Protein Structural Class and Ligand Binding Prediction Using Image Based Feature

Abstract

Proteins are the building blocks of all cells in both human and all our living creatures of the world. Most of the work in the living organism is performed by Proteins. Proteins are polymers of amino acid monomers which are biomolecules or macromolecules. The tertiary structure of protein represents the three-dimensional shape of a protein. The functions, classification and binding sites are governed by protein’s tertiary structure. If two protein structures are alike then the two proteins can be of the same kind. To detect the similarity of proteins accurately in real time is crucial in the research. In this thesis, we present an analysis on local binary pattern histogram, Wavelet transformed Local Binary Pattern Histogram, Separate Row Multiplication Matrix with Uniform Local Binary Pattern Histogram, Neighbor Block Subtraction Matrix with Uniform Local Binary Pattern Histogram and Atom Bond for protein structural class prediction. We have used them on the distance matrix of α carbons of proteins which are used as an image for feature extraction. The experiments were done on a 40 percent reduced dataset of PDB files. We have demonstrated the usefulness of this feature over a large variety of supervised machine learning algorithms. We propose the use of Random Forest as the best performing classifier on this dataset using the selected features. Secondly, Protein-Ligand binding is accountable for managing the tasks of biological receptors that helps to cure diseases and many more. So, binding prediction between protein and ligand is important for understanding a protein’s activity or to accelerate docking computations in virtual screening-based drug design. Protein-Ligand Binding Prediction requires three-dimensional tertiary structure of the target protein to be searched for ligand binding. In this paper, we’ve introduced a supervised learning algorithm for predicting Protein-Ligand Binding which is a Similarity-Based Clustering approach. Our algorithm works better than most popular and widely used machine learning algorithms. So, our work is divided into two parts, Protein Structural Class Prediction & Protein-Ligand Binding Prediction.