search

geneontology / paint

This curation tool allows curators to make precise assertions as to when functions were gained and lost during evolution and record the evidence (e.g. experimentally supported GO annotations and phylogenetic information including orthology) for those assertions.
Other
4 stars 4 forks source link

notes persisting on loading new family is back #45

Closed krchristie closed 7 years ago

krchristie commented 7 years ago

Hi Suzi,

This issue of notes persisting from the previous family when loading a new one seems to be back.

I was using v2.28.1. The note shown is from the previous family, PTHR38601, not the PTHR15431 family that has just been loaded.

20161130-2 28 1-persistantnote

thanks,

-Karen

selewis commented 7 years ago

Neither of these families appear to be in the svn repository so I can't test with exactly the same thing. In the mean time I tried loading existing+existing and existing+new and the problem is not appearing. Please provide an example I can use directly to work on this.

krchristie commented 7 years ago

PTHR38601 was already in SVN:

bash-3.2$ svn log PTHR38601/PTHR38601.gaf
------------------------------------------------------------------------
r37420 | kchris | 2016-11-30 14:23:35 -0800 (Wed, 30 Nov 2016) | 1 line

not new annotations, examined but unannotated Panther family
------------------------------------------------------------------------
r37411 | kchris | 2016-11-30 11:04:54 -0800 (Wed, 30 Nov 2016) | 1 line

new PAINT annotations
------------------------------------------------------------------------

PTHR15431 is not in SVN and I am not ready to commit yet, so I will send you a zipped copy of my directory.

Additional details required to duplicate this are that I opened the second family (PTHR15431) using "Open using server...". Possibly also relevant is the fact that PTHR15431 had info in its notes file, but I haven't tested this carefully by opening a family that does not have notes.

When opening a subsequent family as "New", the notes field is cleared.

selewis commented 7 years ago

Turns out this had nothing to do with the notes, it was the family PTHR15431 itself causing the issue. Turns out one of the PAINT annotations was a negation of a particular term (GO:42384) that is now obsolete. While PAINT was replacing/updating to the new term for the positive annotations, it was missing these.