Matthieu Vilain

This paper develops an unsupervised pipeline to predict viral fitness using sequence data and finds that protein alignment, not model size, is key for accurate prediction.

The paper analyzes SARS-CoV-2 spike evolution in Canada, finding that variant emergence, not vaccine deployment, drove episodic positive selection at the provincial level.

dna-parser encodes DNA and RNA sequences into numerical features efficiently using Rust and Python, offering fast and parallel processing.

This study shows that positive selection on influenza A surface antigens has increased over time, driven by immune escape and surveillance, not climate factors.