CID__norman_s60.csv has 'extra' rows
CID__norman_s60.csv was obtained via getPcCand.trans(' ')
, corresponding to the query
https://pubchem.ncbi.nlm.nih.gov/sdq/sdqagent.cgi?infmt=json&outfmt=csv&query={\"download\":\"*\",\"collection\":\"norman_s60\",\"where\":{\"ands\":[{\"cid\":\" \"}]},\"order\":[\"relevancescore,desc\"],\"start\":1,\"limit\":10000,\"downloadfilename\":\"CID_ _norman_s60.csv\"}"
It is assumed to be the underlying database for PubChem's Transformations section, comprising a merge of Norman SLE S60 + S66 + S68.
However, it seems to have 'extra' rows and must be used with caution in its current state, especially in the absence of info about how a blank cid (as above) translated to a SDQ query gives this output. (Basically, we should ask someone what it really is.)
e.g. Nicotine - it only has 2 TPs, but they are represented by 4 rows with unique gid
.
Values in column cid
should normally match predecessorcid
, but here not the case.
When querying web services using getPcCand.trans(89594)
, one correctly gets only the 2 results (gid 58044806 and 58044808), not 4.