I'm trying to understand some aspects of a pindel run, to try to determine whether I'm getting realistic variant calls or not.
Specifically, when pindel calls a long inversion (e.g.15Kbp), shouldn't I expect to see reads supporting two breakpoints? If I only see one breakpoint for *every* long insertion reported, what should I conclude?
Methods: I have long insert mate pair data, average insert is 8K with std 900. I mapped this to my reference genome with bwa-mem, used sam2pindel to convert to paired end orientation, then ran pindel.
Have I done something wrong?
T.Hattum
Specifically, when pindel calls a long inversion (e.g.15Kbp), shouldn't I expect to see reads supporting two breakpoints? If I only see one breakpoint for *every* long insertion reported, what should I conclude?
Methods: I have long insert mate pair data, average insert is 8K with std 900. I mapped this to my reference genome with bwa-mem, used sam2pindel to convert to paired end orientation, then ran pindel.
Have I done something wrong?
T.Hattum