PCEvo: Path-Consistent Molecular Representation via Virtual Evolutionary

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Kun, Hu, Longtao, Xiong, Yida, Yu, Jiajun, Zhang, Hongzhi, Chen, Jiameng, Cai, Xiantao, Wu, Jia, Hu, Wenbin
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914283504271360
author Li, Kun
Hu, Longtao
Xiong, Yida
Yu, Jiajun
Zhang, Hongzhi
Chen, Jiameng
Cai, Xiantao
Wu, Jia
Hu, Wenbin
author_facet Li, Kun
Hu, Longtao
Xiong, Yida
Yu, Jiajun
Zhang, Hongzhi
Chen, Jiameng
Cai, Xiantao
Wu, Jia
Hu, Wenbin
contents Molecular representation learning aims to learn vector embeddings that capture molecular structure and geometry, thereby enabling property prediction and downstream scientific applications. In many AI for science tasks, labeled data are expensive to obtain and therefore limited in availability. Under the few-shot setting, models trained with scarce supervision often learn brittle structure-property relationships, resulting in substantially higher prediction errors and reduced generalization to unseen molecules. To address this limitation, we propose PCEvo, a path-consistent representation method that learns from virtual paths through dynamic structural evolution. PCEvo enumerates multiple chemically feasible edit paths between retrieved similar molecular pairs under topological dependency constraints. It transforms the labels of the two molecules into stepwise supervision along each virtual evolutionary path. It introduces a path-consistency objective that enforces prediction invariance across alternative paths connecting the same two molecules. Comprehensive experiments on the QM9 and MoleculeNet datasets demonstrate that PCEvo substantially improves the few-shot generalization performance of baseline methods. The code is available at https://anonymous.4open.science/r/PCEvo-4BF2.
format Preprint
id arxiv_https___arxiv_org_abs_2601_19257
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle PCEvo: Path-Consistent Molecular Representation via Virtual Evolutionary
Li, Kun
Hu, Longtao
Xiong, Yida
Yu, Jiajun
Zhang, Hongzhi
Chen, Jiameng
Cai, Xiantao
Wu, Jia
Hu, Wenbin
Biomolecules
Artificial Intelligence
Machine Learning
Molecular representation learning aims to learn vector embeddings that capture molecular structure and geometry, thereby enabling property prediction and downstream scientific applications. In many AI for science tasks, labeled data are expensive to obtain and therefore limited in availability. Under the few-shot setting, models trained with scarce supervision often learn brittle structure-property relationships, resulting in substantially higher prediction errors and reduced generalization to unseen molecules. To address this limitation, we propose PCEvo, a path-consistent representation method that learns from virtual paths through dynamic structural evolution. PCEvo enumerates multiple chemically feasible edit paths between retrieved similar molecular pairs under topological dependency constraints. It transforms the labels of the two molecules into stepwise supervision along each virtual evolutionary path. It introduces a path-consistency objective that enforces prediction invariance across alternative paths connecting the same two molecules. Comprehensive experiments on the QM9 and MoleculeNet datasets demonstrate that PCEvo substantially improves the few-shot generalization performance of baseline methods. The code is available at https://anonymous.4open.science/r/PCEvo-4BF2.
title PCEvo: Path-Consistent Molecular Representation via Virtual Evolutionary
topic Biomolecules
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2601.19257