[maker-devel] choosing the right gene model

[Xabier Vázquez-Campos]

Actually, it's a fungal genome. Although not very typical, almost half of
it are repeats. Worth mention that Genemark generates a lot of predictions
that overlap LTRs and other complex repeats, something that neither SNAP or
Augustus do. Have you seen this before?

On 14 Oct. 2017 02:56, "Carson Holt" <carsonhh at gmail.com> wrote:

> Both transcript and protein evidence will go into the AED calculation for
> overlap support. So in both cases the chosen model had better overlap
> (protein evidence will not count toward the eAED overlap calculation if it
> is out of frame with the model it is supposed to be supporting). The larger
> merged model generates a clutering affect on it’s evidence, so it’s
> evidence set for AED calculation is slightly different than the SNAP and
> Augustus model would generate. In both cases, I think GeneMark is hurting
> more than it is helping. You may want to just drop it from the analysis
> (unless it’s a fungi, I often find GeneMark can have that affect).
>
> —Carson
>
>
> On Oct 12, 2017, at 12:09 AM, Xabier Vázquez-Campos <xvazquezc at gmail.com>
> wrote:
>
> Hi there,
>
> I was visualising the annotations and I realised that in some cases, what
> it seems to be a gene is splitted according to one of the gene models,
> despite that the other 2, est2genome and prot2genome suggest that it isn't
> the case.
>
> <split-gene.png>
>
> Although the opposite also happens.
>
> <merged-gene.png>
> ​
> For some reason, the "out of place" model is always (or almost) the one
> from Genemark.
>
> How much weight does carry the RNAseq and protein data on this decision
> (if any)?
> How exactly is the final gene selected?
>
> Cheers,
> Xabi
>
> --
> Xabier Vázquez-Campos, *PhD*
> *Research Associate*
> NSW Systems Biology Initiative
> School of Biotechnology and Biomolecular Sciences
> The University of New South Wales
> Sydney NSW 2052 AUSTRALIA
> _______________________________________________
> maker-devel mailing list
> maker-devel at box290.bluehost.com
> http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://yandell-lab.org/pipermail/maker-devel_yandell-lab.org/attachments/20171014/fdf7a53e/attachment-0003.html>