A new digital system called FluWarning has been developed to monitor the cross-species transmission of avian influenza, often referred to as bird flu. Created by a research team from the Politecnico di Milano and the University of Milan, this innovative tool detects significant genetic changes in flu viruses that could indicate a potential spillover from birds to other species, including humans and cattle.
The research, published in the journal Science Advances, forms part of the SENSIBLE project, which stands for Small-data Early warNing System for viral pathogens in public hEalth. The team includes key figures such as Anna Bernasconi, Principal Investigator, alongside Stefano Ceri and Tommaso Alfonsi from the Department of Electronics, Information and Bioengineering at the Politecnico di Milano, as well as Matteo Chiara, a professor in the Department of Biosciences at the University of Milan.
FluWarning utilizes data from GISAID, a global platform dedicated to sharing virus sequences and related metadata. This system is particularly informed by the H1N1 swine flu pandemic in 2009, a notable instance of virus transmission from animals to humans. It has also been adapted to track the H5N1 bird flu strain, which has recently begun to infect cattle in the United States.
The software employs a statistical approach to identify anomalies in viral sequences. Depending on its configuration, FluWarning can detect individual anomalies or clusters of them. It learns the typical genetic sequences of influenza viruses and issues alerts when significant changes occur. Upon receiving an alert, virologists can investigate the sequences to confirm or rule out instances of spillover.
With its user-friendly installation and customizable analysis, FluWarning is positioned to be an essential tool for laboratories and regional genomic surveillance institutions. Bernasconi emphasizes its potential, stating, “The system is perfectly operational. It can give feedback on these changes day by day.”
In the period from 2024 to 2025, two different H5N1 genotypes were associated with independent outbreaks in the United States, leading to numerous cases of bird flu among dairy cattle. FluWarning successfully identified clusters of viral activity across several states, particularly in California. A state of emergency was declared on December 18, 2024, due to the heightened risk of bird flu spreading among cattle. Notably, some alerts from FluWarning were issued prior to any official reports, showcasing its early detection capabilities.
Chiara noted that the system identified specific mutations in the hemagglutinin (HA) gene, a critical protein influencing the virus’s ability to infect host cells. By monitoring the virus’s evolution, FluWarning was able to pinpoint distinct markers related to the strains circulating in California.
Ceri concluded that FluWarning represents a significant advancement in the detection of viral changes that could threaten both animal and human health. “By making this technology widely available, we hope to strengthen global surveillance on a health issue of global significance,” he stated.
For further details, refer to the study by Tommaso Alfonsi et al., titled “Lightweight multiscale early warning system for influenza A spillovers,” published in Science Advances in 2025. The DOI for this publication is 10.1126/sciadv.adz7312.
