question about hisat2

[zhanghao]

I want to use hisat2. According to the manual, "Use extract_splice_sites.py (in the HISAT2 package) to extract splice sites from a GTF file", I do it as:
bac@bac-X10DAi:~/hisat2-2.0.0-beta$ python extract_splice_sites.py mus.gtf > s.txt

the mus.gtf is in the hisat2-2.0.0-beta folder, but the output s.txt is empty!!!

I do not know where is wrong???

[GenoMax]

Are there any other errors? Does mus.gtf have content (i.e. it is not empty)?

[zhanghao]

Are there any other errors? Does mus.gtf have content (i.e. it is not empty)?

You are right, it seems the mus.gtf lack something. I just download this gtf file from ensembl at random. I try other gtf files, they are all ok!

But I found an another problem. When I run hisat2-build to index the genome, if I add --ss and --exons parameters, it can not finish with an error "illlegal instruction (core dump)". Is this due to insufficient memory? If I don't add these parameters, it is ok.

Also, I want to know whether adding "--ss" and "--exon" will give us a better result? Is these parameter like -G xx.gtf in tophat? and let the software map reads to transcripts first?

[texy34]

You have to put the --ss and --exon parameters, and --large-index if needed or desired, after the path to your genome sequence even though the online manual and command line help has them before it. Like this:

hisat2-build -p 16 Triticum_aestivum.IWGSC1.0+popseq.29.dna.genome.fa --large-index --ss wheat_splice_sites.txt --exon wheat_exons.txt Triticum_aestivum