exonerate - how to use multiple cores?

[sandesh]

Hello,
I used this command to run exonerate:

exonerate -m p2g -q uniref90_modified.fasta -t all_LGs/PcapLG_18.fasta --softmasktarget yes --showvulgar no --showalignment no --showtargetgff yes --showquerygff no --percent 80 " > LG18_exonerate_pcap.gff

This only use one core. I have 32 cores.
It's getting too slow.

Thanks

[atcghelix]

I would try running it in server-client mode:

http://i.animalgenome.org/bioinfo/resources/manuals/exonerate/server.html

[sandesh]

I tried exonerate-server too. This one is also slow.

[Blahah404]

exonerate is simply a slow program. You can split the query file up into 32 files and run 32 exonerate commands simultaneously, then concatenate the results.

exonerate comes with a a tool, fastasplit, for this purpose

[sandesh]

exonerate is simply a slow program. You can split the query file up into 32 files and run 32 exonerate commands simultaneously, then concatenate the results.

exonerate comes with a a tool, fastasplit, for this purpose

Thanks
I tried that too. I made 100 parts. Running single file took more then 3 days and stopped, which means more then 300 days to go.

[colindaven]

GMAP might be suitable for your purposes (transcript mapping to GFF3), and is a lot quicker with good parallelization.