The study of the RASMS method on the QSAR of three kinds of drugs

Abstract

Random sampling analysis on molecular surface (RASMS) method was applied to three series of drugs' quantitative structure-activity relationship that were molded by multiple linear regression. The correlation coefficients (R 2 cum ) and cross-validation correlation coefficients (Q 2 CV ) obtained by the models built were 0.834 and 0.748 for the 44 antituberculotic derivatives, 0.961 and 0.781 for the 41 HIV-1 reverse transcriptase inhibitors, and 0.824 and 0.773 for the 38 phosphodiesterase inhibitors, respectively. Satisfactory results showed that information related to biological activity can be systemically expressed by RASMS.

Molecular modeling studies of human immunodeficiency virus type 1 protease inhibitors using three‐dimensional quantitative structure‐activity relationship, virtual screening, and docking simulations

“…The protein and active peptides are usually regarded as drug targets; thus, the atoms of 20 kinds of standard natural amino acids were used as probes in the RASMS method. Eight probe atoms (Table S1) are obtained considering the difference of the activity contribution with the different atoms.…”