Quadruple Attention in Many-body Systems for Accurate Molecular Property Predictions

Jiahua Rao, Dahao Xu, Wentao Wei, Yicong Chen, Mingjun Yang, Yuedong Yang
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:51183-51200, 2025.

Abstract

While Graph Neural Networks and Transformers have shown promise in predicting molecular properties, they struggle with directly modeling complex many-body interactions. Current methods often approximate interactions like three- and four-body terms in message passing, while attention-based models, despite enabling direct atom communication, are typically limited to triplets, making higher-order interactions computationally demanding. To address the limitations, we introduce MABNet, a geometric attention framework designed to model four-body interactions by facilitating direct communication among atomic quartets. This approach bypasses the computational bottlenecks associated with traditional triplet-based attention mechanisms, allowing for the efficient handling of higher-order interactions. MABNet achieves state-of-the-art performance on benchmarks like MD22 and SPICE. These improvements underscore its capability to accurately capture intricate many-body interactions in large molecules. By unifying rigorous many-body physics with computational efficiency, MABNet advances molecular simulations for applications in drug design and materials discovery, while its extensible framework paves the way for modeling higher-order quantum effects.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-rao25a, title = {Quadruple Attention in Many-body Systems for Accurate Molecular Property Predictions}, author = {Rao, Jiahua and Xu, Dahao and Wei, Wentao and Chen, Yicong and Yang, Mingjun and Yang, Yuedong}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {51183--51200}, 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/rao25a/rao25a.pdf}, url = {https://proceedings.mlr.press/v267/rao25a.html}, abstract = {While Graph Neural Networks and Transformers have shown promise in predicting molecular properties, they struggle with directly modeling complex many-body interactions. Current methods often approximate interactions like three- and four-body terms in message passing, while attention-based models, despite enabling direct atom communication, are typically limited to triplets, making higher-order interactions computationally demanding. To address the limitations, we introduce MABNet, a geometric attention framework designed to model four-body interactions by facilitating direct communication among atomic quartets. This approach bypasses the computational bottlenecks associated with traditional triplet-based attention mechanisms, allowing for the efficient handling of higher-order interactions. MABNet achieves state-of-the-art performance on benchmarks like MD22 and SPICE. These improvements underscore its capability to accurately capture intricate many-body interactions in large molecules. By unifying rigorous many-body physics with computational efficiency, MABNet advances molecular simulations for applications in drug design and materials discovery, while its extensible framework paves the way for modeling higher-order quantum effects.} }
Endnote
%0 Conference Paper %T Quadruple Attention in Many-body Systems for Accurate Molecular Property Predictions %A Jiahua Rao %A Dahao Xu %A Wentao Wei %A Yicong Chen %A Mingjun Yang %A Yuedong Yang %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-rao25a %I PMLR %P 51183--51200 %U https://proceedings.mlr.press/v267/rao25a.html %V 267 %X While Graph Neural Networks and Transformers have shown promise in predicting molecular properties, they struggle with directly modeling complex many-body interactions. Current methods often approximate interactions like three- and four-body terms in message passing, while attention-based models, despite enabling direct atom communication, are typically limited to triplets, making higher-order interactions computationally demanding. To address the limitations, we introduce MABNet, a geometric attention framework designed to model four-body interactions by facilitating direct communication among atomic quartets. This approach bypasses the computational bottlenecks associated with traditional triplet-based attention mechanisms, allowing for the efficient handling of higher-order interactions. MABNet achieves state-of-the-art performance on benchmarks like MD22 and SPICE. These improvements underscore its capability to accurately capture intricate many-body interactions in large molecules. By unifying rigorous many-body physics with computational efficiency, MABNet advances molecular simulations for applications in drug design and materials discovery, while its extensible framework paves the way for modeling higher-order quantum effects.
APA
Rao, J., Xu, D., Wei, W., Chen, Y., Yang, M. & Yang, Y.. (2025). Quadruple Attention in Many-body Systems for Accurate Molecular Property Predictions. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:51183-51200 Available from https://proceedings.mlr.press/v267/rao25a.html.

Related Material