A poster on the molecular foundation model ZAO and a session with external researchers will explore the evaluation and design of molecular models for AI drug discovery
Tokyo, 26 October 2026 — SyntheticGestalt KK (headquarters: Minato-ku, Tokyo; CEO: Koki Shimada) will take part in the Chem-Bio Informatics Society (CBI) Annual Meeting 2026, held from 26 to 29 October at Tower Hall Funabori in Edogawa-ku, Tokyo. The company will present a poster on ZAO, its molecular foundation model, and host the Focused Session "Frontiers in Constrained Deep Learning: Integrating Geometry, Topology, and Physics for AI Drug Discovery".
The poster compares the hit-finding performance of ZAO with that of existing methods. ZAO has been pre-trained at scale using multiple conformations for each molecule. Starting from known actives, it can search large compound libraries for novel candidates, even when no structure of the target protein is available.
In the session, speakers will discuss how three kinds of constraint – geometric, topological and physical – can be built into models and applied to molecular design, from both theoretical and implementation perspectives.*1 Some properties of molecules are known to hold before any data are seen. Constrained deep learning builds this kind of prior knowledge into the model, which can help it generalise from limited data.
Event details
Meeting: Chem-Bio Informatics Society (CBI) Annual Meeting 2026 (theme: "Quantum Bioscience and the Future of Healthcare")
Dates: Monday 26 – Thursday 29 October 2026
Venue: Tower Hall Funabori (4-1-1 Funabori, Edogawa-ku, Tokyo)
Poster: "ZAO: A Molecular Foundation Model as the Core of a Drug Discovery Agent" (P09-32), Togen (2F); 16:00–17:00 JST on Tuesday 27 October and 17:00–18:00 JST on Wednesday 28 October
Session: "Frontiers in Constrained Deep Learning: Integrating Geometry, Topology, and Physics for AI Drug Discovery" (FS13), Room 406 (4F); 14:00–15:30 JST on Thursday 29 October
Exhibition booth: No. 16
Event website: https://cbi-society.org/taikai/taikai26/index_e.html
About SyntheticGestalt
SyntheticGestalt KK is a molecular AI company based in Tokyo. It develops and deploys molecular AI for the pharmaceutical, chemical, materials, agrochemical, cosmetics and food industries, and since its founding in 2018 has pursued a single goal: to build systems that mass-produce invention through AI. Its research projects to date have identified pharmaceutical lead compounds and lower-impact materials, and in some projects have cut the cost and duration of discovery by up to 90% compared with conventional methods.*2 For more information, visit https://www.syntheticgestalt.com.
*1 The aims of the session and abstracts of each talk are available (in Japanese) on the meeting website: https://cbi-society.org/taikai/taikai26/sessionFS.html
*2 The "up to 90%" reduction refers to results achieved in SyntheticGestalt's past research projects; the actual reduction varies by target and conditions.
Media Enquiries
SyntheticGestalt KK
CEO Office, Sou Minakami
s.minakami@syntheticgestalt.com