2.5 years ago by
That should work, but it would require that you first come up with a good search to find the papers, secondly create or find a good method for recognizing gene names in text, and lastly come up with a scoring scheme. That would be reinventing the wheel, though.
A much easier solution would be to go to one of the various databases of disease-gene associations such as DISEASES, DisGeNet, or MalaCards. The DISEASES database is maintained by my group.
You can view genes associated with kidney cancer in the DISEASES database via this URL:
If you want to retrieve the top-100 genes in JSON format, you can do that via the REST API (not documented yet):
Lastly, you can download all text-mined disease-gene associations from DISEASES here: