Originally posted in the old forum by Kumar on 18.03.2013 - 16:28
Hi,
I are working on an insect genome re-annotation. Initially ~17000 gene-models were predicted with Glean. In my ab-initio predictions with Augustus I found 200 more gene-models in compare to Glean (that was okay). But when I added the hints generated from unassembled RNASeq data, I got almost "2500" more gene-models. I wonder if I have drawn any mistake?
Thank you!
Kumar
Augustus prediction with RNASeq Hints
Moderator: bioinf
Re: Augustus prediction with RNASeq Hints
by katharina on 19.03.2013 - 10:34
To check that, you should inspect the results visually, e.g. in a GBrowse or a UCSC-Browser, in context with the RNA-Seq alignments.
Katharina
To check that, you should inspect the results visually, e.g. in a GBrowse or a UCSC-Browser, in context with the RNA-Seq alignments.
Katharina
Re: Augustus prediction with RNASeq Hints
by Kumar on 19.03.2013 - 14:09
Thanks, but is it possible to have increment of ~2500 gene models while adding the intronic hints? if yes then why? I used the protocol mentioned at http://bioinf.uni-greifswald.de/bioinf/ ... seq.Tophat
Thanks you!
Kumar
Thanks, but is it possible to have increment of ~2500 gene models while adding the intronic hints? if yes then why? I used the protocol mentioned at http://bioinf.uni-greifswald.de/bioinf/ ... seq.Tophat
Thanks you!
Kumar
Re: Augustus prediction with RNASeq Hints
by katharina on 19.03.2013 - 14:23
Obviously, it is possible. The question is whether it is good or whether you should do something about it (e.g. modify parameters in the extrinsic.cfg file). Visual inspection might be most helpful to get an idea.
Katharina
Obviously, it is possible. The question is whether it is good or whether you should do something about it (e.g. modify parameters in the extrinsic.cfg file). Visual inspection might be most helpful to get an idea.
Katharina