-
Notifications
You must be signed in to change notification settings - Fork 89
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
IBAs from PAINT not being applied to GOA files? #1061
Comments
Quickest way forward is probably feedback from @dougli1sqrd So far, no obvious files changes in Makefile and merge code that I can see in the relevant time period, but may be (likely?) missing something. |
If this was not intended, we'll need to make a release note and reclassify this as a showstopper. |
@pgaudet Would you know of any reason over the past couple of months that valid PAINT annotations would not be applied to GOA (goa_*) output? |
Looking at possible changes in ontobio If cannot get traction on this soon, will test to an earlier ontobio version. |
I don't know. I hadn't noticed. |
@kltm Any idea what's going on ? Will this be resolved for the upcoming release ? Thanks, Pascale |
We have no IBA annotations for human - isn't this a showstopper ? I would expect that changes our annotation corpus quite a bit (and any analysis done with the GO data). |
@pgaudet It is indeed a "showstopper", which is why it has been labeled as "showstopper". Release will be suspended until this has been fixed. |
The release build log shows that perhaps goa_chicken has properly merged paint_goa_chicken. Has anyone yet verified goa_chicken? |
It looks like goa_chicken worked for some reason, and not the other goa datasets:
|
In snapshot:
So it looks like PAINT is properly being applied at this point. Reopen if there is still an issue. |
Got a helpdesk email about missing cow annotations, I've narrowed it down to there not being a single IBA in goa_cow.gaf (test: UniProt O18971 definitely has IBAs in Protein2GO). @dougli1sqrd I'm assuming this is the relevant ticket? |
@suzialeksander @dougli1sqrd looking at our incoming products, it seems that there is no paint_goa_cow-src.gaf.gz (or any other paint_goa_cow*) as might be expected. The two possibilities are that: 1) PAINT upstream does not supply cow or that 2) we are somehow not getting them (metadata issues)? |
@suzialeksander did they say what annotations are missing? Were they IBAs? Yeah we filter out IBAs from all non-PAINT sources. We expect IBAs to come in from a PAINT gaf. But as @kltm pointed out, we don't have a corresponding paint_goa_cow*.gaf. It's not in the metadata. |
@dustine32 Is there actually a cow paint file that we're missing upstream? |
@dougli1sqrd Not a full list of errors, but the user provided GO IDs that are missing for PPARG. I can't get others easily as BioMart is down. I have a feeling these are all IBAs but only checked two, then looked at the IBA-less GAF:
|
@kltm @dougli1sqrd Yeah the cow IBAs are in paint_other.gaf. Checking ftp://ftp.pantherdb.org/downloads/paint/presubmission/gene_association.paint_other.gaf.gz:
Assuming 9913 is the taxon ID for cow. |
@dustine32 9913 is what PANTHER uses for cow AKA Bos taurus (short letter code BOVIN) |
Okay, great--thank you for the input. It seems that things are working as intended:
So, in summary, this is all as it should be. In the case of cow, there are two files that one needs to look at. In the future, when we move from resource-centric to species-centric, this friction would disappear. |
Great, thanks for looking into this. |
It seems that from the http://release.geneontology.org/2019-04-17/ release, PAINT files do not seem to be applied to the GOA (goa_*) resources as they are to the others--neither header no data. This is appears to be true for all goa_ and non-goa_ files sampled so far, comparing to http://release.geneontology.org/2019-03-18/ .
As part of this exploration, I also noticed that when unpacking goa_*.gaf.gz files, the "inner" archive name was "goa_*_valid.gaf" rather than the expected "goa_*.gaf", as would be expected. All non-GOA files that had PAINT applied properly did not seem to have this property. Unsure if related, but a sign of trouble and should be fixed if not.
Original reference:
geneontology/helpdesk#206
The text was updated successfully, but these errors were encountered: