dgidb / dgidb-v5

Providing interactions between drugs and genes sourced from a variety of publications and knowledgebases
https://dgidb.org
MIT License
14 stars 2 forks source link

Add retracted notice to PMIDs #492

Open jsstevenson opened 4 months ago

jsstevenson commented 4 months ago
acoffman commented 4 months ago

For this feature, I know @susannasiebert has already done some work on the same issue in CIViC. You may be able to use some of that here. (Mostly the stuff under /server would be the relevant parts for you guys.)

https://github.com/griffithlab/civic-v2/pull/1011

susannasiebert commented 4 months ago

We recently went through this same exercise in CIViC using Retraction Watch as our source for retraction information. PR can be found here: https://github.com/griffithlab/civic-v2/pull/1011. The scraper in particular is here: https://github.com/griffithlab/civic-v2/pull/1011/files#diff-94983e536edb9717f8279d0a5e58b0f82619f1259cff4671d136924323545d73. A few notes that might be helpful if you go that same route:

  1. There are two columns in the Retraction Watch spreadsheet with PMIDs: 'RetractionPubMedID' and 'OriginalPaperPubMedID'. Both should be checked against.
  2. Retraction Watch has the concept of a Retraction Nature which can have values of 'Retraction', 'Expression of concern', 'Reinstatement', or 'Correction'. We decided to display these with varying degrees of importance with 'Retraction' being the only one being displayed in red while the others are displayed in yellow as a warning.
  3. Other interesting columns from Retraction Watch that we decided to import are 'RetractionDate', and 'Reason'.