Led by Professor Yosi Shamay of the Faculty of Biomedical Engineering, the team’s breakthrough centers on a concept they call “meta-synergy.” Unlike traditional combination therapies, where two drugs work together biologically, meta-synergy involves drugs that not only complement each other biologically but also chemically assemble into nanoparticles. This powerful AI model identifies drug pairs capable of forming nanoparticles, which are highly effective in targeting cancer cells while minimizing toxicity and side effects. The system has suggested 1985 drug combinations for 70 types of cancer, including an effective pairing for head and neck cancer using the drugs Bortezomib and Cabozantinib.

The AI system, developed with contributions from Ph.D. student Dana Miron Azagury and M.Sc. student Ben Friedmann, uses text mining to gather data on drug synergies from research publications. It then predicts which combinations could form nanoparticles. These nanoparticles deliver the drugs directly to tumors, maximizing their effectiveness. The computational predictions were confirmed by live experiments, proving the model’s ability to improve treatment outcomes while reducing side effects.