[maker-devel] Alternative Splicing and ipr_update_gff

[Jason Gallant]

Hi Everyone,

Another nitty gritty question, probably directed at Carson once more.

I decided to make one more go at my maker annotation, this time turning on
alt_splice=1.   I have been keeping keep_preds=1 on to export the “max”
dataset as detailed in Campbell et al (2014), with the hopes of “rescuing”
genes that have IPR domains do not have evidence.

Everything works swimmingly when alt_splice=0, but when activated, the run
behaved normally— I ran gff3_merge and fast_merge to obtain proteins and
transcripts, and found that as predicted the resulting fasta files
contained more proteins than the initial run.

I ran IPR scan and now am trying to update my GFF3 file to obtain the
“maker standard” dataset— what I am noticing is a sudden complaint by the
ipr_update_gff script about use of an uninitialized value.

This appears to have happened to others:
https://groups.google.com/d/msg/maker-devel/dM4WvyghYks/BboRZQLmEF8J

And indeed, I can verify that the first protein listed in my fasta file is
only listed as “augustus_masked”  match/match part in the original GFF3
file.

If I understand the ipr_update_gff script correctly, this transcript will
be ignored because it lacks the mRNA type.  Is this expected naming
behavior for the alternative splicing?  I would have expected the
alternative splice variants to be listed as alternative mRNAs under the
same parent gene?  Is there some sort of misconfiguration or am I expecting
incorrectly?

Hopes for any help you all can provide in diagnosing

Best,
Jason Gallant
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://yandell-lab.org/pipermail/maker-devel_yandell-lab.org/attachments/20151113/9a0be353/attachment-0002.html>