R3D Prediction Model: Remote Homology Prediction using Protein Contact Map with Dimensionality Reduction

Abstract

Bioinformatics is one of a developing field that utilizations evaluation to extort information from Biological Data. Bioinformatics approaches are regularly utilized for significant activities that create expansive informational collections. Two basic tremendous scale practices that use bioinformatics are genomics and proteomics. Proteins are the far reaching, complex atoms that are essential for normal working of cells. 20% of the human body is involved proteins. Proteins are involved smaller units called amino acids, which are building squares of proteins. Protein remote homology identification and recognition are focal issues in bioinformatics. Sequence homologies are a vital source of data about proteins. In this research, the framework propose different strategy that diminishes the high dimensionality of the vector representation in remote homology detection by utilizing models that are characterized at the 3D level and consequently are very structurally and practically related. Subsequently, the 3D models are mapped from the protein primary sequence. The framework proposes to address the issue of remote homology identification by reducing 3D structure models. The new technique, called remote homology identification by the Reduction of 3D models (remote-R3D), is introduced and tested on various protein families.