HOGPred: artificial neural network-based model for orphan GPCRs

Abstract

G protein-coupled receptors (GPCRs) are one of the largest protein families of seven-transmembrane domain receptors with 800 members in human genome which act as targets for many drugs. Human GPCRs are a broad class of membrane-bound receptors in eukaryotes and have a signature seven-transmembrane alpha-helical domain structure. They play crucial role in signalling processes in cells and regulation of basic physiological processes. In this paper, we report an artificial neural network model-based HOGPred tool for identification of orphan GPCRs, predicting the location of alpha-helical transmembrane domains, point mutations and physicochemical properties of GPCRs. In this study, ANN-based classification model has been developed for the fingerprinting of human orphan GPCRs of significance in current bioinformatics research. HOGPred scans the orphan GPCRs protein sequence or structure provided by the user against the ANN-based datasets. A cross-validation was done using experimentally determined structures of GPCRs. It has a prediction accuracy of 98.4 %.