Abstract:
The antimalarial activity of a series of 4-anilinoquinolines was modeled with topological and other functional descriptors using feature selection approaches combinatorial protocol in multiple linear regression (CP-MLR) and genetic algorithm (GA). Five models were identified from each approach to explain the activity of the compounds. All these models jointly shared eighteen descriptors. Among the identified descriptors, five (H-052, MATS4m, MATS7e, Mor30p and R7m) were common to both approaches. In partial least squares (PLS) analysis the eighteen descriptors have led to a significant three-component model (r2=0.731, Q2=0.688, r2t=0.676). The common descriptors from CP-MLR and GA approaches were found among the most influential ones to modulate the activity. Among the common descriptors, MATS7e indicates that non-linear and branched molecular topology is favorable for higher activity. MATS4m has also advocated in favor of branching/ non-linearity in the molecule for the activity. The atom-centered fragment H-052 argues that R’CH2-CHX-CH2R fragments (X is halogen) in the scaffold enhance the activity. Apart from these all identified descriptors contributed to the understanding of activity profile of the compounds. Also, selected descriptors from CP-MLR and GA models and the five common descriptors have led to very good predictive models (training r2>0.81; validation r2>0.81; test r2>0.75) in back-propagation artificial neural networks (BP-ANN). The study has offered direction to understand the patterns of the antimalarial activity of anilinoquinolines for exploring potential prototype compounds.
