Ensure that now-present chain identifiers in the GOA GPIs are correctly propagated

[kltm]

Recently, GOA started supplying chain identifiers in their GPIs this ticket is to ensure that they are

• being correctly propagated into the NEO load and
• do not cause any issue with Noctua curation

An example would be: UniProtKB:A0A3B3IS91-PRO_0000454209

tagging @pgaudet @vanaukenk

[kltm]

@pgaudet These identifiers do not seem to appear in my load created on 2025-01-31. Could you confirm the date and location, just to be sure?

[kltm]

@pgaudet @vanaukenk I'm not sure what project this should belong to; I've temporarily added it to the pipeline/data QC rolling project. Please feel free to move it.

[pgaudet]

These identifiers do not seem to appear in my load created on 2025-01-31. Could you confirm the date and location, just to be sure?

That seems right. The latest 'official' file according to our yaml is timestamped 2024-12-21 17:57

The new files for the future GOC-GOA data exchange pipeline do contain the chains, if you want to test.

Thanks, Pascale

[kltm]

@pgaudet Okay, I had assumed that these were in the "main" GAFs and GPIs we were picking up, but this looks like it's a little more in-progress, for the future.
It looks like the files at https://ftp.ebi.ac.uk/pub/contrib/goa/panther_proteomes/ would be for the post-GOA derivative production at the GO site. I'll download the current version there and try it in our derivatives-from-goa pipeline.

[pgaudet]

Sounds good !

[kltm]

Well, thinking through this, even going through the new pipeline, as it doesn't examine GPI files, there is no real test. Can I confirm with you that it only affect GPI files? Is it only new entities added, with no change in current entities (give or take)?

[pgaudet]

The only existing entities that I can think of are the Sars-Cov genes, for which we had somehow loaded chains back in 2020.
But these should be included here, I assume in the same format.