scorbett-ebi
commented
1 year ago CHEMBL1145819 (doc_id = 17066, doi 10.1021/jm030030n) is a review paper. The original paper the quoted measurement is taken from is CHEMBL1135108 (doc_id = 4829, doi 10.1016/s0960-894x(02)00172-5). The value in the original paper is a pKi of 8.3 which has been correctly assigned to compound CHEMBL29410 (molregno = 38499) under the document CHEMBL1135108 (activity_id = 241135).
The compound CHEMBL98511 (molregno = 155411) is only associated with one activity (doc_id = 17066, assay_id = 3664, activity_id = 82690) under document CHEMBL1145819. It does appear that the compound is a fragment of the actual structure .
The simplest solution is to reassign the molregno for the single activity from CHEMBL98511 (molregno 155411) to CHEMBL431298 (molregno = 38445):
UPDATE chembl.activities SET molregno = 38499, potential_duplicate = 1, original_activity_id = 241135 WHERE doc_id = 17066 AND assay_id = 3664 AND activity_id = 82690 AND molregno = 155411
Given that no other activities are associated with the structure CHEMBL98511 this could also be downgraded:
UPDATE chembl.molecule_dictionary SET downgraded = 1, downgrade_reason = 'Incorrect structure' WHERE molregno = 155411 AND chembl_id = 'CHEMBL98511'
Evert-Homan
Emma-Manners
CHEMBL98511 has a Ki value of 5 nM (pKi 8.3) in ChEMBL_31 for HTR6. Based on the literature reference CHEMBL1145819 it appears that this is based on the sentence:
'The corresponding racemic octahydropyrido[1,2-a]pyrazine (Ki = 5 nM) was shown to bind with somewhat reduced affinity.'
It seems that the reported data hase been associated with octahydropyrido[1,2-a]pyrazine, which is a fragment of the actual structure referred to in the paper. Imo CHEMBL98511 should be removed from ChEMBL.
Best wishes/Evert