Large difference between aligned reads and RSEM counted reads

[Chase Mateusiak]

Hi all,

I am trying to understand why there is such a large difference between the input library size and the final RSEM count. Specifically, I'm hoping for documentation that would help me understand better why this happens.

I am attaching the output of `stat`, as well as a multiqc report on the off chance anyone wants to look. I'm looking specifically at sample 10367282_visit_1 (you can filter the multiqc report with the tool bar on the right), though this pattern is consistent with all samples.

Best,

Chase Mateusiak

Bioinformatics Scientist, Washington University in St Louis

[Chase Mateusiak]

Edit: couldn't post the multiqc report (kept getting message for "too long"). But, there are a lot of intronic reads, as well as secondary reads, in this library. I am sure that explains it, but I can't make the numbers come out exactly. I am being asked to explain a discrepancy between the RSEM counts and other methods.