[maker-devel] error in snap training

[Carson Holt]

Make sure you are using 2.31,  and then try the maker2zff filters individually.  If the protein models are not working well, use CEGMA to generate models. It's from the same group as SNAP.  Use cegma2zff for the conversion.

--Carson

Sent from my iPhone

> On Feb 25, 2014, at 2:49 PM, Rebecca Harris <rbharris at uw.edu> wrote:
> 
> Hey - 
> 
> I'm trying to train SNAP and am running into errors. I don't have any EST evidence, just protein. My .gff file reports 10865 genes but when I run maker2zff  -c0 -e0 I get back empty genome files. When I run maker2zff -n, a ton of overlap_prev_exon errors get written to the screen and then with I get to the forge step I get an "impossible error5". Any help would be greatly appreciated.
> 
> Thanks!
> Rebecca
> _______________________________________________
> maker-devel mailing list
> maker-devel at box290.bluehost.com
> http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org