AI and the World's Hard Problems · Medicine & drug discovery · arXiv · 2026-10-23
Mathematical AI Predicts 40% and 57% Influenza Vaccine Effectiveness Against H1N1 and H3N2, Respectively, for the 2026--2027 U.S. Season
Clinical influenza vaccine effectiveness has often been low, so preseason forecasts are useful but uncommon. The authors compared hemagglutinin sequences from U.S. specimens with vaccine strains using two mathematical distances and fitted regression models to 26 observations across 13 seasons. For the 2026–2027 season, the topological model forecast effectiveness of 40.3% against H1N1pdm09, 56.9% against H3N2, and 47.1% against B/Victoria, while the geometric model forecast 34.0%, 56.8%, and 51.1%.