Exploring the intersection of High-Performance Computing, Molecular Dynamics, and Machine Learning to accelerate Drug Discovery.
Extensive experience using the AMBER software suite for simulating protein-ligand dynamics. Analyzing trajectories using cpptraj to calculate RMSD, RMSF, and Hydrogen Bonds to understand stability.
Developing ensemble models (XGBoost, Random Forest) to predict binding affinities. Expert in dataset curation from PDBbind and feature extraction for QSAR modeling.
Full-stack implementation of scientific tools. I bridge the gap between Python backend logic and user-friendly Web interfaces (PHP/JS) hosted on Linux HPC clusters.
A comprehensive scoring function developed to improve the accuracy of protein-ligand binding affinity predictions. Trained on a curated dataset of 11,689 complexes from PDBbind, this tool utilizes an ensemble of machine learning algorithms to outperform classical scoring functions.
An ongoing initiative to map protein-ligand interactions across specific viral targets. This project integrates molecular docking pipelines with machine learning refinement stages to filter false positives in high-throughput screening campaigns.