2026-10-14T05:22:37Z https://api.figshare.com/v2/oai

oai:figshare.com:article/34032116 2026-09-30T12:06:55Z category_1 category_4 category_7 category_21 category_46 category_734 category_931 category_132 portal_63 item_type_6 month_year_09_2026

Speak to a Protein: An Interactive Multimodal Co-Scientist for Protein Analysis Carles Navarro (17276046) Mariona Torrens-Fontanals (25143752) Philipp Tholke (25143755) Stefan Doerr (4332823) Gianni de Fabritiis (5648713) Biophysics Biochemistry Medicine Biotechnology Immunology Biological Sciences not elsewhere classified Information Systems not elsewhere classified Infectious Diseases working mental model tightly coupling language https :// open enables hypothesis generation ai system retrieves advanced structural analysis synthesizes relevant literature live 3d scene inspecting ligand complexes turns protein analysis protein analysis building interactive multimodal co relevant proteins multimodal dialogue ligand data 3d structures protein </ unevenly accessible test ideas referencing crystal posing questions new capability grounds answers freely accessible explaining results expert coscientist conformational changes binding pockets also generates activity relationships Building a working mental model of a protein typically requires weeks of reading, cross-referencing crystal and predicted structures, and inspecting ligand complexes, an effort that is slow, unevenly accessible, and often requires specialized computational skills. We introduce <i>Speak to a Protein</i>, a new capability that turns protein analysis into an interactive, multimodal dialogue with an expert coscientist. The AI system retrieves and synthesizes relevant literature, structures, and ligand data; grounds answers in a live 3D scene; and can highlight, annotate, manipulate, and see the visualization. It also generates and runs code when needed, explaining results in both text and graphics. We demonstrate these capabilities on relevant proteins, posing questions about binding pockets, conformational changes, or structure-activity relationships to test ideas in real time. <i>Speak to a Protein</i> reduces the time from question to evidence, lowers the barrier to advanced structural analysis, and enables hypothesis generation by tightly coupling language, code, and 3D structures. <i>Speak to a Protein</i> is freely accessible at https://open.playmolecule.org. 2026-09-30T00:00:00Z Text Journal contribution 10.1021/acs.jcim.6c01738.s001 https://figshare.com/articles/journal_contribution/Speak_to_a_Protein_An_Interactive_Multimodal_Co-Scientist_for_Protein_Analysis/34032116 CC BY-NC 4.0