prohibiting splicing with --alignIntronMax 1 not working

[Eddie Imada]

Hello,

I'm using STAR to align a small RNA dataset to Hg38 and I'm trying to prohibit splicing by setting --alignIntronMax (1) smaller than --alignIntronMin as suggested in another topic of this group. However in the Log.final.out report I'm still getting around 2200 total number of splicing, shouldn't it be 0?

Also, I'm having a very low unique mapped reads (around 6%) and a very high multiple mapped (~70%) is this normal in smal RNA datasets? It's the first time I'm analyzing sRNA data.

Thank you in advance.

[Alexander Dobin]

Hi Eddie,

--alignIntronMax 1 will only prohibit unannotated junctions, and if you are using genome with annotations, you will see some splices.

The best way to completely prohibit splices is to generate genome without annotations and use it together with --alignIntronMax 1.

Cheers

Alex

[Eddie Imada]

Thank you for your support Alexander.

How about the --sjdbOverhang option? should I use this option while generating the genome?

Cheers,

Eddie.

[Alexander Dobin]

Hi Eddie,

if you do not use annotations, you do not need --sjdbOverhang option at all.

Cheers

Alex

[Eddie Imada]

Thank you for the clarification.

Best,