Application of Machine Learning Based Methods for Studying Protein Protein Interactions and Pathway Modeling of Proteins Related to Cancer

Abstract

Protein-protein interactions (PPIs) regulate the cellular functions, regulating cell newlinecycle, signal transduction and metabolic pathways. These macromolecular newlineinteractions play significant role in all the biological processes. There are several newlineexperimental methods to detect PPI that have made it possible to carry out newlinestructural, interactome and protein interface studies. Despite the high throughput newlinenature of the experimental methods, there still exist many limitations which newlineaffect the validity of PPI studies. In-silico approach is cost effective, time newlineefficient and enables analysis of large heterogeneous data. However, incorrect newlinefeature selection, binding energy parameters, interface properties, protein newlineconformational study in the computational methods can hinder accurate newlinepredictions. newlineThe focus of the research work in this thesis is a complete study of protein protein interactions using existing and new computational approaches to newlinecircumvent some of the above mentioned drawbacks.