Abstract: Owing to the viral genome epidemiological surveys conducted during the COVID-19 pandemic, virology has become one of the life science fields with the richest genome information. By utilizing this big data, is it possible to understand and even predict the patterns of viral evolution and epidemics? We have developed statistical and machine learning models to estimate and predict the fitness of viruses based on genomic data. Using these models, we have predicted future variants that are likely to spread next, identified mutations contributing to increased viral fitness, and even predicted the evolution of viruses. In this seminar, I would like to discuss the breakthroughs that such data-driven virology could bring to the fields of bioinformatics and evolutionary biology.