[maker-devel] Mapping gene names

[Torsten Seemann]

Sajeet,

Brief test of barrnap suggests that it does not perform well on rRNA genes
> with introns such as those found in fungal mitochondria. Setting a lower
> threshold for --reject and --evalue helps, but is not enough.
> Looks like I cannot abandon rnammer for now.
> FYI - if you want to test barrnap with fungal mitochondria, use --kingdom
> bacteria because they have 23S and 16S unlike the human mitochondria.
>

This is good feedback. Paul Gardner also mentioned the intron issue. A
"fungi" kingdom is clearly needed. I am not a mycologist so any assistance
is coming up with a detailed rRNA architecture for eukaryotict phyla etc is
something I have started but need assistance with. Adjustment of nhmmer
alignment parameters could be done to improve the intronic rRNAs too.

Here is what I have so far in terms of models:
https://github.com/Victorian-Bioinformatics-Consortium/barrnap/blob/master/README.md#data-sources-for-hmm-models

- do i need to split euk into protist / plant / animal / fungi?
- should the current 'mito' be places inside the current 'euk' ? as mito
data is likely to end up in assemblies, but keep separate for mito-only
data?
- plastids, chloroplasts, apicoplasts; i am not sure of the subtleties of
these organelles' rRNA but am willing to learn.

Thank you again for testing. Any help appreciated,

-- 

*--Dr Torsten Seemann--Victorian Bioinformatics Consortium, Monash
University, AUSTRALIA*

*--Life Sciences Computation Centre, VLSCI, Parkville, AUSTRALIA
--http://www.bioinformatics.net.au/ <http://www.bioinformatics.net.au/>*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://yandell-lab.org/pipermail/maker-devel_yandell-lab.org/attachments/20140516/6206e421/attachment-0003.html>