davecap
commented
10 years ago We're working on it now. I'm not sure exactly how long it will take, but probably around 1-2 weeks to get a preliminary version of Medline up.
The pass-through links to the full-text articles might be a bit more involved (except for articles that are included in PubMed Central).
Currently we just display (completely and totally inadequate) pubmed ID's for articles. This requires users to click out to pubmed to even see the title or abstract. We should Have some sort of compact way to render these in the UI. This likely requires both some design help from SolveBio as well as some API access. In my mind, we could go pretty far with just the following:
If we could somehow detect open access articles and slurp in the contents (either as an HTML page or PDF document) we could lay the groundwork for experimenting with assistive technology for rapidly screening manuscripts such as highlighting pertinent sections that match a patient's phenotype, HPO terms associated with the gene or disease, etc... This is something we'd love some help from SolveBio on.
@davecap any thoughts on how feasible it would be to get this pubmed stuff going? I think clean access to open access text would be huge, even if all you do is give us a URL where we know we can reliably get it.