CCR5 receptor antagonists modeling and support vector regression in QSARs

In the first section of this dissertation, Quantitative Structure-Retention Relationship (QSAR) and Quantitative Structure-Property Relationship (QSPR) models are developed to predict Caco-2 permeability of drugs and protein retention times in anion-exchange chromatography systems. The physicochemical, topological, subdivided surface area, and Transferable Atom Equivalent (TAE) electron-density-based descriptors are computed directly from molecular structural geometries. A novel algorithm based on Support Vector Machine (SVM) regression was employed to obtain mathematical prediction models. A visualization scheme was also used to display the relative importance of each selected descriptor in the final models. Once these predictive models are validated, they can be used as an automated prediction tool for virtual high-throughput screening (VHTS).; Recently, the CCR5 chemokine receptor has been found to play a crucial role in the viral entry stage of HIV infection. This discovery has motivated intensive efforts for the development of small molecule CCR5 antagonists as a new class of anti-HIV therapeutics. In the second section of this dissertation, both ligand-based and structure-based approaches were performed to model the interaction between the CCR5 receptor and its small molecular antagonists at the atomic level. The initial three-dimensional (3D) structure of the CCR5 receptor was constructed using a homology modeling approach. Subsequently, recently published CCR5 piperidine-based antagonists were docked into potential binding site of the CCR5 receptor and the resulting complexes were refined by means of classical molecular dynamics (MD) simulation in a lipid bilayer environment. The binding modes between the receptor and antagonists were analyzed. In addition, 3D-QSAR, Comparative Molecular Field Analysis (CoMFA) and Comparative Similarity Indices Analysis (CoMSIA) studies were conducted on a series of piperidine-based CCR5 antagonists. Such computer-aided techniques were used to provide insight into the physicochemical factors crucial for CCR5 receptor-antagonist binding, which could in turn be used to guide the rational design of potential new anti-viral drugs. This procedure is also applicable to other GPCRs protein targets.