A novel fractal approach for predicting G-protein–coupled receptors and their subfamilies with support vector machines

Abstract

Abstract. G-protein-coupled receptors (GPCRs) are seven membrane-spanning proteins and regulate many important physiological processes, such as vision, neurotransmission, immune response and so on. GPCRs-related pathways are the targets of a large number of marketed drugs. Therefore, the design of a reliable computational model for predicting GPCRs from amino acid sequence has long been a significant biomedical problem. Chaos game representation (CGR) reveals the fractal patterns hidden in protein sequences, and then fractal dimension (FD) is an important feature of these highly irregular geometries with concise mathematical expression. Here, in order to extract important features from GPCR protein sequences, CGR algorithm, fractal dimension and amino acid composition (AAC) are employed to formulate the numerical features of protein samples. Four groups of features are considered, and each group is evaluated by support vector machine (SVM) and 10-fold cross-validation test. To test the performance of the present method, a new non-redundant dataset was built based on latest GPCRDB database. Comparing the results of numerical experiments, the group of combined features with AAC and FD gets the best result, the accuracy is 99.22% and Matthew's correlation coefficient (MCC) is 0.9845 for identifying GPCRs from non-GPCRs. Moreover, if it is classified as a GPCR, it will be further put into the second level, which will classify a GPCR into one of the five main subfamilies. At this level, the group of combined features with AAC and FD also gets best accuracy 85.73%. Finally, the proposed predictor is also compared with existing methods and shows better performances.

The application of artificial intelligence to accelerate G protein‐coupled receptor drug discovery

“…AI models have been used to accurately distinguish GPCRs from non‐GPCRs and to further classify GPCRs into families, subfamilies, sub‐subfamilies and subtypes. The models use a variety of data for the input, including amino acid sequences or structural data derived from X‐ray crystallography, cryo‐electron microscopy (cryoEM) or molecular dynamics simulation experiments (Ao et al, 2020; Begum et al, 2020; Gu et al, 2020; Li et al, 2017; Liao et al, 2016; Naveed & Khan, 2012; Nie et al, 2015; Peng et al, 2010; Qiu et al, 2021; Wang et al, 2022; Zia‐Ur‐Rehman & Khan, 2012). A machine learning model, trained using the helical and loop information from X‐ray crystal structures, can distinguish active and inactive states of class A GPCRs (Bemister‐Buffington et al, 2020).…”

ALUminating the Path of Atherosclerosis Progression: Chaos Theory Suggests a Role for Alu Repeats in the Development of Atherosclerotic Vascular Disease

“…Specifically, Yu et al studied a large number of protein sequences that were derived from corresponding complete genomes, and demonstrated that these protein sequences were, in fact, not completely random in nature [ 41 ]. Further developments allowed for the prediction of novel structures of G-protein-coupled receptors (GPCRs) from amino acid sequences, despite the poor degree of homology among them [ 42 ], as well as the construction of phylogenetic trees for bacteria, through the use of protein sequences from complete genomes and CGR-based modeling [ 41 ].…”

EMCBOW-GPCR: A method for identifying G-protein coupled receptors based on word embedding and wordbooks

“…The classification algorithms mainly based on statistics and machine learning methods, including Artificial Neural Network (ANN) [12] , [13] , Random Forest(RF) [5] , [14] , intimate sorting [15] , K-Nearest Neighbor(KNN) [16] , [17] , etc. The methods of feature representation for predicting GPCRs contain amino acid composition (AAC) [16] , [18] , 400D [5] , N-gram [5] , [13] , [19] , SVM-Prot [14] , etc. Zou [14] proposed a novel method in which the GPCRs were represented by a 188D feature vectors of SVM-Prot and the synthetic minority oversampling technique (SMOTE) [20] , [21] , [22] algorithm was used to generate some new positive samples to balance the training datasets.…”