Pharmacophore modeling and 3D-QSAR analysis of phosphoinositide 3-kinase p110alpha inhibitors.

Yiping Li,Yawen Wang,Fuqiang Zhang
DOI: https://doi.org/10.1007/s00894-010-0659-y
IF: 2.172
2010-01-01
Journal of Molecular Modeling
Abstract:Pharmacophore modeling studies were undertaken for a series of compounds belonging several groups of phosphoinositide 3-kinase (PI3K) p110alpha inhibitors: 4-morpholino-2-phenylquinazolines derivatives, pyrido[3',2':4,5]furo-[3,2-d]pyrimidine derivatives, imidazo[1,2-a]pyridine derivatives, sulfonylhydrazone substituted imidazo[1,2-a]pyridines, and LY294002. A five-point pharmacophore with three hydrogen bond acceptors (A), one hydrophobic group (H), and one aromatic ring (R) as pharmacophore features was developed. The pharmacophore hypothesis yielded a statistically significant 3D-QSAR model, with a correlation coefficient of R (2) = 0.95 for training set compounds. The model generated showed excellent predictive power, with a correlation coefficient of Q (2) = 0.88 and r (pret) (2) = 0.95 for a test set of 14 compounds. Furthermore, the structure-activity relationships of PI3K p110alpha inhibitors were elucidated and the activity differences between them discussed. Docking studies were also carried out wherein active and inactive compounds were docked into the active site of the PI3K p110alpha crystal structure to analyze PI3K p110alpha-inhibitor interactions. The results provide insights that will aid optimization of these classes of PI3K p110alpha inhibitors for better activity, and may prove helpful for further lead optimization and virtual screening.
What problem does this paper attempt to address?