RETRO SYNFLOW: Discrete Flow Matching for Accurate and Diverse Single-Step Retrosynthesis

TL;DR

RETRO SYNFLOW introduces a novel discrete flow-matching framework for single-step retrosynthesis, significantly improving accuracy and diversity of predicted reactions by leveraging reaction center identification and steering techniques.

Contribution

It presents a new discrete flow-matching approach with reaction center-based intermediate structures and inference steering, advancing the state-of-the-art in single-step retrosynthesis prediction.

Findings

Achieves 60.0% top-1 accuracy, outperforming previous methods by 20%.

FK-steering improves top-5 round-trip accuracy by 19%.

Maintains competitive top-k accuracy while enhancing diversity and feasibility.

Abstract

A fundamental problem in organic chemistry is identifying and predicting the series of reactions that synthesize a desired target product molecule. Due to the combinatorial nature of the chemical search space, single-step reactant prediction -- i.e. single-step retrosynthesis -- remains challenging even for existing state-of-the-art template-free generative approaches to produce an accurate yet diverse set of feasible reactions. In this paper, we model single-step retrosynthesis planning and introduce RETRO SYNFLOW (RSF) a discrete flow-matching framework that builds a Markov bridge between the prescribed target product molecule and the reactant molecule. In contrast to past approaches, RSF employs a reaction center identification step to produce intermediate structures known as synthons as a more informative source distribution for the discrete flow. To further enhance diversity and feasibility of generated samples, we employ Feynman-Kac steering with Sequential Monte Carlo based resampling to steer promising generations at inference using a new reward oracle that relies on a forward-synthesis model. Empirically, we demonstrate \nameshort achieves $60.0 \%$ top-1 accuracy, which outperforms the previous SOTA by $20 \%$. We also substantiate the benefits of steering at inference and demonstrate that FK-steering improves top-$5$ round-trip accuracy by $19 \%$ over prior template-free SOTA methods, all while preserving competitive top-$k$ accuracy results.