 50 bp. It may be worth mentioning it in the documentation.
Best,
L
There is no consensus on that actually. Arbitrary thresholds have been used, including much larger ones such as1000 bp.
I agree it would be good to have it documented though.
Dear authors,
While I was filtering a vcf to grab only its indels I noticed that the output also reports "indels" that are longer than 50 bp. Is this an expected behaviour? Shouldn't variants longer than 50 bp be classified as "other"?
Thanks for your help.
Best regards,
Lorenzo