Artificial Intelligence and Big Data in Chemistry: From Molecular Data to Autonomous Discovery
Valentine P. Ananikov, Jun Li, Tong ZhuArtificial intelligence and big data are transforming chemistry: the field is shifting from a discipline in which observations are mainly interpreted after experiments into a domain in which data, models, and experimental research form a unified, interconnected system of scientific discovery. This Perspective defines the chemical data-to-knowledge cycle and examines molecular representation, predictive modeling, synthesis planning, drug and materials discovery, large language models, and autonomous (self-driving) laboratories. The central message is that progress will depend not so much on the nominal size of a dataset or model as on chemical meaning, data provenance, uncertainty estimation, experimental validation, and the ability to convert computational results into reproducible knowledge.