Shifted genomic positions of breakends

[Paranchai87]

Hi there

I see that for many breakends, calls for events like INV, etc, have genomica positions that are not corresponding to the reads I see in bam file.
Do you maybe have an explanation for this observation?

Best
Paranchai