COMPUTATIONAL BIOLOGY IN HOST-PATHOGEN INTERACTION PREDICTION

Abstract

 In the case of making discoveries on how to treat infections and understand their working mechanisms, the interaction between proteins of host and pathogen requires us to know how they interact with each other intricately.  The present paper provides a full computational framework to predict a host-pathogen interaction (HPI) that involves functional annotation, structural bioinformatics, and machine learning.  We used human host and key pathogen protein datasets (Salmonella enterica, SARS-CoV-2, Candida albicans) and trained known HPI datasets up to a group of classifiers, i.e., Support Vector Machines, Random Forest, and Gradient Boosting. We employed such characteristics as sequence similarity, domain-domain interaction and subcellular localization.  There were nine different computer models that made 180 high-confidence HPIs.  The consensus ensemble scoring was implemented to increase the likelihood of interactions. After that, we performed structural docking simulations with the highest-ranking pairs.  The binding free energy calculated indicated that over 65 percent of the expected complexes possessed favorable 6G (< =8 kcal/mol) values, a fact that demonstrated the feasibility of the structures.  The top contacts were demonstrated to be highly impacted with effector and immune modulator proteins based on functional analysis.  Using measurements over networks, central hub proteins that are well connected were identified and may be good targets of therapy.  All line plots, bar charts, scatter graphs, pie charts, and hybrid figures indicated that the same trends occurred in all functional annotations, docking confidence, and interaction probability.  An Integration of the comments of experts on infectious diseases ensured that the computer calculations made biological sense, and could be well-understood.  The findings indicate that computational biology is an effective method that can easily and speedily forecast HPI and may be utilised to test and apply on new illnesses in the future.