Alliance - migrating WormBase papers to Alliance ABC

Agenda and Minutes

• We didn't the minutes, so here's an email sent out covering what was discussed.
  • A bit over a week ago you asked what we have in Caltech postgres not in acedb, and I thought that anything user-facing was in acedb, but it turns out corresponding author flags are not. It'd be mildly helpful to display them on the reference webpages users would see, but it'd be pretty important for Alliance to do community outreach, so that when author first pass or something else wants user data, we know who to contact.
  • During the call earlier we talked about the ?Paper model having
    • Author ?Author XREF Paper #Affiliation
  • And Affiliation having
    • ?Affiliation Affiliation_address UNIQUE Text
    • Person ?Person
  • So we suggested adding a 'Corresponding' tag/flag.
  • The Affiliation_address data is also modeled for, but not populated, so as part of author cleanup that Kimberly, Cecilia, and I are doing, we'd populate that when available from the pubmed XML.
  • Kimberly and I discussed more things about cleanup afterwards, and realized that we can also get ORCID for authors from pubmed, so we should also model for that. In ?Person that's modeled like
    • Database ?Database ?Database_field ?Text
  • And at that point, sticking that into the #Affiliation inside an ?Author seemed like a pretty deep/complex structure.
  • So Kimberly and I are thinking of updating the way the ?Author model works. Right now the ?Author object name is the literal author name, but many different people can publish under the same author name, so the author objects are a fairly useless mish-mash of multiple authors's data. In Caltech postgres, the authors are keyed by a sequential author ID, so if we could change ?Authors from names to WBAuthorIDs, we'd be able to properly assign everything that is in #Affiliation into the ?Author object, and further add more things we'd want (Corresponding, ORCID). Kimberly is planning to bring it up tomorrow at the site-wide call.
  • Separately, other things we talked about were:
  • Which papers don't get into Alliance, do we want a species filter (will come up tomorrow)
  • A QC step for Kimberly to look at the DQM submission, probably at the FMS and rounds of feedback (but maybe a pre-submission location)
  • The lag between data being updated at Caltech -> Built into acedb -> released into Alliance.

Wednesday, March 17th

• Review process for generating Reference.json from WB
• QC checks
  • • WBPapers without a PubMed XML at the Alliance
      • Were all papers from WB included in the Reference.json?
    • Author clean-up
      • Author corresponding flag, add to #Affiliation so we can connect in Paper's Author tag ?
        • Author ?Author XREF Paper #Affiliation
        • ?Affiliation Affiliation_address UNIQUE Text
        • Person ?Person
        • Corresponding
      • is Affiliation_address being dumped for authors ? Data not in Caltech postgres (Affiliation in postgres tied to paper).
    • Duplicate PMIDs
• WormBase vs Alliance release schedule
  • Corrections made in our Caltech postgres curation database now - how quickly can we get them to the Alliance?