Modeling All-Atom Glycan Structures via Hierarchical Message Passing and Multi-Scale Pre-training

Minghao Xu, Jiaze Song, Keming Wu, Xiangxin Zhou, Bin Cui, Wentao Zhang
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:69675-69688, 2025.

Abstract

Understanding the various properties of glycans with machine learning has shown some preliminary promise. However, previous methods mainly focused on modeling the backbone structure of glycans as graphs of monosaccharides (i.e., sugar units), while they neglected the atomic structures underlying each monosaccharide, which are actually important indicators of glycan properties. We fill this blank by introducing the GlycanAA model for All-Atom-wise Glycan modeling. GlycanAA models a glycan as a heterogeneous graph with monosaccharide nodes representing its global backbone structure and atom nodes representing its local atomic-level structures. Based on such a graph, GlycanAA performs hierarchical message passing to capture from local atomic-level interactions to global monosaccharide-level interactions. To further enhance model capability, we pre-train GlycanAA on a high-quality unlabeled glycan dataset, deriving the PreGlycanAA model. We design a multi-scale mask prediction algorithm to endow the model about different levels of dependencies in a glycan. Extensive benchmark results show the superiority of GlycanAA over existing glycan encoders and verify the further improvements achieved by PreGlycanAA. We maintain all resources at https://github.com/kasawa1234/GlycanAA.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-xu25z, title = {Modeling All-Atom Glycan Structures via Hierarchical Message Passing and Multi-Scale Pre-training}, author = {Xu, Minghao and Song, Jiaze and Wu, Keming and Zhou, Xiangxin and Cui, Bin and Zhang, Wentao}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {69675--69688}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/xu25z/xu25z.pdf}, url = {https://proceedings.mlr.press/v267/xu25z.html}, abstract = {Understanding the various properties of glycans with machine learning has shown some preliminary promise. However, previous methods mainly focused on modeling the backbone structure of glycans as graphs of monosaccharides (i.e., sugar units), while they neglected the atomic structures underlying each monosaccharide, which are actually important indicators of glycan properties. We fill this blank by introducing the GlycanAA model for All-Atom-wise Glycan modeling. GlycanAA models a glycan as a heterogeneous graph with monosaccharide nodes representing its global backbone structure and atom nodes representing its local atomic-level structures. Based on such a graph, GlycanAA performs hierarchical message passing to capture from local atomic-level interactions to global monosaccharide-level interactions. To further enhance model capability, we pre-train GlycanAA on a high-quality unlabeled glycan dataset, deriving the PreGlycanAA model. We design a multi-scale mask prediction algorithm to endow the model about different levels of dependencies in a glycan. Extensive benchmark results show the superiority of GlycanAA over existing glycan encoders and verify the further improvements achieved by PreGlycanAA. We maintain all resources at https://github.com/kasawa1234/GlycanAA.} }
Endnote
%0 Conference Paper %T Modeling All-Atom Glycan Structures via Hierarchical Message Passing and Multi-Scale Pre-training %A Minghao Xu %A Jiaze Song %A Keming Wu %A Xiangxin Zhou %A Bin Cui %A Wentao Zhang %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-xu25z %I PMLR %P 69675--69688 %U https://proceedings.mlr.press/v267/xu25z.html %V 267 %X Understanding the various properties of glycans with machine learning has shown some preliminary promise. However, previous methods mainly focused on modeling the backbone structure of glycans as graphs of monosaccharides (i.e., sugar units), while they neglected the atomic structures underlying each monosaccharide, which are actually important indicators of glycan properties. We fill this blank by introducing the GlycanAA model for All-Atom-wise Glycan modeling. GlycanAA models a glycan as a heterogeneous graph with monosaccharide nodes representing its global backbone structure and atom nodes representing its local atomic-level structures. Based on such a graph, GlycanAA performs hierarchical message passing to capture from local atomic-level interactions to global monosaccharide-level interactions. To further enhance model capability, we pre-train GlycanAA on a high-quality unlabeled glycan dataset, deriving the PreGlycanAA model. We design a multi-scale mask prediction algorithm to endow the model about different levels of dependencies in a glycan. Extensive benchmark results show the superiority of GlycanAA over existing glycan encoders and verify the further improvements achieved by PreGlycanAA. We maintain all resources at https://github.com/kasawa1234/GlycanAA.
APA
Xu, M., Song, J., Wu, K., Zhou, X., Cui, B. & Zhang, W.. (2025). Modeling All-Atom Glycan Structures via Hierarchical Message Passing and Multi-Scale Pre-training. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:69675-69688 Available from https://proceedings.mlr.press/v267/xu25z.html.

Related Material