Particle swarm optimization and genetic algorithm as feature selection techniques for the QSAR modeling of imidazo[1,5-a]pyrido[3,2-e]pyrazines, inhibitors of phosphodiesterase 10A.

摘要: Quantitative structure-activity relationship (QSAR) modeling was performed for imidazo[1,5-a]pyrido[3,2-e]pyrazines, which constitute a class of phosphodiesterase 10A inhibitors. Particle swarm optimization (PSO) and genetic algorithm (GA) were used as feature selection techniques to find the most reliable molecular descriptors from large pool. Modeling between selected pIC50 activity data achieved by linear [multiple regression (MLR)] non-linear [locally weighted (LWR) based on both Euclidean (E) Mahalanobis (M) distances] methods. In addition, stepwise MLR model built using only limited number quantum chemical descriptors, because their correlation with . The not found interesting. It concluded that LWR model, distance, applied PSO has best prediction ability. However, some other models behaved similarly. root-mean-squared errors (RMSEP) test sets obtained PSO/MLR, GA/MLR, PSO/LWRE, PSO/LWRM, GA/LWRE, GA/LWRM 0.333, 0.394, 0.313, 0.421, 0.424, respectively. PSO-selected resulted in models, non-linear.