search

ExpressionAnalysis / STAR-SEQR

RNA Fusion Detection and Quantification
Other
16 stars 12 forks source link

Runtime far exceeding Wiki-suggested runtime #23

Open dstueckm opened 4 years ago

dstueckm commented 4 years ago

Hi there,

I've been trying to use this software to analyze single-cell RNA-seq data. I was able to successfully run the STAR Index generation, and starseqr.py does not appear to crash but is taking far longer than ~20 minutes as suggested by the Wiki. The apply_get_rna_support function took ~3 hours using 24 CPUs, and the "Getting fusion homology scores" step failed to complete in ~30 hours.

I am wondering if this behavior has been seen before (and if I should commit 24 CPUs to multiple days), or if this likely indicates an error has occurred.

Thanks in advance!

bdvorsky commented 3 years ago

Hi, I have the same problem - in my case this step "Getting fusion homology scores", is taking already 72+ hours. Well, some my samples are done in 40-180 minutes. This one is "bigger" but any other fusion finder (Arriba, deFuse) finished procession of this sample under 20 hours. Also it seems that this step is using only one processor core/thread. Also I have no indication if Star-SEQR is still working or if it is stuck...

Do you have any advice?

Thank you.

rsiddaway commented 3 years ago

I am having the same problem. Most samples run pretty fast, but a few are getting stuck on the "getting fusion homology scores" step and timing out after 200+ hours.

Is there a solution to this problem?

Thanks.