Hi again,
I think I've detected a bug on PubChem's side.
e.g. Predecessor Codeine (CID 5284371) is known to have the Transformation Product/Successor Morphine (CID 5288826).
Structured Data Query of 5284371 gives 5288826 correctly, as it should. (using getPcCand.trans)
However, query of 5288826 erroneously returns...ITSELF!
https://pubchem.ncbi.nlm.nih.gov/sdq/sdqagent.cgi?infmt=json&outfmt=csv&query={\"download\":\"*\",\"collection\":\"norman_s60\",\"where\":{\"ands\":[{\"cid\":\"5288826\"}]},\"order\":[\"relevancescore,desc\"],\"start\":1,\"limit\":10000,\"downloadfilename\":\"CID_5288826_norman_s60.csv\"}"
As described in https://git-r3lab.uni.lu/eci/pubchem/-/issues/3, the cid value should match the predecessorcid value in the query output, but this is not the case here.
Noticeably, the gids are different.
Same issue has been detected for Diazepam and Nordiazepam (CID 2997),
https://pubchem.ncbi.nlm.nih.gov/sdq/sdqagent.cgi?infmt=json&outfmt=csv&query={\"download\":\"*\",\"collection\":\"norman_s60\",\"where\":{\"ands\":[{\"cid\":\"2997\"}]},\"order\":[\"relevancescore,desc\"],\"start\":1,\"limit\":10000,\"downloadfilename\":\"CID_2997_norman_s60.csv\"}"
and Caffeine and Theophylline (CID 2153).
https://pubchem.ncbi.nlm.nih.gov/sdq/sdqagent.cgi?infmt=json&outfmt=csv&query={\"download\":\"*\",\"collection\":\"norman_s60\",\"where\":{\"ands\":[{\"cid\":\"2153\"}]},\"order\":[\"relevancescore,desc\"],\"start\":1,\"limit\":10000,\"downloadfilename\":\"CID_2153_norman_s60.csv\"}"