Key facts
- Artificial intelligence is being used to develop predictive models for disease outbreaks and viral evolution.
- Researchers at the University of Florida developed an algorithm to predict which COVID-19 variant is most likely to become dominant.
- The EVEscape AI tool, developed by Harvard Medical School and the University of Oxford, can predict viral variants likely to occur.
- AI can analyze open-source data for early warning systems to detect infectious disease events.
- The US CDC launched Insight Net to enhance infectious disease outbreak analytics using AI and machine learning.
- The WHO maintains a list of over 30 pathogens with pandemic potential.
Scientists are increasingly turning to artificial intelligence to enhance global health security by predicting and potentially preventing future pandemics. AI-driven modeling and predictive analytics offer the potential to identify emerging threats before they escalate, moving from a reactive containment strategy to a proactive prevention approach.
Researchers at the University of Florida’s Emerging Pathogens Institute have developed an algorithm capable of forecasting which circulating COVID-19 variant is most likely to become dominant within the next three months. This tool successfully identified 11 out of 11 variants up to 10 weeks before official labeling by the CDC, by training on publicly available genetic sequences of SARS-CoV-2.
Lessons learned from the COVID-19 pandemic have underscored the importance of agile tools for predicting and controlling disease spread. The World Health Organization (WHO) has updated its list of priority pathogens that could spark the next pandemic to include over 30 microorganisms based on transmissibility, virulence, and limited treatment options.
In response, public health organizations are integrating AI into their surveillance programs. The US Centers for Disease Control and Prevention (CDC) launched Insight Net in 2023, a network combining machine learning and AI to improve infectious disease outbreak analytics. Similarly, the WHO Hub for Pandemic and Epidemic Intelligence is working on implementing AI in its surveillance efforts.
Researchers from Harvard Medical School and the University of Oxford have developed an AI tool named EVEscape. This tool builds on existing generative models to predict viral protein mutations that do not interfere with the virus's function, incorporating biological and structural details to forecast likely variants. Studies have shown EVEscape to be as accurate as experimental methods in anticipating SARS-CoV-2 variations and generalizable to other viruses like influenza and HIV.
Beyond predicting viral evolution, AI is also proving valuable in disease surveillance, particularly in disaster contexts. AI algorithms can analyze open-source data at high speeds to detect signals of infectious disease events, enhancing early-warning systems.