Document how to put PRO IDs in the GPI file

[ValWood]

Please describe your question, idea, or concern.

I wanted to do some Noctua curation using modified PRO forms, but these were not available for PomBAse because they are not currently in our GPI file.
It isn't clear from the GPI spec how to include these. Could you provide an example? @pgaudet mentioned that MGI have imported PRO IDs.

Thanks
Val

[deustp01]

@ukemi @nataled @vanaukenk [warning: scope and feature creep!] I wonder if doing something in this direction, e.g., coining PRO IDs for all Alliance model organism proteins identified in the Reactome computationally projected pathways would be a reasonable chore for us / PRO and a good starting point for model organism GO-CAM curators?

[nataled]

@deustp01 I figured this idea would come up eventually. It's certainly something that can be done, but to do it on a large scale would have to come later, or maybe even left to a second grant cycle...?

@ValWood With each release (and even between releases when an internal update is made) we generate a GPI file for all PRO terms. It would be easy to grab the desired terms from there for any organism of interest The release file is http://purl.obolibrary.org/obo/pr/gpi and the interim update file is https://proconsortium.org/download/development/pro_wv.gpi.

[ValWood]

@vanaukenk has added this to the GO agenda.

tagging @kltm @krchristie

[deustp01]

maybe even left to a second grant cycle.

Problwm - if you're thinking of the pathways2GO project, there will not be a second grant cycle. The original application was stuctured as a one-time project to clean up discrepancies and errors, to provide a harmonized, integrated view of material from Reactome, GO, PRO, Rhea to outside users, and to generate procedures, automated as much as possible, that the participating groups could use to maintain this harmonious state in the future. One explicit aim is to test the use of GO-CAMs derived from Reactome model organism pathways as templates for efficient expert annotation of model organism biological processes, using PRO as the source of model organism protein identifiers, per the current GO-CAM standard.

[ukemi]

We do this for mouse proteoforms because we use them in models. Here are a few examples from our GPI file.

PR:Q19LI2 mA1BG alpha-1B-glycoprotein (mouse) mA1BG|alpha-1-B glycoprotein (mouse) PR:000000001 NCBITaxon:10090 MGI:MGI:2152878 UniProtKB:Q19LI2
PR:Q5YD48 mA1CF APOBEC1 complementation factor (mouse) mA1CF|APOBEC1-stimulating protein (mouse) PR:000000001 NCBITaxon:10090 MGI:MGI:1917115 UniProtKB:Q5YD48
PR:Q5YD48-1 mA1CF/iso:m1 APOBEC1 complementation factor isoform m1 (mouse) mA1CF/iso:m1 PR:000000001 NCBITaxon:10090 MGI:MGI:1917115 UniProtKB:Q5YD48-1
PR:Q5YD48-2 mA1CF/iso:m2 APOBEC1 complementation factor isoform m2 (mouse) mA1CF/iso:m2 PR:000000001 NCBITaxon:10090 MGI:MGI:1917115 UniProtKB:Q5YD48-2
PR:Q5YD48-3 mA1CF/iso:m3 APOBEC1 complementation factor isoform m3 (mouse) mA1CF/iso:m3 PR:000000001 NCBITaxon:10090 MGI:MGI:1917115 UniProtKB:Q5YD48-3
PR:000036218 mATP5ME/Ac:1 ATP synthase subunit e, mitochondrial acetylated 1 (mouse) mATP5ME/Ac:1|UniProtKB:Q06185, Met-1, MOD:00058 PR:000000001 NCBITaxon:10090 MGI:MGI:106636

[ValWood]

I think this can close PRO IDs are there now.