Gaurav Awasthi

Gaurav Awasthi is a graduate student in chemical engineering whose research leverages machine learning to design and engineer improved antibodies against globally relevant diseases. Conventional directed evolution approaches underexplore the combinatorial space of multi-mutant variants, while existing models capture only general fitness information rather than antigen-specific properties. He develops machine learning models trained on high-throughput yeast display data to learn interactions between amino acid mutations and improve the breadth and potency of therapeutic antibodies. Using MATLAB’s Statistics and Machine Learning Toolbox, Gaurav implements regularized regression for supervised Potts models and employs built-in solvers from the Global Optimization Toolbox for bound-constrained optimization. He applies Pareto search algorithms to co-optimize different antibody functional properties simultaneously. As a MathWorks Fellow, Gaurav will accelerate therapeutics development through enhanced computational antibody design and optimization workflows. His work could fundamentally reshape the landscape of antibody therapy development, offering new possibilities for diseases like HIV, malaria, and influenza.