At the Advanced Biocomputing Lab, BIT Mesra, I independently built and ran a virtual-screening pipeline for Amyloid-β aggregation targets. It screened 500,000 marine bioactive compounds and shortlisted four candidates for in-vitro validation.
At CSIR-Institute of Himalayan Bioresource Technology (HiCHiCoB-BIC), I engineered a PyTorch model combining graph and transformer encoders to predict drug–target binding interactions. At the same internship, I unified five benchmark datasets into a curated pipeline and incorporated per-residue attention maps for binding-site interpretability.
