Re: Segmentation Fault with Small Genome

[Alexander Dobin]

Hi Joshua,

I run genome generation with your parameters, and then aligned some fly reads to it, and could not replicate the seg-fault.

Couple of quick things to do first:

1. Try the latest patch ftp://ftp2.cshl.edu/gingeraslab/tracks/STARrelease/Alpha/STAR_2.3.1n.tgz

2. Try to reduce  --genomeSAindexNbases to 4,3,2.

If neither of these solve the problem, please send me the minimal set of reads on which you still experience the seg-fault, and also the Log.out file from the failed run.

Cheers

Alex

On Thursday, May 2, 2013 7:50:07 PM UTC-4, Joshua Welch wrote:

Hello,

I'm trying to align reads to the attached sequence, and I'm getting a segmentation fault. I found the thread on this group where someone else experienced a similar problem, and tried re-generating the genome index file using the following command. Since the sequence is about 4700 bases long, I set --genomeSAindexNbases to floor(lg(4700)/2-1) = 5.

star --runMode genomeGenerate --genomeDir /Genomes/FlyBase_Histone_Region/ --genomeFastaFiles FlyBase_Histone_Region.fasta --genomeSAindexNbases 5 --runThreadN 6

When I try to align reads, I still get a segmentation fault. Any idea what might be going wrong?

Thanks,

Joshua

[Joshua Welch]

Updating to the latest version of STAR and generating the genome with --genomeSAindexNbases 2 seems to have fixed the problem. Thanks!