Addressing Long-Tailed Drug–Drug Interactions Through Minimisation of Predictive Uncertainty and Loss Sharpness

Chao Liu , Xizhao Wang , Farhad Pourpanah , Sam Kwong

CAAI Transactions on Intelligence Technology ›› 2026, Vol. 11 ›› Issue (5) : 1613 -1625.

PDF (1374KB)
CAAI Transactions on Intelligence Technology ›› 2026, Vol. 11 ›› Issue (5) :1613 -1625. DOI: 10.1049/cit2.70166
ORIGINAL RESEARCH
research-article
Addressing Long-Tailed Drug–Drug Interactions Through Minimisation of Predictive Uncertainty and Loss Sharpness
Author information +
History +
PDF (1374KB)

Abstract

Drug–Drug Interaction (DDI) prediction is critical for ensuring patient safety, particularly under long-tailed distributions, where frequent (head) interactions dominate whereas rare (tail) interactions remain underrepresented. Conventional loss functions such as cross-entropy often tend to overfit head classes while they underperform on rare but clinically important classes. Recent studies have shown that sharpness-aware minimisation (SAM) improves generalisation by encouraging solutions that lie in flatter regions of the loss landscape. Given that the loss function plays a fundamental role in model training and that uncertainty estimation is crucial for handling ambiguous or difficult samples, we propose the uncertainty and SAM (USAM) loss for improving DDI prediction under long-tail scenarios. The USAM loss integrates predictive uncertainty with SAM via two key components: (i) a dynamic reweighting strategy based on uncertainty, which leverages prediction entropy to emphasise difficult samples and (ii) a class-balanced SAM regularisation term, which encourages flatter minima and enhances generalisation across imbalanced classes. Extensive experiments on four long-tailed DDI benchmarks demonstrate that the USAM loss consistently outperforms existing methods in terms of both F1 score and recall. These results highlight the effectiveness of incorporating uncertainty modelling with sharpness-aware optimisation in addressing long-tailed problems.

Keywords

drug–drug interaction prediction / dynamic reweighting / long-tailed distribution / predictive uncertainty / sharpness-aware minimisation

Cite this article

Download citation ▾
Chao Liu, Xizhao Wang, Farhad Pourpanah, Sam Kwong. Addressing Long-Tailed Drug–Drug Interactions Through Minimisation of Predictive Uncertainty and Loss Sharpness. CAAI Transactions on Intelligence Technology, 2026, 11 (5) : 1613-1625 DOI:10.1049/cit2.70166

登录浏览全文

4963

注册一个新账户 忘记密码

References

[1]

Z. Du, L. Yang, H. He, X. Wu, X. Qi, and Y. Zhang, “ Artificial Intelligence for the Noninvasive Diagnosis of Clinically Significant Portal Hypertension,” EngMedicine 2, no. 2 (2025): 100069, https://doi.org/10.1016/j.engmed.2025.100069.

[2]

Y. Xia, A. Xiong, Z. Zhang, Q. Zou, and F. Cui, “ A Comprehensive Review of Deep Learning-Based Approaches for Drug-Drug Interaction Prediction,” Briefings in Functional Genomics (2025).

[3]

S. Chen, T. Li, L. Yang, et al., “ Artificial Intelligence-Driven Prediction of Multiple Drug Interactions,” Briefings in Bioinformatics 23, no. 6 (2022): bbac427, https://doi.org/10.1093/bib/bbac427.

[4]

L. Yue, D. Tian, W. Chen, X. Han, and M. Yin, “ Deep Learning for Heterogeneous Medical Data Analysis,” World Wide Web 23, no. 5 (2020): 2715-2737, https://doi.org/10.1007/s11280-019-00764-z.

[5]

L. N. Zheng, C. G. Dong, W. E. Zhang, X. Chen, L. Yue, and W. Chen, “ Devil in the Tail: A Multi-Modal Framework for Drug-Drug Interaction Prediction in Long Tail Distinction,” in Proceedings of the 33rd ACM International Conference on Information and Knowledge Management, (Association for Computing Machinery, 2024), 3395-3404.

[6]

S. Lin, W. Chen, G. Chen, S. Zhou, D. Wei, and Y. Xiong, “ MDDI-SCL: Predicting Multi-Type Drug-Drug Interactions via Supervised Contrastive Learning,” Journal of Cheminformatics 14, no. 1 (2022): 81, https://doi.org/10.1186/s13321-022-00659-8.

[7]

S. Lin, Y. Wang, L. Zhang, et al., “ MDF-SA-DDI: Predicting Drug-Drug Interaction Events Based on Multi-Source Drug Fusion, Multi-Source Feature Fusion and Transformer Self-Attention Mechanism,” Briefings in Bioinformatics 23, no. 1 (2022): bbab421, https://doi.org/10.1093/bib/bbab421.

[8]

Y. Zhang, B. Kang, B. Hooi, S. Yan, and J. Feng, “ Deep long-tailed Learning: A Survey,” IEEE Transactions on Pattern Analysis and Machine Intelligence 45, no. 9 (2023): 10795-10816, https://doi.org/10.1109/tpami.2023.3268118.

[9]

D. Wood, T. Papamarkou, M. Benatan, and R. Allmendinger, “ Model-Agnostic Variable Importance for Predictive Uncertainty: An Entropy-Based Approach,” Data Mining and Knowledge Discovery 38, no. 6 (2024): 4184-4216, https://doi.org/10.1007/s10618-024-01070-7.

[10]

Y. Cui, M. Jia, T. Lin, Y. Song, and S. Belongie, “ Class-Balanced Loss Based on Effective Number of Samples,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, (IEEE, 2019), 9268-9277.

[11]

T. Lin, P. Goyal, R. Girshick, K. He, and P. Dollár, “ Focal Loss for Dense Object Detection,” in Proceedings of the IEEE International Conference on Computer Vision, (IEEE, 2017), 2980-2988.

[12]

P. Foret, A. Kleiner, H. Mobahi, and B. Neyshabur, “ Sharpness-Aware Minimization for Efficiently Improving Generalisation,” in Proceedings of the International Conference on Learning Representations, (OpenReview.net, 2021).

[13]

Y. Zhou, Y. Qu, X. Xu, and H. Shen, “ ImbSAM: A Closer Look at Sharpness-Aware Minimization in Class-Imbalanced Recognition,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (IEEE, 2023): 11345-11355.

[14]

Y. Deng, X. Xu, Y. Qiu, J. Xia, W. Zhang, and S. Liu, “ A Multi-modal Deep Learning Framework for Predicting Drug-Drug Interaction Events,” Bioinformatics 36, no. 15 (2020): 4316-4322, https://doi.org/10.1093/bioinformatics/btaa501.

[15]

A. Vaswani, N. Shazeer, N. Parmar, et al., “ Attention Is all You Need,” in Proceedings of the Advances in Neural Information Processing Systems, (Curran Associates, Inc., 2017).

[16]

J. Zhu, C. Che, H. Jiang, J. Xu, J. Yin, and Z. Zhong, “ SSF-DDI: A Deep Learning Method Utilising Drug Sequence and Substructure Features for Drug-Drug Interaction Prediction,” BMC Bioinformatics 25, no. 1 (2024): 39, https://doi.org/10.1186/s12859-024-05654-4.

[17]

D. Niu, L. Xu, S. Pan, L. Xia, and Z. Li, “ SRR-DDI: A Drug-Drug Interaction Prediction Model With Substructure Refined Representation Learning Based on Self-Attention Mechanism,” Knowledge-Based Systems 285 (2024): 111337, https://doi.org/10.1016/j.knosys.2023.111337.

[18]

X. Lin, Z. Quan, Z. Wang, T. Ma, and X. Zeng, “ KGNN: Knowledge Graph Neural Network for Drug-Drug Interaction Prediction,” in Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence, (2020), 2739-2745.

[19]

Y. Hong, P. Luo, S. Jin, and X. Liu, “ LaGAT: Link-Aware Graph Attention Network for Drug-Drug Interaction Prediction,” Bioinformatics 38, no. 24 (2022): 5406-5412, https://doi.org/10.1093/bioinformatics/btac682.

[20]

Z. Li, S. Zhu, B. Shao, X. Zeng, T. Wang, and T. Liu, “ DSN-DDI: An Accurate and Generalised Framework for Drug-Drug Interaction Prediction by Dual-View Representation Learning,” Briefings in Bioinformatics 24, no. 1 (2023): bbac597, https://doi.org/10.1093/bib/bbac597.

[21]

T. Lyu, J. Gao, L. Tian, Z. Li, P. Zhang, and J. Zhang, “ MDNN: A Multi-modal Deep Neural Network for Predicting Drug-Drug Interaction Events,” in Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence, (2021), 3536-3542.

[22]

Y. Chen, T. Ma, X. Yang, J. Wang, B. Song, and X. Zeng, “ MUFFIN: Multi-Scale Feature Fusion for Drug-Drug Interaction Prediction,” Bioinformatics 37, no. 17 (2021): 2651-2658, https://doi.org/10.1093/bioinformatics/btab169.

[23]

M. Jiang, G. Liu, Y. Su, W. Jin, and B. Zhao, “ Hierarchical Multi-Relational Graph Representation Learning for Large-Scale Prediction of Drug-Drug Interactions,” IEEE Transactions on Big Data 11, no. 3 (2025): 961-975, https://doi.org/10.1109/tbdata.2025.3536924.

[24]

M. Jiang, G. Liu, B. Zhao, Y. Su, and W. Jin, “ Relation-Aware Graph Structure Embedding With Co-Contrastive Learning for Drug-Drug Interaction Prediction,” Neurocomputing 572 (2024): 127203, https://doi.org/10.1016/j.neucom.2023.127203.

[25]

Y. Zhong, G. Li, J. Yang, et al., “ Learning Motif-Based Graphs for Drug-Drug Interaction Prediction via Local-Global Self-Attention,” Nature Machine Intelligence 6, no. 9 (2024): 1094-1105, https://doi.org/10.1038/s42256-024-00888-6.

[26]

Y. Wang, Z. Yang, and Q. Yao, “ Accurate and Interpretable Drug-Drug Interaction Prediction Enabled by Knowledge Subgraph Learning,” Communication and Medicine 4, no. 1 (2024): 59, https://doi.org/10.1038/s43856-024-00486-y.

[27]

Z. Jiang, P. Ding, C. Shen, and X. Dai, “ Geometric Molecular Graph Representation Learning Model for Drug-Drug Interactions Prediction,” IEEE Journal of Biomedical and Health Informatics 28, no. 12 (2024): 7623-7632, https://doi.org/10.1109/jbhi.2024.3453956.

[28]

Y. Chen, J. Wang, Q. Zou, et al., “ DrugDAGT: A Dual-Attention Graph Transformer With Contrastive Learning Improves Drug-Drug Interaction Prediction,” BMC Biology 22, no. 1 (2024): 233, https://doi.org/10.1186/s12915-024-02030-9.

[29]

H. Du, Q. Yao, J. Zhang, Y. Liu, and Z. Wang, “ Customised Subgraph Selection and Encoding for Drug-Drug Interaction Prediction,” in Proceedings of the Advances in Neural Information Processing Systems, (Curran Associates, Inc., 2024), 109582-109608.

[30]

Y. Gan, W. Liu, G. Xu, C. Yan, and G. Zou, “ DMFDDI: Deep Multi-modal Fusion for Drug-Drug Interaction Prediction,” Briefings in Bioinformatics 24, no. 6 (2023): bbad397, https://doi.org/10.1093/bib/bbad397.

[31]

L. Meng, Y. He, C. Sun, L. Huang, T. Hu, and F. Yang, “ Learning Personalised Drug Features and Differentiated Drug-Pair Interaction Information for Drug-Drug Interaction Prediction,” Neural Networks 181 (2025): 106828, https://doi.org/10.1016/j.neunet.2024.106828.

[32]

J. Ren, C. Yu, S. Sheng, et al., “ Balanced Meta-Softmax for Long-Tailed Visual Recognition,” in Proceedings of the Advances in Neural Information Processing Systems, (Curran Associates, Inc., 2020), 4175-4186.

[33]

K. Cao, C. Wei, A. Gaidon, N. Arechiga, and T. Ma, “ Learning Imbalanced Datasets With Label-Distribution-Aware Margin Loss,” in Proceedings of the Advances in Neural Information Processing Systems, (Curran Associates, Inc., 2019).

[34]

M. Li, Y. Cheung, and Y. Lu, “ Long-Tailed Visual Recognition via Gaussian Clouded Logit Adjustment,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, (IEEE, 2022), 6929-6938.

[35]

J. Chen and B. Su, “ Transfer Knowledge From Head to Tail: Uncertainty Calibration Under Long-Tailed Distribution,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, (IEEE, 2023), 19978-19987.

[36]

M. Li, Z. Hu, Y. Lu, W. Lan, Y. Cheung, and H. Huang, “ Feature Fusion From Head to Tail for Long-Tailed Visual Recognition,” in Proceedings of the AAAI Conference on Artificial Intelligence, (AAAI, 2024), 13581-13589, https://doi.org/10.1609/aaai.v38i12.29262.

[37]

Y. Jin, M. Li, Y. Lu, Y. Cheung, and H. Wang, “ Long-Tailed Visual Recognition via Self-Heterogeneous Integration With Knowledge Excavation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), (IEEE, 2023), 23695-23704.

[38]

S. Park, Y. Hong, B. Heo, S. Yun, and J. Choi, “ The Majority Can Help the Minority: Context-Rich Minority Oversampling for Long-Tailed Classification,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, (IEEE, 2022), 6887-6896.

[39]

J. Gao, H. Zhao, Z. Li, and D. Guo, “ Enhancing Minority Classes by Mixing: An Adaptative Optimal Transport Approach for Long-Tailed Classification,” in Proceedings of the Advances in Neural Information Processing Systems, (Curran Associates, Inc., 2023), 60329-60348.

[40]

J. Shi, T. Wei, Z. Zhou, J. Shao, X. Han, and Y. Li, “ Long-Tail Learning With Foundation Model: Heavy Fine-Tuning Hurts,” in Proceedings of the 41st International Conference on Machine Learning, (PMLR, 2024), 45014-45039.

[41]

M. Li, Y. Liu, Y. Lu, Y. Zhang, Y. Cheung, and H. Huang, “ Improving Visual Prompt Tuning by Gaussian Neighborhood Minimization for Long-Tailed Visual Recognition,” in Proceedings of the Advances in Neural Information Processing Systems, (Curran Associates, Inc., 2024), 103985-104009.

[42]

B. Dong, P. Zhou, S. Yan, and W. Zuo, “ LPT: Long-Tailed Prompt Tuning for Image Classification,” in Proceedings of the the Eleventh International Conference on Learning Representations, (OpenReview.net, 2023).

[43]

Z. Zhou, L. Li, P. Zhao, P. Heng, and W. Gong, “ Class-Conditional Sharpness-Aware Minimization for Deep Long-Tailed Recognition,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, (IEEE, 2023), 3499-3509.

[44]

S. Li, Q. Xu, Z. Yang, et al., “ Focal-SAM: Focal Sharpness-Aware Minimization for Long-Tailed Classification,” in Proceedings of the Forty-second International Conference on Machine Learning, (PMLR, 2025).

[45]

X. Lyu, Q. Xu, Z. Yang, S. Lyu, and Q. Huang, “ SSE-SAM: Balancing Head and Tail Classes Gradually Through Stage-Wise Sam,” in Proceedings of the AAAI Conference on Artificial Intelligence, (AAAI, 2025), 19278-19286, https://doi.org/10.1609/aaai.v39i18.34122.

[46]

X. Wang, X. Liu, P. Huang, P. Huang, S. Hu, and H. Zhu, “ U-MedSAM: Uncertainty-Aware Medsam for Medical Image Segmentation,” in Proceedings of the Medical Image Segmentation Challenge (Springe, 2024), 206-217.

[47]

Y. Ma, J. Yang, and R. Yan, “ Sharpness-Aware Gradient Alignment for Domain Generalization With Noisy Labels in Intelligent Fault Diagnosis,” IEEE Transactions on Instrumentation and Measurement 73 (2024): 10-19, https://doi.org/10.1109/tim.2024.3413128.

PDF (1374KB)

0

Accesses

0

Citation

Detail

Sections
Recommended

/

〈 〉