The AIRIS project aims to revolutionize biomedical research by developing an AI platform that integrates mechanistic models of disease.
Current generative AI systems often struggle with accurately modeling diseases, as they primarily focus on statistical correlations rather than the biological mechanisms that drive them. This limitation can lead to unreliable predictions, particularly for diseases with poorly understood mechanisms. The AIRIS initiative, involving 21 partners from Europe, the U.S., and Canada, aims to address this issue by creating a generative AI platform that focuses on causal and dynamical modeling in biomedical research.
The Research Institute of the McGill University Health Centre stands as the sole Canadian partner in this ambitious project, which has secured €16.9 million in funding from the European Union’s Horizon Europe Programme. Over four years, the AIRIS platform will integrate various biological and clinical data types, including medical imaging, genomic data, and patient records. This integration aims to help researchers identify unknown disease pathways and generate new scientific hypotheses.
By acting as a virtual collaborator, AIRIS will enhance the research process by harmonizing complex data and enabling the exploration of disease mechanisms. Unlike many existing AI systems that operate as opaque "black boxes," AIRIS will ensure that generated hypotheses are checked for plausibility and evidence, fostering a more transparent and reliable research environment. The project will focus on five disease areas with complex biological mechanisms, such as Pulmonary Fibrosis and Cardiovascular Disease, with findings validated through experimental testing.
This initiative represents a significant shift toward a more mechanistically grounded and trustworthy approach in biomedical discovery, aligning with European values and health priorities. By providing a robust AI tool, AIRIS aims to transform how researchers understand and treat complex diseases, ultimately improving patient outcomes.