Title:Ligand-based Molecular Modeling of HDL Receptor SR-BI Inhibitors as Potent Anti-hyperlipidemic Agents
Volume: 21
Issue: 16
Author(s): Swati Verma*Sarvesh Paliwal
Affiliation:
- Department of Pharmacy, ITS College of Pharmacy, Muradnagar, Ghaziabad, India
- Department of Pharmacy,
Banasthali Vidyapith, 304022, Banasthali, India
Keywords:
HDL, SR-BI, QSAR, MLR, PLS, FFNN, TSAR, drug design.
Abstract:
Introduction: The High-density lipoprotein (HDL) receptor, Scavenger receptor class B,
type I (SRBI) plays a crucial role in lipoprotein metabolism, cholesterol homeostasis, and atherosclerosis.
In the present study, a quantitative structure-activity relationship study (QSAR) investigation
was conducted on a data set of 31 novel indolinyl thiazole-based inhibitors of SR-BI mediated lipid
uptake.
Methods: To build the QSAR model, Multiple linear regression analysis (MLR), partial least square
analysis (PLS), and neural analysis (NN) were performed which were further evaluated internally as
well as externally for the prediction of activity. The best QSAR model for MLR was selected with a
correlation coefficient (r2) of 0.937, cross-validation r2cv of 0.908, and a standard error (S) value of
0.253. For PLS, r2 was 0.937 and for FFNN r2 was 0.961 (for the training set). This was further
evaluated externally by a test set having r2 values 0.870 (MLR), 0.863(PLS), and 0.933(neural network)
analysis.
Results: The final model comprised hydrophobic parameters (Lipole Z component) and steric parameters
(molar refractivity and K alpha2 index).
Conclusion: All these descriptors generated comparable results which prove that the model generated
is sound and has good predictability.
