Papers › MARS: A Motif-based Autoregressive Model for Retrosynthesis Prediction

MARS: A Motif-based Autoregressive Model for Retrosynthesis Prediction

27 Sep 2022arXiv:2209.13178archive 2025-07-28

Jiahan Liu, Chaochao Yan, Yang Yu, Chan Lu, Junzhou Huang, Le Ou-Yang, Peilin Zhao

Retrosynthesis is a major task for drug discovery. It is formulated as a graph-generating problem by many existing approaches. Specifically, these methods firstly identify the reaction center, and break target molecule accordingly to generate synthons. Reactants are generated by either adding atoms sequentially to synthon graphs or directly adding proper leaving groups. However, both two strategies suffer since adding atoms results in a long prediction sequence which increases generation difficulty, while adding leaving groups can only consider the ones in the training set which results in poor generalization. In this paper, we propose a novel end-to-end graph generation model for retrosynthesis prediction, which sequentially identifies the reaction center, generates the synthons, and adds motifs to the synthons to generate reactants. Since chemically meaningful motifs are bigger than atoms and smaller than leaving groups, our method enjoys lower prediction complexity than adding atoms and better generalization than adding leaving groups. Experiments on a benchmark dataset show that the proposed model significantly outperforms previous state-of-the-art algorithms.

PaperPDF

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

No code repository is listed for this paper in the archive or in Syntology's graph.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Drug DiscoveryGraph GenerationPredictionRetrosynthesisSingle-step retrosynthesis

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Single-step retrosynthesis USPTO-50k MARS (reaction class as prior) Top-1 accuracy 66.2 #3 of 35 Archive leaderboard report
Single-step retrosynthesis USPTO-50k MARS (reaction class as prior) Top-10 accuracy 92.9 #3 of 35 Archive leaderboard report
Single-step retrosynthesis USPTO-50k MARS (reaction class as prior) Top-3 accuracy 85.8 #3 of 35 Archive leaderboard report
Single-step retrosynthesis USPTO-50k MARS (reaction class as prior) Top-5 accuracy 90.2 #3 of 35 Archive leaderboard report
Single-step retrosynthesis USPTO-50k MARS (reaction class unknown) Top-1 accuracy 54.6 #15 of 35 Archive leaderboard report
Single-step retrosynthesis USPTO-50k MARS (reaction class unknown) Top-10 accuracy 88.5 #15 of 35 Archive leaderboard report
Single-step retrosynthesis USPTO-50k MARS (reaction class unknown) Top-3 accuracy 76.4 #15 of 35 Archive leaderboard report
Single-step retrosynthesis USPTO-50k MARS (reaction class unknown) Top-5 accuracy 83.3 #15 of 35 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections