[maker-devel] Most scaffolds fail with BadParameter error, Maker on TACC Lonestar

[Card, Daren C]

Hello,

I’ve tried to run Maker on TACC Lonestar (4, trying to squeeze some last things in before deprecation), but I haven’t had much success. I get Maker to run, but only 28 proteins/transcripts are annotated and most scaffolds fail to finish properly, according to the …master_datastore_index.log. In my STDERR, I see a consistent error show up for many scaffolds:

------------- EXCEPTION: Bio::Root::BadParameter -------------
MSG: ' 7.5' is not a valid score
VALUE: 7.5
STACK: Error::throw
STACK: Bio::Root::Root::throw /opt/apps/maker/2.30/bin/../perl/lib/Bio/Root/Root.pm:486
STACK: Bio::SeqFeature::Generic::score /opt/apps/maker/2.30/bin/../perl/lib/Bio/SeqFeature/Generic.pm:468
STACK: GFFDB::_ary_to_features /opt/apps/maker/2.30/bin/../lib/GFFDB.pm:891
STACK: GFFDB::phathits_on_chunk /opt/apps/maker/2.30/bin/../lib/GFFDB.pm:534
STACK: Process::MpiChunk::_go /opt/apps/maker/2.30/bin/../lib/Process/MpiChunk.pm:756
STACK: Process::MpiChunk::run /opt/apps/maker/2.30/bin/../lib/Process/MpiChunk.pm:341
STACK: main::node_thread /opt/apps/maker/2.30/bin/maker:1433
STACK: threads::new /opt/apps/maker/2.30/bin/../perl/lib/forks.pm:799
STACK: /opt/apps/maker/2.30/bin/maker:901
--------------------------------------------------------------
--> rank=18, hostname=c304-113.ls4.tacc.utexas.edu
ERROR: Failed while doing repeat masking
ERROR: Chunk failed at level:0, tier_type:1
FAILED CONTIG:scaffold279|size418813

The ‘7.5’ value can vary between errors, but other than that and the scaffold ID, the rest of the error message is the same. I obviously don’t have the expertise to diagnose the issue here, but I’m hoping someone can help me sort this out.

A quick, unrelated question, is whether the Yandell lab (or anyone else) has a script that will produce a CDS file (multi-FASTA file) from a GFF annotation and FASTA genome sequence. I’m trying to produce a CDS from some NCBI genomes (annoying that it isn’t already included from NCBI), but the script I produced to do this is giving some suspect results. I figured if anyone had a well-tested script for this purpose, it would be someone on this list.

Best,
Daren


Daren Card
Ph.D. Candidate
Castoe Lab
University of Texas at Arlington
dc...@uta.edu
www.darencard.net


_______________________________________________
maker-devel mailing list
maker...@box290.bluehost.com
http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org

[Carson Holt]

It’s failing because there is something wrong with the format of the input GFF file. It might not be GFF3, it may be GTF format, it may have mixed types (not just gene/mRNA/exon/CDS models), or it may have a missing Parent= or ID= tag required to generate the proper feature relationship.

You can try and use GAL (http://www.sequenceontology.org/software/GAL.html) to help validate of convert the format.

Also note the message —> MSG: ' 7.5' is not a valid score

There is an extra whitespace inside the single quotes which probably means you have contaminating whitespace before the value. GFF3 is tab delimited, space characters are not permitted, and if required must be escaped following URI escaping convention.

—Carson