Extreme large insert size of BWA alignment

Question

[Darren Lin]

The BAM file I am dealing with consist some alignments with extremely large insert size.

For example:

H06HDADXX130110:1:1214:13197:7463   145 20  1361711 60  41S147M62S  =   61590101    60228245
H06HDADXX130110:1:1214:13197:7463   97  20  61590101    60  242M8S  =   1361711 -60228245

What's going on with this kind of alignment?

Thanks!

[Medhat]

This should be changed to a question.
It could be; bad alignment, specially if it is a repetitive genome and this sequence exists in different location on the genome, or may be any kind of structure variations.

try to add to your question the BWA command that you used for alignment.

[Darren Lin]

hm....thanks, but I can not just filter "properly paired" flag. I can probably set a threshold to limit the maximum deletion.

[Pierre Lindenbaum]

yes you can with samtools view -F 2

Answer 1

[Pierre Lindenbaum]

you don't have to deal with that, unless you're looking for large deletions, those reads with flag 145 & 97 are not mapped as "properly paired" ( see https://broadinstitute.github.io/picard/explain-flags.html )