A Sequence-Graph Fusion Framework via BiMamba and Fourier-KAN for Interpretable Drug-Target Affinity Prediction

Xibo Li, Lian Chen, Dingyuan Chen, Yichuan Zhao, Li Zhou, Dongxi Li
Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence, PMLR 337:3685-3704, 2026.

Abstract

Accurate prediction of drug-target affinity (DTA) is crucial for accelerating drug discovery. Existing methods struggle to model global long-range dependencies of sequences at linear complexity, and multi-layer perceptrons relying on fixed activation functions in graph neural networks face limitations in fitting complex non-linear topologies. We propose SGFF-DTA, a sequence-graph fusion framework based on bidirectional Mamba ({BiMamba}) and {Fourier}-{Kolmogorov}-Arnold Networks ({Fourier-KAN}). The framework employs {BiMamba} to capture global contextual semantics of sequences at linear complexity through bidirectional selective state space scanning. It integrates {Fourier-KAN} into graph neural networks to model high-order topological interactions using learnable non-linear transformations in the spectral space. To achieve cross-modal semantic alignment, we introduce pre-trained feature transfer and a cross-gated fusion module. On three benchmark datasets, SGFF-DTA significantly outperforms state-of-the-art methods in terms of mean squared error, demonstrating robust generalization capabilities under cold-start settings. Visual analysis further confirms model interpretability in accurately locating key binding sites and pharmacophores.

Cite this Paper

BibTeX
@InProceedings{pmlr-v337-li26g, title = {A Sequence-Graph Fusion Framework via {BiMamba} and {Fourier-KAN} for Interpretable Drug-Target Affinity Prediction}, author = {Li, Xibo and Chen, Lian and Chen, Dingyuan and Zhao, Yichuan and Zhou, Li and Li, Dongxi}, booktitle = {Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence}, pages = {3685--3704}, year = {2026}, editor = {Perković, Emilija and Malinsky, Daniel}, volume = {337}, series = {Proceedings of Machine Learning Research}, month = {17--21 Aug}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v337/main/assets/li26g/li26g.pdf}, url = {https://proceedings.mlr.press/v337/li26g.html}, abstract = {Accurate prediction of drug-target affinity (DTA) is crucial for accelerating drug discovery. Existing methods struggle to model global long-range dependencies of sequences at linear complexity, and multi-layer perceptrons relying on fixed activation functions in graph neural networks face limitations in fitting complex non-linear topologies. We propose SGFF-DTA, a sequence-graph fusion framework based on bidirectional Mamba ({BiMamba}) and {Fourier}-{Kolmogorov}-Arnold Networks ({Fourier-KAN}). The framework employs {BiMamba} to capture global contextual semantics of sequences at linear complexity through bidirectional selective state space scanning. It integrates {Fourier-KAN} into graph neural networks to model high-order topological interactions using learnable non-linear transformations in the spectral space. To achieve cross-modal semantic alignment, we introduce pre-trained feature transfer and a cross-gated fusion module. On three benchmark datasets, SGFF-DTA significantly outperforms state-of-the-art methods in terms of mean squared error, demonstrating robust generalization capabilities under cold-start settings. Visual analysis further confirms model interpretability in accurately locating key binding sites and pharmacophores.} }
Endnote
%0 Conference Paper %T A Sequence-Graph Fusion Framework via BiMamba and Fourier-KAN for Interpretable Drug-Target Affinity Prediction %A Xibo Li %A Lian Chen %A Dingyuan Chen %A Yichuan Zhao %A Li Zhou %A Dongxi Li %B Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2026 %E Emilija Perković %E Daniel Malinsky %F pmlr-v337-li26g %I PMLR %P 3685--3704 %U https://proceedings.mlr.press/v337/li26g.html %V 337 %X Accurate prediction of drug-target affinity (DTA) is crucial for accelerating drug discovery. Existing methods struggle to model global long-range dependencies of sequences at linear complexity, and multi-layer perceptrons relying on fixed activation functions in graph neural networks face limitations in fitting complex non-linear topologies. We propose SGFF-DTA, a sequence-graph fusion framework based on bidirectional Mamba ({BiMamba}) and {Fourier}-{Kolmogorov}-Arnold Networks ({Fourier-KAN}). The framework employs {BiMamba} to capture global contextual semantics of sequences at linear complexity through bidirectional selective state space scanning. It integrates {Fourier-KAN} into graph neural networks to model high-order topological interactions using learnable non-linear transformations in the spectral space. To achieve cross-modal semantic alignment, we introduce pre-trained feature transfer and a cross-gated fusion module. On three benchmark datasets, SGFF-DTA significantly outperforms state-of-the-art methods in terms of mean squared error, demonstrating robust generalization capabilities under cold-start settings. Visual analysis further confirms model interpretability in accurately locating key binding sites and pharmacophores.
APA
Li, X., Chen, L., Chen, D., Zhao, Y., Zhou, L. & Li, D.. (2026). A Sequence-Graph Fusion Framework via BiMamba and Fourier-KAN for Interpretable Drug-Target Affinity Prediction. Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research 337:3685-3704 Available from https://proceedings.mlr.press/v337/li26g.html.

Related Material