find_structures
ChatGPTSearch for molecular structures by name and get their IDs: proteins and nucleic acids in the PDB, AlphaFold predicted models, and small molecules in PubChem. Use it for requests like "hemoglobin", "insulin hexamer", "caffeine" or "human p53 AlphaFold". Pass a short English name, not a sentence: translate names from other languages and leave out words like "show", "3D" or "ball and stick". AlphaFold DB has no models for most viral proteins and for very long proteins; search kind "pdb" for those. Curated MolViewer catalogs are searched first, then RCSB, UniProt and PubChem. Returns IDs to pass to show_structure or get_structure_details, best match first: the first result is the canonical, complete structure, so prefer it unless the user asked for a specific state, variant or organism.
