Ansh Gupta

Molecule Optimizer with ECFP6 and GROVER Fingerprints: Built ML models using ECFP6 and GROVER molecular fingerprints to predict molecular activity and optimize compounds across 19 models. Implemented hyperparameter tuning and feature selection to enhance predictive accuracy for protein/gene activity using the MIMOSA and GenAI frameworks. Protein–Protein Interaction ML Models: Curated and preprocessed multi-species protein-protein interaction datasets comprising over 1M+ data points, ensuring high data quality and integrity for training robust machine learning models, and improving accuracy.