CAPTURING META-SYNERGYISTIC NANOMEDICINE WITH AUTOMATION AND AI
Yosi Shamay
Assistant Professor of Biomedical Engineering, Technion-IIT
Combination therapy is widely used in cancer medicine due to the benefits of drug synergy and the reduction of acquired resistance. To minimize emergent toxicities, nanomedicines containing drug combinations are being developed, and they have shown encouraging results. Dye-stabilized drug nanoformulations with high drug loading have shown promise based on molecular structures, but their use has been limited by relatively low long-term stability and the ability to stabilize only a small range of drugs. To address these limitations, we developed an automated system for the synthesis and selection of novel dye stabilizers that improve both the stability and compatibility of a wider array of drugs. Using spectral scanning to identify drugs with strong aggregation-induced emission (AIE) properties, we employed a liquid handling robot to automate the dye synthesis process and selection of stabilizers. Through this process, we discovered a novel stabilizer, R595, which significantly enhances nanoparticle stability. In preclinical models R595-based nanoparticles demonstrated superior stability and efficacy compared to previously published stabilizers. We then used artificial intelligence (AI) to address the complexity of multi-drug nanoparticle formulations. Using a machine learning model based on decision trees, together with the automated workflow, we mapped the self-assembly outcomes of 77 small molecule drugs and drug pairs with IR783 dye, achieving high accuracy with F1-scores of 89.3% and 87.2% in training and testing, respectively. This model enabled us to predict which drug pairs would form stable NPs, an essential step toward overcoming the unpredictability of multi-drug nanoformulations. To enhance our discovery pipeline, we performed literature text mining to identify drug pairs that exhibit both biological and chemical synergy, creating a database of 1,985 drug pairs across 70 cancers. Using this database, we developed an online tool to identify cancer-specific meta-synergistic drug combinations. Among the validated combinations, we discovered a novel pair, bortezomib-cabozantinib, which formed stable NPs with enhanced biodistribution, efficacy, and reduced toxicity in an in vivo head and neck cancer model.
In conclusion, the integration of automated synthesis with AI-driven drug pair discovery represents a powerful approach to developing stable, multi-drug nanomedicines and opens new pathways for high-complexity cancer therapies.