Question

Exonerate Affine:Overlap Alternative

0

Entering edit mode

10.7 years ago

jomaco ▴ 200

Hi,

In order to see if any gaps are closed, I am trying to map cDNA query sequences to an assembly. I have been using exonerate with the affine:overlap model - I have found this method to be very slow.

Are there any alternatives? Ideally I'd like to have a standard looking blast style output showing query/target sequnce alignment, percent identity etc., rather than using an assembly program such as AMOS minimus.

Thanks,

Jon.

overlap • 2.7k views

ADD COMMENT • link updated 10.7 years ago by Michael 55k • written 10.7 years ago by jomaco ▴ 200

0

Entering edit mode

I am trying to align ESTs using method est2genome, but it appears to me that exonerate just is very slow. With standard settings it took over 80 minutes to align a single EST against a 600Mb genome, but I have 100.000 to align.

ADD REPLY • link 10.7 years ago by Michael 55k

score 1 · Answer 1 · 2014-04-04

exonerate --model est2genome --bestn 5 --showtargetgff 1 --showcigar 1 --hspfilter 1 --subopt 0 --dnawordlen 16

Using the above settings reduces the running time for a single query by a factor of 10 compared to default for my test case. The most efficient reduction is by increasing the wordlength from 12 to 16, sacrificing some sensitivity. Maybe this works also for your case?