Re-evaluating Retrosynthesis Algorithms with Syntheseus

Published: 28 Oct 2023, Last Modified: 07 Dec 2023NeurIPS2023-AI4Science PosterEveryoneRevisionsBibTeX
Keywords: retrosynthesis, reaction prediction, chemistry, drug design, science
TL;DR: We build a framework for evaluating retrosynthesis algorithms, and use it to correct many previously published results
Abstract: The planning of how to synthesize molecules, also known as retrosynthesis, has been a growing focus of the machine learning and chemistry communities in recent years. Despite the appearance of steady progress, we argue that imperfect benchmarks and inconsistent comparisons mask systematic shortcomings of existing techniques. To remedy this, we present a benchmarking library called syntheseus which promotes best practice by default, enabling consistent meaningful evaluation of single-step and multi-step retrosynthesis algorithms. We use syntheseus to re-evaluate a number of previous retrosynthesis algorithms, and find that the ranking of state-of-the-art models changes when evaluated carefully. We end with guidance for future works in this area.
Submission Track: Attention
Submission Number: 157