From Holo Pockets to Electron Density: GPT-style Drug Design with Density
Researchers have developed EDMolGPT, a new AI model for drug design that utilizes electron density maps as a conditioning signal. This approach moves beyond traditional methods that only consider empty binding pockets, instead incorporating information from filler molecules and solvents. The model can process electron density data from both computational and experimental sources, enabling it to generate molecules with 3D conformations that better reflect the binding environment. AI
IMPACT Introduces a novel conditioning signal for generative AI in drug discovery, potentially improving molecule design accuracy.