Harnessing genomic epidemiology and machine learning for enhanced poliovirus surveillance and response

The main threat to achieving the eradication of poliovirus is the spread of circulating vaccine-derived poliovirus (cVDPV) – rare event in which the oral poliovirus vaccine becomes capable of causing outbreaks of acute flaccid paralysis (AFP). Since 2016, 4934 cVDPV type 2 cases were reported across 80 outbreaks in 44 countries, especially Africa and Asia, with recent circulation in the USA and Europe. Although every polio AFP case is routinely sequenced, the use of sequences to guide public health strategies remains very limited and underexploited. Here, I will analyze over 4,000 unpublished and novel (2025-2030) genetic sequences to maximize the impact of genomic epidemiology for surveillance and response to cVDPV2 outbreaks. I will apply phylogenetic methods to: (i) reconstruct VDPV spread at different geographic scales and identify key drivers of spread; (ii) extract indicators to inform outbreak response, quantify the impact of interventions and assess the added value of wastewater surveillance; and (iii) couple genetic analysis with artificial intelligence to forecast the spread of VDPVs and inform vaccination campaigns. This will be underlined by knowledge transfer initiatives and findings will be shared regularly with the Global Polio Eradication Initiative to support the integration of genomics epidemiology into routine real-time poliovirus response.