[maker-devel] Maker failed to annotated whole gene

[Carson Holt]

Look at the evidence and annotation in a browser (Apollo, IGV, etc). If the evidence is not bridging the exons, then that is why. Look at the evidence alignment to see if HSP’s repeat the same alignment to multiple spots (looks like 4 exons but is essentially a duplication of identical HSPs). Look at the gene predictor performance across the genome to see if perhaps the gene predictor is not trained well. Zoom in on the evidence alignments in a browser, do you see long strings of NNNN that mean parts of the assembly are missing so a working annotation cannot be made to bridge the exons. Are there exonerate alignments with canonical splice site support. Or there may be assembly errors generating early stop codons, so the gene predictor cannot create a single model covering the entire locus, but rather generates multiple broken loci.

—Carson


> On Nov 24, 2020, at 7:58 PM, Diana Moreno Santillán <dianad.mosa at gmail.com> wrote:
> 
> Hello, 
> I noticed that some genes on my maker runs were annotated like fragmented pieces instead of a single gene. 
> 
> For example,  for a gene composed by 4 exons, I was expecting  to have the 4 exons concatenated in a single protein sequence. I performed annotations in several species and for some of them I have only one gene annotated, i.e with the 4 exons merged in a single protein sequence. But for other species, with the same protein evidence and maker.ctl parameters I got 4 "genes" evidence instead of one. Actually on the blast results is pretty clear how they are part of the same gene at different positions. This is an issue because I'm doing gene families expansions and contraction and the analysis detects as this gene is being expanded, as it has 3 more copies, but in reality they are part of the same gene. 
> 
> Have you seen this before? Could you help me to seek for a solution?
> 
> Thank you
> _______________________________________________
> maker-devel mailing list
> maker-devel at yandell-lab.org
> http://yandell-lab.org/mailman/listinfo/maker-devel_yandell-lab.org

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 1376 bytes
Desc: not available
URL: <http://yandell-lab.org/pipermail/maker-devel_yandell-lab.org/attachments/20210108/e2d68da5/attachment-0002.p7s>