Re: mismatch_profile

[Liguo Wang]

Hi Philipp,

It only has two filters: read_number and MAPQ. I check the code (v4.0.0), it looks OK. Maybe the Galaxy used the old version? 

You can install RSeQC and run it locally. Please let me know if you still have problems. 

Thanks

Liguo 

On Wed, Jan 27, 2021 at 7:04 AM Philipp Trnka <phi....@gmail.com> wrote:

​Hey everyone,

I try to use the mismatch_profile tool to compare the mismatch profiles in various samples. I'm running the tool on usegalaxy.eu. 

Whatever I give as input BAM-file, the tool is never looking at all reads, but at a subset?. I did not find any comment on subsampling or a setting to run the tool on all reads.

My BAM files have around 10 million reads that have a mismatch in the mapped sequence. The tool ends up using around 50000 or less reads (depending on the MAPQ setting). 

I mainly want to understand why this is happening - is it by default only considering a subset? Is there another filtering step that I overlooked? e.g. base-quality scores?

Thanks in advance. Best,

Philipp

--
You received this message because you are subscribed to the Google Groups "rseqc-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rseqc-discus...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/rseqc-discuss/3bcbe407-bdcd-405c-bea5-9d0bc051adfbn%40googlegroups.com.