Segmentation fault for different input with a same genome

[taochu...@gmail.com]

Hi Alex,

I  have a wild strawberry genome and some sample's reads (pair-end).  When I aligned my samples to the genome one by one, a part of the alignment were successed, othes got segmentation fault.

I place the log in the attachment, 1-CK-12 is the successed one, and the 1-CK-24 is the fail one.

The file size of 1-CK-12's read1 is 7.9G, 1-CK-24 is 8.8G.

My STAR version is  2.5.3a

Thanks,

May

[Alexander Dobin]

Hi May,

first, please try to reduce --genomeSAindexNbases to 12 or even 10 and see of the seg-fault goes away.

Cheers

Alex

[taochu...@gmail.com]

Hi Alex

I reduce --genomeSAindexNbases to 12, then the succeed samples reduced from 4 to 1(total 18 samples). I'm very very confused.

Need more suggestion. 

Thank you very much.

May

在 2017年6月22日星期四 UTC+8上午5:21:12，Alexander Dobin写道：

[Alexander Dobin]

Hi May,

have you re-generated the genome with  --genomeSAindexNbases 12?

In the 1-CK-12.Log.out file, it still says 

genomeSAindexNbases           14     

Also, please try to run without setting

seedPerWindowNmax                 10000

alignTranscriptsPerReadNmax       100000

alignTranscriptsPerWindowNmax     10000

These parameters need to be tweaked for STARlong, but defaults are OK for STAR.

Cheers

Alex