This project will analyze publicly available whole genome sequences of COVID-19 isolates collected in India to bioinformatically predict epitopes recognized by human immune cells. Predicting epitopes from sequence is useful in vaccine design, as they could reveal the parts of the viral proteins that elicit immune response in human immune cells. We intend to validate these epitopes experimentally in collaboration with Dr. Aravindhan Vivekanandhan at the University of Madras and other co-workers. The bioinformatics analysis would involve applying reverse vaccinology methods based on machine learning such as Vaxign-ML, IEDB on COVID-19 sequences to predict epitopes, and sequence similarity methods to study if epitopes are conserved across different isolates and whether they are non-self (i.e., do no trigger auto-immune response against our own human proteins). If we identify successful epitopes from this study and characterize their evolution in Indian vs. non-Indian sequences, we hope these results would enable design of second-generation COVID-19 vaccines and inform the effectiveness of the first-generation COVID-19 vaccines currently under development in the Indian population.