3 ms·
I agree with your guess that Watson covers 6/7. From the clinical geneticists I've talked to, variants of unknown significance (where a person differences from
by dekhn 12y ago
I agree with your guess that Watson covers 6/7. From the clinical geneticists I've talked to, variants of unknown significance (where a person differences from the reference, the impact of the difference) are investigated manually through literature searches- often just basic text matches on the variant name.
Regarding paper parsing: there are companies that employ hundreds of PhDs that read papers all day long and enter the data into a database based on ontology (Ingenuity). IBM would license that database and use it as a prior.