Computational chemistry is a unique method in the drug discovery process?? Explain Why?. In this study 109 molecules containing the isatin backbone were subjected to quantitative structure-activity relationship analysis to find the structure requirements for ligand binding. The structures were sketched and optimized in Hyperchem. The structural invariants used in this study were those obtained from whole molecular structures: by both hyperchem and dragon software (16 types of descriptors). Four chemometrics methods including MLR, FA-MLR, PCR and GA-PLS were employed to make connections between structural parameters and anticancer effects. MLR models revealed the effects of constitutional, functional, geometrical, WHIM and GETAWAY descriptors having higher impact on anticancer activity of the compounds. GA-PLS showed functional, constitutional and chemical descriptor indices to be the most significant parameters on anticancer activity. Moreover, the result of FA-MLR analysis revealed the effects of functional descriptors on the anticancer activity. A comparison between the different statistical methods employed and the results indicated that GA-PLS represented superior results and could explain and predict 81% and JPRI, 27(5): 1-22, 2019; Article no.JPRI.48076 2 78% variances in the PIC 50 data, respectively. Docking studies of these compounds were also investigated and promising results were obtained showing that some compounds were introduced as a good candidate for cancer agents.
Original Research Article