[Bioperl-l] GFF3 Specification

Barry Moore barry.utah at gmail.com
Thu Mar 29 15:29:48 EDT 2012

Hi All,

There has been active discussion on the song-devel mailing list over the past 12 months about various ambiguities and unresolved issues with the GFF3 specification.  The SO group is initiating a process to resolve these issues so that GFF3 can continue to serve it's role of unifying genome annotations in a format that promotes collaboration between genome projects and comparison of datasets across a wide variety of genomes.  With the rapid acceleration of genome sequencing, a simple, standard format for genome annotation and comparative genomics is more critical than ever.

Several issues have been raised, and they range from simple requests for clarification to more fundamental questions about the structure of the specification.  We can't address all of these issues in one update to the spec, so we've started the ball rolling with three steps:

Incorporate all the minor changes into a GFF3 1.21 candidate spec (http://www.sequenceontology.org/resources/gff3_1.21.html).
Organize remaining unresolved issues onto a wiki page and start working through those issues one by one (http://www.sequenceontology.org/wiki/index.php?title=GFF3_Developement).
Develop a set of wiki pages to describe 'GFF3 Best Practices' and existing community usage (http://www.sequenceontology.org/wiki/index.php?title=GFF3_best_practices).

We will work through the unresolved issues one at a time - soliciting feedback from the community, and clarify/update the GFF3 spec in a backwards compatible way with existing tools and datasets, adding documenting wiki pages as needed.

We welcome and encourage feedback from the genomics community and gratefully acknowledge those who have been active in the discussion thus far.  Please have a look at the pages described above and join the conversation.  The best place for discussion of all things GFF3 is the SO mailing list (song-devel at lists.sourceforge.net).  Please feel free to re-post this message to relevant mailing lists so that all interested parties can be involved.

On behalf of the SO developers - Thanks.

Barry Moore
Research Scientist
Dept. of Human Genetics
University of Utah
Salt Lake City, UT 84112
(801) 585-3543

More information about the Bioperl-l mailing list