Molecular Information

Molecular information research focuses on leveraging computational methods, particularly large language models (LLMs) and graph neural networks (GNNs), to represent, analyze, and generate molecular data. Current efforts concentrate on multimodal approaches, integrating textual descriptions with various molecular representations (e.g., SMILES strings, 2D and 3D structures) to improve property prediction, retrosynthetic planning, and drug discovery. These advancements significantly impact fields like drug design and materials science by accelerating the identification and optimization of molecules with desired properties.

Papers