Syntheseus (Microsoft Research, Faraday Discussions 2024)
github.com/microsoft/syntheseusOpen-source framework for end-to-end retrosynthetic planning that combines multi-step search algorithms (MCTS, retro*, and retro-fallback) with wrappers for state-of-the-art single-step reaction models behind a standardized API, enabling benchmarking and rapid prototyping of synthesis-planning pipelines; used by RetroGFN, TANGO, and SimpRetro (arXiv 2310.19796, 217+ stars, MIT License, actively maintained)
Sourced from
- Awesome AI for Science — github.com/microsoft/syntheseus
Related resources
AstraZeneca's industrial-grade retrosynthetic planning tool using MCTS to recursively decompose molecules into purchasable precursors, with multi-step route scoring and support for custom one-step models (v4.0, 2024)
AI for chemical reaction prediction and synthesis planning
Generative AI system for antibiotic discovery that searches billions of synthesizable molecules by combining molecular building blocks through real chemical reactions, experimentally validating novel compounds active against drug-resistant bacteria