Prediction of viral evolution and exploration of next-pandemic viruses

One of the major challenges in controlling viral infectious diseases is that viruses continuously alter their properties through evolution. During the COVID-19 pandemic, for example, variants with enhanced immune escape and increased fitness emerged successively, thereby making epidemic control substantially more difficult. In this seminor, I will introduce our research on understanding and predicting viral evolution and epidemic dynamics by integrating protein language models, massive viral genome sequence data, and large-scale experimental datasets to model the relationships among viral genotypes, antigenicity, and fitness.

Another major factor complicating the control of viral infectious diseases is the cross-species transmission of viruses harbored by wild animals to humans and livestock, leading to the emergence of novel infectious diseases. The COVID-19 pandemic, for instance, is thought to have originated from a coronavirus carried by horseshoe bats that subsequently spilled over into humans. To prepare for future pandemics, it is essential to comprehensively identify and systematically catalog viruses circulating in wildlife populations. In this seminar, I will also present our research on efficiently discovering novel viruses from massive public RNA-seq datasets by predicting viral infection based on host immune responses.