Binary Particle Swarm Optimization Based Prediction of G-Protein-Coupled Receptor Families with Feature Selection

摘要

G-protein-coupled receptors (GPCRs), the largest family of membrane protein, play an important role in production of therapeutic drugs. The functions of GPCRs are closely correlated with their families. It is crucial to develop powerful tools to predict GPCRs families. In this study, Binary particle swarm optimization (BPSO) algorithm, which has a better optimization performance on discrete binary variables than particle swarm optimization (PSO), is applied to extract effective feature for amino acids pair compositions of GPCRs protein sequence. Ensemble classifier is used as prediction engine, of which the basic classifier is the fuzzy K-nearest neighbor (FKNN). Each basic classifier is trained with different feature sets. The results obtained by jackknife test are quite encouraging, indicating that the proposed method might become a potentially useful tool for GPCR prediction, or play a complimentary role to the existing methods in the relevant areas.