The protein_variant table has a column for pubmed ids, although there are only 10 for variants in the therapy_genotype table. It's also not clear what the source is supporting (amino acid coordinates, funcational impact, reference transcript, etc.).
As a side note, the first variant I checked, BRAF G466V missense mutation, PMID:15035987 is not referenced in the article, although G465V is discussed. These may need to be reviewed manually.
The protein_variant table has a column for pubmed ids, although there are only 10 for variants in the therapy_genotype table. It's also not clear what the source is supporting (amino acid coordinates, funcational impact, reference transcript, etc.).
As a side note, the first variant I checked, BRAF G466V missense mutation, PMID:15035987 is not referenced in the article, although G465V is discussed. These may need to be reviewed manually.