Prediction of Plasma Membrane Cholesterol from 7-Transmembrane Receptor Using Hybrid Machine Learning Algorithm

Abstract

The researches have been made on G-protein coupled receptors (GPCRs) over the long-ago decades. GPCR is also named as 7-transmembrane (7TM) receptor. According to biological prospective GPCRs consist of large protein family with respective subfamilies and are mediated by different physiological phenomena like taste, smell, vision etc. The main functionality of these 7TM receptors is signal transduction among various cells. In human genome, cell membrane plays significant role. All cells are made up of trillion of cells and have dissimilar functionality. Cell membrane composed of different components. GPCRs are reported to be modulated by membrane cholesterol by interacting with cholesterol recognition amino acid consensus (L/V-X (1-5) -Y-X (1-5) -R/K) (CRAC) or reverse orientation of CRAC (R/K-X (1-5) -Y-X (1-5) -L/V) (CARC) motifs present in the TM helices. Among all, cholesterol is one who is regulated by membrane proteins. Here we took GPCR as membrane proteins and this protein modulates membrane cholesterol. According to cell biology, GPCR regulates a wide diversity of vital cellular processes and are targeted by a huge fraction of approved drugs. In this paper we have concentrated our investigation on membrane protein with membrane cholesterol. A hybrid algorithm consisting of spectral clustering and support vector machine is proposed for prediction of membrane cholesterol with GPCR. Spectral clustering uses graph nodes for calculating the cluster points and also it considers other concept such as similarity matrix, low-dimensional space for projecting the data points and upon this parameter at last construct the cluster centre. Supervised learning method is used for solving regression and classification problems. From the analysis we found that our result shows better prediction accuracy in terms of time complexity when compared with two existing models such as fuzzy c-means (FCM) and rough set with FCM model.