[maker-devel] gff pass through

[Carson Holt]

I see one issue that may be causing the error.  The way structures are modeled in the GFF3 you sent, while technicallye correct under strict GFF3 format specifications, do go against some more common usage guidelines which can be found in the formal GFF3 specifications ( http://www.sequenceontology.org/gff3.shtml ).  They are essentially degenerate gene models in the format transcript/exon, which are expected to be in the format gene/mRNA/exon.  The two level structures given in your GFF3 file are common for evidence but are expected to occur using non-gene model terms (i.e. other than exon).  One quick fix is just to do a replacement of the term transcript for match and exon for match_part, or you can just replace exon with transcript_region and it will probably work.  I'll make some changes to the MAKER code to handle cases where gene model terms are being used for evidence so it won't be an issue in the future.

Copy the following line of text into a terminal window and it will do the replacements for you.  Just remember to give the proper filename and location for the gff3 file.

perl -ane '$_ =~ s/\ttranscript\t/\tmatch\t/; $_ =~ s/\texon\t/\tmatch_part\t/; print $_'  est-part.gff

This command will print the fixed GFF3 contents to the screen. Just redirect the output into a file to make a new file.

Let me know if this change solves the issue.

Thanks,
Carson

On 2/3/10 9:47 AM, "Reith, Michael" <Michael.Reith at nrc-cnrc.gc.ca> wrote:

Hi all,

I'm trying to use the gff pass through method to add the transcript/exon data generated by Cufflinks as EST information.  Maker works fine if I use a fasta file for the EST input, but dies (with "Bus error") if I try to use a gff3 file that I've built from the Cufflinks gtf output and validated with the on-line gff3 validator.  I was wondering:  1). if the gff pass through files referred to in the Maker tutorial (pass.tar.gz) might be available to test our set up (running on Solaris); 2). whether anyone has any bright ideas on why maker would be crashing only with the gff input.

Thanks,
Mike

-----------------------------------------------------
Michael Reith
Principal Research Officer
Functional Genomics Group Leader
NRC Institute for Marine Biosciences
1411 Oxford St.
Halifax, N.S.    B3H 3Z1
Canada

phone:  (902) 426-8276
fax:       (902) 426-9413
email:   michael.reith at nrc.ca
-----------------------------------------------------------
The information contained in this e-mail may contain confidential information intended for a specific individual and purpose. The information is private and is legally protected by law. If you are not the intended recipient, you are hereby notified that any disclosure, copying, distribution or the taking of any action in reliance on the comments of this information is strictly prohibited. If you have received this communication
in error, please notify the sender immediately by telephone or return e-mail.
Thank you.




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://box290.bluehost.com/pipermail/maker-devel_yandell-lab.org/attachments/20100203/42a00e1f/attachment.html>