MAINCE will use AI approaches to identify new and much-needed therapeutics in immunology. Clues to the effect of therapeutics through state-of-the-art imaging techniques will be linked by AI to laboratory experiments to accelerate drug development and make it more efficient.
In pharmaceutical research, the early detection of undesirable or even absent effects of therapeutics represents the greatest hurdle to successful drug development. At JGU Mainz, Prof. Dr. Czodrowski's interdisciplinary team, which includes research groups from biology, chemistry, computer science, medicine, pharmacy and physics, is training an AI to assist in identifying and planning the most promising experiments. Three data sources will be used for this purpose: 1.) high-resolution microscopic image data (from both standard cell lines and primary cells), 2.) synthesis and structural data, and 3.) scientifically collected text data.
The neurosymoblical AI applied in the project thereby combines classical machine learning and generative models with the ability to establish explanatory chains. This makes it possible to limit the number of possible outcomes in order to perform only the laboratory experiments with the greatest gain in knowledge in real terms. This will be used to develop much needed therapeutics for immunology.
In summary, multi-modal generative AI will be used to achieve the breakthrough of developing innovative immunotherapeutics starting from high-resolution cellular images, which will be directly tested experimentally for their effect and side effect.