search

JiekaiLab / scTE

MIT License
97 stars 27 forks source link

"outSAMmultNmax" STAR argument for multimapping reads? #55

Open rargelaguet opened 1 year ago

rargelaguet commented 1 year ago

Hello, I've been running scTE in a few data sets and it is working very well, thanks!

I had a question about the alignment. In the instructions you suggest to preserve multi-mapping reads by using the following STAR arguments:

--winAnchorMultimapNmax 100 (to increase sensitivity and detection of multi-mapped reads)
--outFilterMultimapNmax 100 (to control whether multi-mapping reads are included in the output BAM file)
--outSAMmultNmax 1 (to report only the best alignment).

Could you explain why should we constrain outSAMmultNmax to 1, instead of -1 (reporting all alignment for multimappers)?

Thanks a lot, Ricard

jphe commented 1 year ago

Outputting all alignment for multimappers will excessively amplify the signal. This issue has been extensively discussed in the these review paper: doi:10.1038/nrg3117, https://doi.org/10.1038/ s41576-020-0251-y, https://doi.org/10.1038/ s41576-018-0050-x.