Modelling the Anticancer Activity of 4-Alkoxy Cinnamic Analogues using 3D-Descriptors and Genetic Algorithm-Multiple Linear Regression (GA-MLR) Method

Abstract

QSAR modelling as anticancer of 4-alkoxy cinnamic analogues has been done against their 3D descriptors. This modelling aims to develop the 4-alkoxy cinnamic analogues in better activity. We employed Kennard and Stone’s algorithm to split the data set into training and test set. Genetic Algorithm (GA) and Multiple Linear Regression (MLR) are used to select the influencing descriptors and to carry out statistically robust model.pIC50 = 2.7350 (±1.0653) ×RDF145u − 2.2276 (±0.9655) × RDF120m + 1.8369 (±1.1295) × E2m − 1.5611 (±0.4518)(R2training = 0.7436, R2test = 0.9812)Coefficient of Y-randomization (cR2p) as one of model validation for the selected model gave a value of 0.6569 that greater than 0.5 identifying the model is powerful and not inferred by chance. Descriptors which influencing the activity are RDF145u, RDF120m, and E2m. RDF145u is radial distribution function-145/unweighted, RDF120m is radial distribution function-120/weighted by atomic mass and E2m is 2nd component accessibility directional WHIM index, weighted by atomic mass. Selected QSAR model used to design and predict some of new 4-alkoxy cinnamic analogues. There are 10 newly designed compounds that have a better prediction activity and have no violation in the drug likeness property based on the Lipinski rule.

Keywords

• QSAR modelling
• 3D Descriptors
• 4-alkoxy cinnamic