Open sabrinatoro opened 8 years ago
Hello, We are still getting the "invalid number of columns (expected 15, got 17) warning in our gaf-validation-report. The latest one: [http://build.berkeleybop.org/job/gaf-check-zfin/134/]
Could someone have a look at it and tell us if there is an actual problem with out GAF file, please? Thank you! Sabrina
CC: @cmungall
I seem to recall another open ticket about this (incorrect format version identification)--I'll take a look around for it. I've assigned @cmungall for further comment when he gets back.
I looked in my notes, and found a the following comment:
18:49:30,958 WARN owltools.gaf.GAFParser.loadNext(GAFParser.java:171) - Got invalid number of columns for row (expected 15, got 17). The '4245' row is ignored. : MGI MGI:2685243 Zscan4c GO:0045950 MGI:MGI:4440771|PMID:20336070 IMP P zinc finger and SCAN domain containing 4C LOC245109 protein taxon:10090 20111115 MGI regulates_o_occurs_in(CL:0002322) VEGA:OTTMUSP00000044032
IIRC, we decided it was harmless (for my use case) and that we should look into better warning suppression in some cases. This may be a place to start for getting the reports more reasonable.
Just a thought: do you declare that this is a GAF 2.1? (line 1, !gaf-version: 2.1) GAF 1.0 specifies 15 columns rather than 17, so is this checker thinking this is a GAF 1.0?
Thank you all! @hdrabkin : yes, we do declare that it is a GAF 2.1 file. line 1 : !gaf_version: 2.1
@kltm : I thought it should be harmless too (especially since it is 'only' a warning. But I wanted to make sure that this is not creating other problems which I am not aware.
Thanks again!
Even if harmless, it does cause rather spectacular log files. We should aim to suppress unneeded warnings no matter.
Hello, ZFIN updated their GAF file to support the "AND/OR" logic (comma/pipe) in column 8. Since this update, our GAF validation checks have been unstable. See the last one here: http://build.berkeleybop.org/job/gaf-check-zfin/128/
The gaf-validation-report.txt reports WARNING : Got invalid number of columns (expected 15, got 17)
Is this a bug in the GAF checking process related to our updated GAF version? Or is there actually a problem with our GAF file?
Thank you very much Sabrina