Computational Biology & Machine Learning
AI Architectures, Bioinformatics Pipelines, and In-Silico Modeling.
General Research Overview: My research in this area focuses on the development and application of advanced computational frameworks to decode complex biological data. By leveraging Artificial Intelligence (AI) and Deep Learning, I build predictive models that identify bioactive molecules and disease biomarkers with high precision. My work spans from architectural innovations in neural networks to high-resolution Molecular Dynamics (MD) simulations, providing a "dry-lab" foundation for drug discovery and precision medicine.
Core Areas of Impact:
- Deep Learning & Architecture Innovation: I have contributed to the development of novel neural network models, such as SE-Res-U-Net, designed for efficient biological signal detection and state classification.
- Explainable AI (XAI) for Therapeutics: My work includes building interpretable machine learning models like StackAHTPs, which identifies antihypertensive peptides using stacked learning approaches to ensure that computational predictions are biologically meaningful.
- In Silico Drug Discovery: I utilize Molecular Docking and MD simulations to evaluate the efficacy of bioactive compounds from natural sources (e.g., Nigella sativa, Allium sativum) against critical targets like PTEN and Neuropilins in oncology.
- Viral Informatics: I have led structural and functional analyses of viral proteins, including the SARS-CoV-2 RNA-dependent RNA polymerase, to identify potential binding sites for synthetic and complementary drug candidates.
Technical Expertise:
- AI Development: Advanced proficiency in designing and training Machine Learning and Deep Learning models for protein sequencing, sleep state detection, and disease association.
- Molecular Modeling: Expert use of GROMACS and docking tools to perform structural analysis and protein-ligand interaction studies.
- Bioinformatics Pipelines: Experience in developing multi-block evolutionary information consensus sequences and heterogeneous feature extraction methods for biological identifiers.