Question: Issues with using Exonerate 2.2.0
1
gravatar for ykel
4.0 years ago by
ykel10
United States
ykel10 wrote:

 

I have been able to use exonerate to map a large set of transctiptome-derived cDNA sequences (~10,000) to an assembled genome (~300 Mb). But I am only able to use the default parameters for exonerate. For instance, if I were to specity anything using parameters --bestn and --percent, I do not get any output, and the program seems to be stuck in its paths. If I include --showquerygff or --showtargetgff, the program runs with no problem.

I am using the exonerate version 2.2.0-x86_64 on a linux cluster, after securing a node with 16 processors.

Any help is appreciated.

(My first post to BioStars)

debug exonerate • 1.5k views
ADD COMMENTlink modified 4.0 years ago by SES8.1k • written 4.0 years ago by ykel10
1
gravatar for Woa
4.0 years ago by
Woa2.7k
United States
Woa2.7k wrote:

Can you try this version:
http://www.ebi.ac.uk/~guy/exonerate/exonerate-2.4.0.tar.gz

ADD COMMENTlink written 4.0 years ago by Woa2.7k

@Wao: Thank you for your quick suggestion. I will try this one (currently having trouble installing automake-1.4, which seems to be required for installing exonerate from source). Do you have link to binaries of exonerate-2.4.0?

 

ADD REPLYlink written 4.0 years ago by ykel10
0
gravatar for SES
4.0 years ago by
SES8.1k
Vancouver, BC
SES8.1k wrote:

I have experienced the same issue in the past. My assumption was that hits need to be stored in memory when you choose to do this type of filtering, and it could take a long time for output to be written to a file. Though, it's curious you get output when combining other options, so it may be a bug. My workaround has been to just set the alignment constraints and formatting you want, then process the output with a script. This may have been fixed in version 2.4, but if you are comfortable with writing a script, then it may be faster just to take that approach (and the results shouldn't be any different).

ADD COMMENTlink written 4.0 years ago by SES8.1k

The post-processing approach will likely also be easier than trying to figure out how to build automake :)

ADD REPLYlink written 4.0 years ago by SES8.1k

Yes, For --bestn option, I think you may be right: hits may be stored in memory first before being written out. However, the result for setting either of percent and bestn was the same for version 2.2: no output for a 24 hr job!

Our admin installed the version 2.4. Issue with --bestn still persists (which is OK), but --percent is working fine.

Thanks for your responses!

ADD REPLYlink written 4.0 years ago by ykel10

That is what I was referring to when I said I've seen the same issue. You will have to wait a long time, more than a day. Running the program with no filtering, and then post-processing the results is much faster, though maybe not as convenient for everyone. It does allow you to inspect all your results though, like hit number, not just the top hits.

ADD REPLYlink written 4.0 years ago by SES8.1k

Agreed. Weaker hits are looking interesting too, for biological reasons.

ADD REPLYlink written 4.0 years ago by ykel10
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 820 users visited in the last hour