Marie lataretu/issue23

[MarieLataretu]

• modularized alignment handling for easy bam file filtering
• bam soft-clipped filter option:
  • by the total number of soft-clipped positions or
  • relative number of soft-clipped positions (relative to the read length)

[hoelzer]

cool @MarieLataretu !

So one main part is now the additional script you wrote to identify soft-clipped reads? I think that's really a feature others are lacking or can not consider if they just run minimap2 and some downstream commands manually. So that's cool.

You added this to prevent false read mappings just bc/ minimap2 soft clips huge parts of a read to get it somehow aligned to the reference?

[MarieLataretu]

So one main part is now the additional script you wrote to identify soft-clipped reads? I think that's really a feature others are lacking or can not consider if they just run minimap2 and some downstream commands manually. So that's cool.

Yes, one can filter by the total number of soft-clipped nts or a relative number (relative to the read length). (And I removed a lot of redundant code.)

You added this to prevent false read mappings just bc/ minimap2 soft clips huge parts of a read to get it somehow aligned to the reference?

We saw that in at least one test data set. So with the soft-clip parameter, one can filter the reads more stringent by not trusting the mapped reads with lots of soft-clipped nts.

---

Ah, what might be missing are the fastq files from the bam output from the soft-clip filter process

[hoelzer]

We saw that in at least one test data set. So with the soft-clip parameter, one can filter the reads more stringent by not trusting the mapped reads with lots of soft-clipped nts.

cool!

[MarieLataretu]

Should work now :)

Feel free to test or merge directly!

[hoelzer]

cool, in the advent of buisiness - I merge :D We anyway need to do then a larger master merge and testing for a release I guess