Prediction Of G-protein Coupled Receptors And Their Coupling Specificity

G-protein coupled receptors (GPCRs) represent one of the most important classes of drug targets for pharmaceutical industry and play important roles in cellular signal transduction. They belong to one of the largest superfamilies of membrance proteins and an important class of cell surface receptor proteins that are central to the signaling network. In the human genome, it is now assumed that at least 1% genes encode GPCRs, and only about 900 GPCRs were found so far, so there is still fairly large space of identifying new GPCRs as the candidate of drug target. Prediction of GPCRs and their coupling specificity are vital for the research and pharmaceutical industry. Using bioinformatics methods to study GPCRs will help for further understanding the mechanism of signal transduction and the function of the receptors within a cell, which can provide new clues for pharmaceutical research.Feature extraction and module construction is one of the key issues in bioinformatics applications. In this study, the features of protein sequence, such as dipeptide composition, length of sequence, and physiochemical properties of amino acid, including hydrophobicity, polarity, charge of amino acids, have been analyzed and extracted. Based on these features, classifiers were constructed for identification and functional prediction of GPCRs using the method of support vector machines (SVMs). As a result of cross-validation, GPCRs could be correctly identified with an accuracy of 97%, 99% at family classification and 90% at coupling prediction, respectively. The results were all better than previous methods. Finally, GPCRFP software was integrated using the classifiers developed by the above procedures, then used it to analyze the human genome, and some meaningful results were obtained. The GPCRFP can be available at http://moe.hgrp.cn/tools/GPCR/index.html.