Tango*: Constrained synthesis planning using chemically informed value functions

Published: 03 Mar 2025, Last Modified: 09 Apr 2025AI4MAT-ICLR-2025 PosterEveryoneRevisionsBibTeXCC BY 4.0
Submission Track: Full Paper
Submission Category: Automated Synthesis
Keywords: Chemical Synthesis, Machine Learning, Cheminformatics, Search Algorithm
TL;DR: We present Tango*, general purpose algorithm to address the starting-material constrained synthesis planning problem using chemically informed value functions and an existing, uni-directional search algorithm, Retro*.
Abstract:

Computer-aided synthesis planning (CASP) has made significant strides in generating retrosynthetic pathways for simple molecules in a non-constrained fashion. Recent work introduces a specialised \textit{bidirectional} search algorithm with forward and retro expansion to address the starting material-constrained synthesis problem, allowing CASP systems to provide synthesis pathways from specified starting materials, such as waste products or renewable feed-stocks. In this work, we introduce a simple guided search that allows us to solve the starting material-constrained synthesis planning problem using an existing unidirectional search algorithm, Retro*. We show that by optimising a single hyperparameter, Tango* outperforms existing methods in terms of efficiency and solve rate. We find that the Tango* cost function catalyses strong improvements for the bidirectional DESP methods. Our method also achieves lower wall clock times while proposing synthetic routes of similar length, a common metric for route quality.

AI4Mat Journal Track: Yes
Submission Number: 39
Loading