get location for analysed sites

[ZKai0801]

Hi @Beifang and @owehann ,

Thank you for developing such a wonderful tool!

I would like to get all locations of analysed sites and I found someone asked the same question here: https://github.com/ding-lab/msisensor/issues/49

I tried to retrieve all counted locations in your mentioned manner (by default i.e. sites where reads depth >= 20), however, it didn't generate correct output as it does not match the number shown on .prefix file.

So I assume MSIsensor and MSIsensor2 used different criteria for selecting readily analysed sites. Is there a way I could retrieve those sites?

Thank you very much!

[owehann]

Thanks for your attention on MSIsensor2 Actually, MSIsensor and MSIsensor2 used the same criteria for selecting microsatellites. You can easily get all analysed sites with their distributions in prefix_dis file. Samtools can also used to get the detail information of a microsatellite in bam files, and the premise is to pay attention that left and right flank must match with the reference.

[ZKai0801]

Hi @owehann ,

Thank you so much for your reply!

That seems quite odds, because what I did is simply scanning through the prefix_dis file. And whenever I encounter a location contain a number that >= 20, then its cooresponding location will be directed to a output file.

Here is a Python script I used to retrieve sites:

import sys

if __name__ == "__main__":
    with open(sys.argv[1], "r") as fh:
        for line in fh:
            if line.startswith("chr"):
                coordinate_line = line
                continue
            depths = [int(i) for i in line.strip().lstrip("T:").split() if int(i) >= 20]
            if not depths:
                continue
            print(coordinate_line, end="")

To be specific, with the use of this script, sometimes I get less analysed locations than the number shown on prefix file. Is there something wrong with my script? Or would you mind if I sent you an example BAM file?

Thanks for your helping!