[maker-devel] MAKER v3 Fgenesh ERROR

[Carson Holt]

If you you pass in a gff3, MAKER uses the name in the gff3.

—Carson

Sent from my iPhone

> On Mar 11, 2019, at 11:22 AM, Kapeel Chougule <kapeelc at gmail.com> wrote:
> 
> Thanks Carson,
> 
> I will try the settings you suggested.
> 
> Also, for those failed contigs I ran fgenesh standalone and passed the results to pred_gff in maker_opts.ctl. The run was successful. The only issue with this was the name attribute in column predicted by MAKER had a very different ID compared to other genes predicted.
> 
> E.g
> Chr10   maker   mRNA    33029   45468   .       -       .       ID=maker-Chr10-pred_gff_Fgenesh-gene-0.5-mRNA-1;Parent=maker-Chr10-pred_gff_Fgenesh-gene-0.5;Name=Chr10mRNA_1;_AED=0.21;_eAED=0.21;_QI=76|0.88|1|1|1|1|10|315|418
> 
> Correct Name attribute :
> Chr10   maker   mRNA    33029   45468   .       -       .       ID=maker-Chr10-pred_gff_Fgenesh-gene-0.5-mRNA-1;Parent=maker-Chr10-pred_gff_Fgenesh-gene-0.5;Name=maker-Chr10-pred_gff_Fgenesh-gene-0.5-mRNA-1;_AED=0.21;_eAED=0.21;_QI=76|0.88|1|1|1|1|10|315|418
> 
> This was also reflected in the fasta headers. 
> 
> Best
> 
> Kapeel
> 
> 
> 
>> On Mon, Mar 11, 2019 at 12:55 PM Carson Holt <carsonhh at gmail.com> wrote:
>> Sorry for the slow reply. This got lost somehow. The file from the mailing list is 5 years old. Don’t use it. You are actually downgrading, not upgrading. You will break things.
>> 
>> Run maker outside of MPI (if you are using MPI). There might be additional error message, that are further back that are getting covered by extra MPI output.  I don’t actually have a license for fgenesh to test with any more, so debugging it for you may be difficult.  My guess is that a more recent version has slightly different output format which is breaking the parser. Also you need to restore the previous fgenesh.pm file before you can run any tests (may need to reinstall MAKER).  
>> 
>> Also try MAKER3 vs MAKER2. They both run fgenesh slightly differently.
>> 
>> Basically after a couple of tests/error messages I may be able to send you something that would at least allow you to run fgenesh outside of MAKER to see what’s going on.
>> 
>> —Carson
>> 
>> 
>> 
>>> On Feb 7, 2019, at 12:43 PM, Kapeel Chougule <kapeelc at gmail.com> wrote:
>>> 
>>> Hi, Carson
>>> 
>>> I have been getting this error with fgenesh tool within MAKER. It runs ok with most of the assembly contigs but seems to fail on one contig or part of the contig with the below error
>>> 
>>> Widget::fgenesh:
>>> /mnt/grid/ware/hpc/home/data/mcampbel/applications/maker/bin/../lib/Widget/fgenesh/fgenesh_wrap /mnt/grid/ware/hpc_norepl/data/data/programs/fgenesh_v8/fgenesh_suite_v8.0.0a/fgenesh /sonas-hs/ware/hpc_norepl/data/programs/fgenesh_v8/fgenesh_suite_v8.0.0a/Zeamays.mpar.dat.new /tmp/uge/53139300.1.primary.q/maker_j3ttxX/6/6_1.600610-613023.Zeamays.mpar.dat.new.auto_annotator.fgenesh.fasta -exon_table:/tmp/uge/53139300.1.primary.q/maker_j3ttxX/6/6_1.600610-613023.Zeamays.mpar.dat.new.auto_annotator.xdef.fgenesh > /tmp/uge/53139300.1.primary.q/maker_j3ttxX/6/6_1.600610-613023.Zeamays.mpar.dat.new.auto_annotator.fgenesh
>>> #-------------------------------#
>>>  ...processing 9 of 24
>>>  ...processing 8 of 28
>>>  ...processing 10 of 24
>>>  ...processing 9 of 28
>>>  ...processing 11 of 24
>>>  ...processing 10 of 28
>>>  ...processing 12 of 24
>>>  ...processing 11 of 28
>>> deleted:0 genes
>>> ERROR: FgenesH failed
>>> --> rank=14, hostname=bnbcompute50
>>> ERROR: Failed while annotating transcripts
>>> ERROR: Chunk failed at level:1, tier_type:4
>>> FAILED CONTIG:Super-Scaffold_14.2_contig2
>>> 
>>> I updated the perl module fgenesh.pm as suggested in the previous threads. Attached are the  maker_opts.ctl and STDERR log file.
>>> 
>>> Thanks
>>> 
>>> Kapeel
>>> 
>>>  
>>> 
>>> -- 
>>> 
>>> Kapeel Chougule
>>> Computational Scientist Developer II
>>> One Bungtown Road Cold Spring Harbor, NY 11724
>>> http://www.warelab.org/
>>> 
>>> <maker_opts.ctl><stderr.log>_______________________________________________
>>> maker-devel mailing list
>>> maker-devel at box290.bluehost.com
>>> http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org
>> 
> 
> 
> -- 
> 
> Kapeel Chougule
> Computational Scientist Developer II
> One Bungtown Road Cold Spring Harbor, NY 11724
> http://www.warelab.org/
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://yandell-lab.org/pipermail/maker-devel_yandell-lab.org/attachments/20190311/bcbebead/attachment-0003.html>