You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, thanks for the amazing (and fast) tool to handle graphs!
I have been writing some code myself to parse GFA files (v1.0 based on https://github.com/GFA-spec/GFA-spec). However, I found an irregularity in the GFA output of odgi view. This is how I convert my .og files (from Minigraph-Cactus specifically) to GFA:
The issue I found is that some of the links implied in the path don't exist. E.g.: a P-line might have the sequence 23231-,23232+ but this does not exist as L-line. When searching myself, I do however find L 23232 - 23231 + 0M; thus with the two nodes switched. I observed this for multiple GFAs created by odgi; but see example yeast-pg.full.og.gz and resulting example yeast.mc.gfa.gz (GZIP'ed as per GitHub's request).
I tested this with someone else's code too to confirm it is not an issue of my own code (using gfatk: https://github.com/tolkit/gfatk):
$ gfatk path -a yeast.mc.gfa
>DBVPG6044#0#chrI#0
Error: This link: 23231-|23232+ - does not occur in the input GFA. Perhaps re-consider the input path?
By the way, I'm using odgi as installed from bioconda:
$ odgi version
v0.8.6-0-ge647844f
Is this a bug in odgi view and fixable? Or perhaps is it a more common misinterpretation of the GFA v1.0 format?
The text was updated successfully, but these errors were encountered:
Hi, thanks for the amazing (and fast) tool to handle graphs!
I have been writing some code myself to parse GFA files (v1.0 based on https://github.com/GFA-spec/GFA-spec). However, I found an irregularity in the GFA output of
odgi view
. This is how I convert my.og
files (from Minigraph-Cactus specifically) to GFA:odgi view -gi /path/to/graph.full.og > graph.mc.gfa
The issue I found is that some of the links implied in the path don't exist. E.g.: a P-line might have the sequence
23231-,23232+
but this does not exist as L-line. When searching myself, I do however findL 23232 - 23231 + 0M
; thus with the two nodes switched. I observed this for multiple GFAs created byodgi
; but see example yeast-pg.full.og.gz and resulting example yeast.mc.gfa.gz (GZIP'ed as per GitHub's request).I tested this with someone else's code too to confirm it is not an issue of my own code (using
gfatk
: https://github.com/tolkit/gfatk):By the way, I'm using
odgi
as installed from bioconda:Is this a bug in
odgi view
and fixable? Or perhaps is it a more common misinterpretation of the GFA v1.0 format?The text was updated successfully, but these errors were encountered: