Researchers at the University of Hong Kong developed a machine-learning classifier that analyzes influenza A virus genomes to predict which strains pose the highest risk of spreading between mammals. The tool identifies genetic markers that indicate when bird flu viruses can jump to mammals and potentially to humans.
Bird flu regularly infects poultry and wild birds but rarely transmits to people. When it does, the consequences prove severe. Understanding which viral mutations enable cross-species spread offers early warning before outbreaks occur. The classifier examines specific genomic sequences that facilitate mammalian infection, providing scientists with a genetic blueprint of danger.
This work matters because it shifts pandemic surveillance from reactive to proactive. Rather than waiting for bird flu cases to appear in humans, researchers can now screen circulating viruses for high-risk characteristics. The tool could guide public health monitoring and inform vaccine development priorities.
The team's next steps involve testing the classifier against newly emerging virus variants and expanding its accuracy. Integrating this technology into global surveillance systems could help health agencies detect dangerous mutations early, potentially preventing the next pandemic.