How does STAR handle repeat masked genomes?

[Joshua Bradley]

I've seen multiple posts on biostar that say "never align to a repeat masked genome" which makes sense. So my question is:

Does STAR recognize a genome that has soft-masked (lowercase) repeats or is the sequence matching process completely case-insensitive?

[Alexander Dobin]

Hi Joshua,

low or upper case A,C,G,T, AllOther are treated the same (actually converted to 0-5) for both the genome or read sequences.

Cheers

Alex

[César Lizárraga]

Wait, does that mean if you are using lower or uppercase A,C,G,T are all scored the same as well as Other (N)? Or is N treated differently?

Thanks.

Cesar.

[Alexander Dobin]

Hi Cesar,

a=A, c=C, g=G, t=T for both the genome and read sequences.

All other characters (including IUPAC di/tri-nucleotides) are converted into "N", i.e. unknown.

Cheers

Alex

[Zhu Zhuo]

Hi Alex,

Continuing with this question, I'm wondering how STAR scores when the base is aligned to "N" in the genome. Will it be considered as match, or mismatch?

Thank you!

Zhu

[Alexander Dobin]

Hi Zhu,

a base mapped to N is considered neutral, neither match or mismatch. Nothing is added to the alignment score, and it's not counted in the number of matches or mismatches.

Cheers

Alex

[David Soong]

Hi Alex,

There are several regions in hg19 that contain long stretches of A's (or T's). And STAR tends to all long A reads in my sample to these regions, e.g.

chr6:160,521,742-160,521,853

chr12:66,451,367-66,451,458

chr15:77,910,857-77,910,956

Should I mask out those regions with N's or is there a way for STAR to handle this situation?

Thanks.

[Alexander Dobin]

Hi David,

masking the genome with Ns is the way to go, there is no other way to handle it with STAR.

Cheers

Alex