Chimeric Reads in Log.final.out

[Felix Schlesinger]

Hi,

Are chimeric reads not included in the "Uniquely mapped reads %" reported in Log.final.out? For runs with many fusions the percentages do not add up to 100 for me. Or maybe they should be reported separately?

Thanks

  Felix

[Alexander Dobin]

Hi Felix,

the chimeric reads that do not have a valid non-chimeric alignment are counted as unmapped, even if they are output into the main file.

If chimeric reads also have a valid non-chimeric alignments, they counted as mapped.

I guess I should count separately the two categories.

Cheers

Alex

[Felix Schlesinger]

If every chimeric read is counted as either aligned or unaligned, I am confused why the percentages in my log file do not add up to 100%. Is any category of reads excluded from either Uniquely mapped reads %, % of reads mapped to multiple loci, % of reads mapped to too many loci and the UNMAPPED READS section?

Do you count chimeric alignments the same way in Log.final.out when using the combined BAM output for chimeric alignments?

Felix

[Alexander Dobin]

Hi Felix,

you are right, it's a bug. I have checked the code and it looks like if the --chimOutType WithinBAM is set, the reads that have chimeric alignments are not counted at all.

Will fix it tomorrow.

Cheers

Alex

[Alexander Dobin]

Hi Felix,

the chimeric reads are now counted in the Log.final.out file.

With the --chimOutType WithinBAM  option, chimeric reads are not counted as mapped or unmapped, so their % has to be added to get to 100%

Cheers

Alex

[Felix Schlesinger]

Great, thanks!