Researchers at MIT create an AI-powered tool aimed at enhancing the accuracy of flu vaccine strains selection.

In a significant breakthrough for global health, scientists at the Massachusetts Institute of Technology (MIT) have developed an AI system called VaxSeer. This innovative technology is designed to predict dominant flu strains and identify the most protective vaccine candidates, months ahead of the flu season.

VaxSeer uses deep learning models trained on decades of viral sequences and lab test results to simulate how the flu virus might evolve and how vaccines will respond. The system has shown remarkable accuracy, with its predicted coverage scores aligning closely with public health data on flu-related illnesses and medical visits prevented by vaccination.

Every year, global health experts must decide which influenza strains will be included in the next seasonal vaccine. If the selected strains match those that circulate, the vaccine will likely be highly effective. VaxSeer's predictions showed strong correlation with real-world vaccine effectiveness estimates as reported by the Centers for Disease Control and Prevention (CDC), Canada's Sentinel Practitioner Surveillance Network, and Europe's I-MOVE program.

In a 10-year retrospective study, VaxSeer's choices for A/H3N2 outperformed the World Health Organization's in nine out of 10 seasons. The system also outperformed or matched the WHO in six out of 10 seasons for A/H1N1. Notably, in one case, for the 2016 flu season, VaxSeer identified a strain that wasn't chosen by the WHO until the following year.

The implications of the predictive modeling go far beyond influenza. Potentially, VaxSeer could apply to antibiotic-resistant bacteria and drug-resistant cancers. However, applying the system to other viruses would require large, high-quality datasets that track both viral evolution and immune responses.

The challenge of predicting influenza strains became more familiar during the Covid-19 pandemic due to the emergence of new variants. VaxSeer's ability to predict dominant flu strains and identify the most protective vaccine candidates could provide valuable insights for similar challenges in the future.

The paper, supported, in part, by the U.S. Defense Threat Reduction Agency and MIT Jameel Clinic, was published today in Nature Medicine. The team is currently working on the methods that can predict viral evolution in low-data regimes building on relations between viral families.

As we continue to navigate the ever-changing landscape of global health, technologies like VaxSeer hold immense potential for improving our ability to predict and respond to emerging threats.

Researchers at MIT create an AI-powered tool aimed at enhancing the accuracy of flu vaccine strains selection.