Improving enzyme regulatory protein classification by means of SVM-RFE feature selection

Abstract

Enzyme regulation proteins are very important due to their involvement in many biological processes that sustain life. The complexity of these proteins, the impossibility of identifying direct quantification molecular properties associated with the regulation of enzymatic activities, and their structural diversity creates the necessity for new theoretical methods that can predict the enzyme regulatory function of new proteins. The current work presents the first classification model that predicts protein enzyme regulators using the Markov mean properties. These protein descriptors encode the topological information of the amino acid into contact networks based on amino acid distances and physicochemical properties. MInD-Prot software calculated these molecular descriptors for 2415 protein chains (350 enzyme regulators) using five atom physicochemical properties (Mulliken electronegativity, Kang-Jhon polarizability, vdW area, atom contribution to P) and the protein 3D regions. The best classification models to predict enzyme regulators have been obtained with machine learning algorithms from Weka using 18 features. K* has been demonstrated to be the most accurate algorithm for this protein function classification. Wrapper Subset Evaluator and SVM-RFE approaches were used to perform a feature subset selection with the best results obtained from SVM-RFE. Classification performance employing all the available features can be reached using only the 8 most relevant features selected by SVM-RFE. Thus, the current work has demonstrated the possibility of predicting new molecular targets involved in enzyme regulation using fast theoretical algorithms.

Set of Approaches Based on Position Specific Scoring Matrix and Amino Acid Sequence for Primary Category Enzyme Classification

“…In other words, this tool compares PSSM profiles to discover related, though sometimes remote, homologous proteins or DNA. Descriptors based on PSSM have been shown to improve the prediction performance of both the structural and functional properties of proteins across a range of bioinformatics problems [27], including the prediction of protein structural classes [28], protein fold recognition [29], protein-protein interactions [30], protein subcellular localization [31], RNA-binding sites [32] and, relevant here, protein functions [33][34][35][36]. In [35], for instance, 1D descriptors taken from PSSM were classified using probabilistic neural networks (PNN), kNN, decision tree, multi-layer perceptron, and SVM.…”

Classification of mild cognitive impairment and Alzheimer’s Disease with machine-learning techniques using 1H Magnetic Resonance Spectroscopy data

“…ANNs have been widely used for different problems, such as data analysis, non-invasive diagnosis in the medical field (Ding et al, 2014;S. Resino et al, 2011), linking chemical knowledge (GomezCarracedo et al, 2007;Gómez-Carracedo et al, 2007) to protein function prediction (C. Fernandez-Lozano et al, 2013a, 2014a, 2014b or, as mentioned before, for AD classification with morphometric measures (Aguilar et al, 2013;Escudero et al, 2011). ANNs consist of many simple, interconnected computational units.…”

Classification of lung cancer using ensemble-based feature selection and machine learning methods

“…Many types of feature selection methods were proposed based on machine learning framework or information theory framework. For example, Fernandez-Lozano et al employed Support Vector Machine Recursive Feature Elimination(SVM-RFE) to classify enzyme regulatory proteins or predict transport proteins 23,24 . Li et al proposed mRMR feature selection approach to predict protein cleavage sites or protein domain 25,26 .…”