From the about page:
EVEX is a text mining resource built on top of PubMed abstracts and PubMed Central full text articles. It contains over 40 million bio-molecular events among more than 76 million automatically extracted gene/protein name mentions. The text mining data further has been enriched with gene identifiers and gene families from Ensembl and HomoloGene, providing homology-based event generalizations. EVEX presents both direct and indirect associations between genes and proteins, enabling explorative browsing of relevant literature.
Ok, it’s not web-scale but it is important information. 😉
What I find the most interesting is the “…direct and indirect associations between genes and proteins, enabling explorative browsing of the relevant literature.”
See their tutorial on direct and indirect associations.
I think part of the lesson here is that no matter how gifted, a topic map with static associations limits a user’s ability to explore the infoverse.
That may work quite well where uniform advice, even if incorrect, is preferred over exploration. However, in rapidly changing areas like medical research, static associations could be more of a hindrance than a boon.