文摘
In this paper, a novel pseudo receptor-based QSAR method is proposed as pseudo receptor probes (PRP), including 6 pseudo-probe types defined as empty, electrostatic, steric, hydrophobic, hydrogen bond donor and receptor probes. On that basis, pseudo probe‑ligand interaction is simulated by Gaussian distance-dependent function, predictability of the constructed model is taken as fitness function, and characteristic parameters and arrangements of pseudo probes are optimized by mutational particle swarm optimization (MPSO) algorithm, thereby yielding a “profile” imitating receptor active site. Then by applying PRP, a statistical quantitative structure‑activity relationship (QSAR) model is constructed over 36 newly-discovered selective vascular endothelial growth factor-2 receptor (VEGFR-2) inhibitors of benzoxazole derivatives. Subsequently, the distribution plot of virtual probes is contoured, interpreted from points of statistics and molecular graphics. Thereby, PRP is confirmed of competence to construct robust and predictable pseudo receptor-based QSAR model which possibly provides valuable information on receptor active sites, amenable to definite physicochemical significance.