MIT’s VaxSeer: AI Revolutionizes Flu Vaccine Selection

August 28, 2025

MIT’s CSAIL and Jameel Clinic have launched VaxSeer, an AI tool transforming flu vaccine strain selection. Using advanced machine learning, VaxSeer predicts dominant flu strains and identifies top vaccine candidates, enhancing flu season readiness.

Overcoming Flu Challenges

The flu’s rapid mutations complicate vaccine selection, and mismatches reduce effectiveness. Unlike traditional methods, VaxSeer uses deep learning, trained on decades of viral data, to predict flu evolution accurately.

How VaxSeer Works

VaxSeer’s dual engines:

1. Dominance Predictor: Forecasts dominant strains using protein models and equations. 

2. Antigenicity Predictor: Evaluates vaccine effectiveness via antibody response simulations. 

These produce a coverage score for vaccine-strain matching.

Proven Success

Over 10 years, VaxSeer outperformed WHO’s A/H3N2 picks in 9 of 10 seasons and matched or beat A/H1N1 in 6. Its 2016 prediction preempted WHO’s choice by a year, aligning with CDC data.

Future Impact

Focused on flu’s hemagglutinin protein, VaxSeer could expand to other pathogens, offering cost-effective strain screening to reduce flu cases and boost global health.

Read more in Nature Medicine here.