We looked at the gene-disease associations (GDA), but the info below probably applies to the variant-disease associations (VDA) as well.
the "Evidence for GDAs" tab has "association type" (type of gene-disease relationship) and supporting sentence info for PMIDs ➡️ this looks particularly interesting!
the "Summary of GDAs" tab has the EL (evidence level), and EI (evidence index) for each gene-disease relationship, and a pLI measure.
MyDisease is using the tsv downloads, which lack this info (look at the README). However, we realized that the SQLite file has this info! So it may be a good idea to rewrite the parser to use the SQLite data and provide more info. This may involve changing the structure of the documents/objects.
(Side note: we do have DSI/DPI in MyDisease right now, but these are actually gene-level properties...not directly related to the gene-disease relationship)
@DylanWelzel and I discussed today (2/22) how MyDisease's DisGeNET info lacks some of the useful info that we see on DisGeNET's website (example: https://www.disgenet.org/browser/0/1/1/C0001080/_a/_b./)
We looked at the gene-disease associations (GDA), but the info below probably applies to the variant-disease associations (VDA) as well.
MyDisease is using the tsv downloads, which lack this info (look at the README). However, we realized that the SQLite file has this info! So it may be a good idea to rewrite the parser to use the SQLite data and provide more info. This may involve changing the structure of the documents/objects.
(Side note: we do have DSI/DPI in MyDisease right now, but these are actually gene-level properties...not directly related to the gene-disease relationship)