A new article was published by colleagues from University Paris Cité, Autorité de Sûreté Nucléaire et de Radioprotection, Hospital Clínic of Barcelona, and Stockholm University, working in WP4 (Effects and Risks) of RadoNorm. In their paper titled Comprehensive computational analysis via Adverse Outcome Pathways and Aggregate Exposure Pathways in exploring synergistic effects from radon and tobacco smoke on lung cancer in the journal Frontiers in Public Health, the authors developed a machine learning–assisted approach to systematically organize and analyse toxicological evidence on how combined exposure to radon and tobacco smoke contributes to lung cancer development. By integrating data from scientific literature and toxicological databases into the Adverse Outcome Pathway (AOP) and Aggregate Exposure Pathway (AEP) frameworks, they were able to link exposure patterns with the sequential biological changes leading to adverse health outcomes. This approach generated an integrated AOP/AEP network that identified the known distinct mechanisms for each exposure, while also highlighting the nodal points of interaction that can explain the sub-multiplicative nature of the combined effects. The machine learning–based method proved effective in consolidating dispersed evidence across multiple data sources, but it also revealed critical knowledge gaps, particularly the scarcity of studies at environmentally relevant radon concentrations and the limited mechanistic data on combined exposures at low doses.

Overall, the findings demonstrate that the current AOP/AEP framework, supported by computational tools, can efficiently integrate complex data on combined exposures. However, further development is required to advance these models into quantitative AOPs, which would enable more accurate prediction of exposure–response relationships and strengthen their application in risk assessment.

This and more publications can be found on the RadoNorm website.