VirusT5: Harnessing Large Language Models to Predicting SARS-CoV-2 Evolution
This work addresses the challenge of forecasting viral evolution for public health and research, introducing a novel 'mutation-as-translation' concept that could lead to new tools against viruses.
The study tackled the problem of predicting SARS-CoV-2 evolution by treating mutation as a translation task, training a transformer model called VirusT5 to capture mutation patterns, and demonstrated its feasibility in identifying hotspots and predicting future variants.
During a virus's evolution,various regions of the genome are subjected to distinct levels of functional constraints.Combined with factors like codon bias and DNA repair efficiency,these constraints contribute to unique mutation patterns within the genome or a specific gene. In this project, we harnessed the power of Large Language Models(LLMs) to predict the evolution of SARS-CoV-2. By treating the mutation process from one generation to the next as a translation task, we trained a transformer model, called VirusT5, to capture the mutation patterns underlying SARS-CoV-2 evolution. We evaluated the VirusT5's ability to detect these mutation patterns including its ability to identify mutation hotspots and explored the potential of using VirusT5 to predict future virus variants. Our findings demonstrate the feasibility of using a large language model to model viral evolution as a translation process. This study establishes the groundbreaking concept of "mutation-as-translation," paving the way for new methodologies and tools for combating virus threats