search

BrendelGroup / AEGeAn

Integrated toolkit for analysis and evaluation of annotated genomes
http://brendelgroup.github.io/AEGeAn
ISC License
24 stars 10 forks source link

src/core/AgnTranscriptClique.c #155

Closed alpapan closed 8 years ago

alpapan commented 9 years ago

Hello

This happens using a particular prediction set. It was created via WebApollo so it may have a bug from the export. How can we debug it?

Assertion failed: (gt_range_contains(&locusrange, &transrange) && gt_range_length(&locusrange) == strlen(modelvector)), function clique_vector_update, file src/core/AgnTranscriptClique.c, line 539. This is a bug, please report it at https://github.com/standage/AEGeAn/issues.

thanks a

alpapan commented 9 years ago

Command was ~/software/JAMg/3rd_party/bin/parseval -d -V -x NCBI -y CURATIONS ncbi.gff3.canon annotations.gff.Approved_Accepted.cleanedup.gff3.canon > CURATIONS_VS_NCBI.txt

alpapan commented 9 years ago

Ignore me, this is almost certainly a bug from somewhere upstream... sorry!

alpapan commented 9 years ago

i'm sorry, but this still happens even after i solved all the upstream issues i could think.

it also happens in the reciprocal search.

If it was possible to print which loci are being compared (with the debug flag) then it would be easier to find out what is happening

alpapan commented 9 years ago

Hello Found that with a very high --delta (800) the crash goes away. so I manually identified 5 genes (out of 1400) that caused the error. they are good proteins so i don't know why that happened. let me know if you'd like the files. thanks a

standage commented 8 years ago

Indeed, it would be much easier for me to help you debug if you posted your small (5 genes) data set.

alpapan commented 8 years ago

thnx, emailed!

standage commented 8 years ago

Ok, sorry for the delayed response. I've had a look at the example data you provided, and it's failing on a gene where the range of the mRNA feature extends beyond the range of the gene feature (NW_004523309.1 from 9714850 to 9718050).

So this is clearly a problem with the data file (and probably WebApollo's GFF3 export feature). But ParsEval should do a better job of handling this. I'll work on a fix for this soon.

standage commented 8 years ago

I just pushed a new commit to the master branch. Please let me know if this solves your problem.

standage commented 8 years ago

I'm going to go ahead and close this issue. Feel free to open it if you are still experiencing related issues.