ToolMol: Evolutionary Agentic Framework for Multi-objective Drug Discovery
Researchers have developed ToolMol, an evolutionary agentic framework designed to improve drug discovery using large language models. This framework combines a genetic algorithm with an LLM operator that iteratively refines potential drug candidates. ToolMol utilizes a toolbox of RDKit-backed functions for precise molecular modifications, achieving state-of-the-art results in binding affinity and Absolute Binding Free Energy scores, outperforming existing methods. AI
IMPACT This framework could accelerate the discovery of new drugs by improving the efficiency and quality of LLM-driven molecular generation.