Open erikyao opened 3 years ago
@everaldorodrigo @andrewsu @newgene
Was this issue ever addressed? It may be useful since BTE uses this info
In practice though (based on the query/result below), it doesn't look like the duplicated drug mechanisms result in duplicated results in BTE. So I think this can remain a lower-priority fix.
E.g. on
CHEMBL1201631
Compound Report Card, there is only one drug mechanism record. However, http://mychem.info/v1/chem/CHEMBL1201631 shows two.The duplication comes from the original source file, https://www.ebi.ac.uk/chembl/api/data/mechanism.json.
Same problem may occur to drug indication records.
Suggested fix: remove duplicates after chembl_parser.py#L52; pay attention to nested lists when hashing records.