Assessing the Extrapolation Capability of Template-Free Retrosynthesis Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chen, Shuan, Jung, Yousung
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911790443528192
author Chen, Shuan
Jung, Yousung
author_facet Chen, Shuan
Jung, Yousung
contents Despite the acknowledged capability of template-free models in exploring unseen reaction spaces compared to template-based models for retrosynthesis prediction, their ability to venture beyond established boundaries remains relatively uncharted. In this study, we empirically assess the extrapolation capability of state-of-the-art template-free models by meticulously assembling an extensive set of out-of-distribution (OOD) reactions. Our findings demonstrate that while template-free models exhibit potential in predicting precursors with novel synthesis rules, their top-10 exact-match accuracy in OOD reactions is strikingly modest (< 1%). Furthermore, despite the capability of generating novel reactions, our investigation highlights a recurring issue where more than half of the novel reactions predicted by template-free models are chemically implausible. Consequently, we advocate for the future development of template-free models that integrate considerations of chemical feasibility when navigating unexplored regions of reaction space.
format Preprint
id arxiv_https___arxiv_org_abs_2403_03960
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Assessing the Extrapolation Capability of Template-Free Retrosynthesis Models
Chen, Shuan
Jung, Yousung
Chemical Physics
Machine Learning
Despite the acknowledged capability of template-free models in exploring unseen reaction spaces compared to template-based models for retrosynthesis prediction, their ability to venture beyond established boundaries remains relatively uncharted. In this study, we empirically assess the extrapolation capability of state-of-the-art template-free models by meticulously assembling an extensive set of out-of-distribution (OOD) reactions. Our findings demonstrate that while template-free models exhibit potential in predicting precursors with novel synthesis rules, their top-10 exact-match accuracy in OOD reactions is strikingly modest (< 1%). Furthermore, despite the capability of generating novel reactions, our investigation highlights a recurring issue where more than half of the novel reactions predicted by template-free models are chemically implausible. Consequently, we advocate for the future development of template-free models that integrate considerations of chemical feasibility when navigating unexplored regions of reaction space.
title Assessing the Extrapolation Capability of Template-Free Retrosynthesis Models
topic Chemical Physics
Machine Learning
url https://arxiv.org/abs/2403.03960