@article{ART003212658},
author={Jong-Hyeon Ko and Kyuri Jo},
title={Leveraging the Mamba Architecture for Multi-Class Drug-Drug Interaction Prediction},
journal={Journal of The Korea Society of Computer and Information},
issn={1598-849X},
year={2025},
volume={30},
number={6},
pages={47-54}
TY - JOUR
AU - Jong-Hyeon Ko
AU - Kyuri Jo
TI - Leveraging the Mamba Architecture for Multi-Class Drug-Drug Interaction Prediction
JO - Journal of The Korea Society of Computer and Information
PY - 2025
VL - 30
IS - 6
PB - The Korean Society Of Computer And Information
SP - 47
EP - 54
SN - 1598-849X
AB - In this study, we propose a multi-class drug–drug interaction (DDI) prediction framework based on the Mamba sequence modeling architecture. The proposed model consists of three modules—a global autoencoder (AE1), a Siamese-based encoder (AE2), and a dual encoding module using Mamba (MambaSeparate)—each of which independently captures diverse drug pair features and integrates them for final classification. A subsequent Mamba layer further enhances the representation of drug–drug interactions by enabling information fusion across modules. Notably, the Mamba architecture leverages selective state space modeling to dynamically control information flow based on input significance, offering efficient long-range dependency modeling with lower computational complexity compared to transformers. The use of intermediate embeddings from each module facilitates representation learning at multiple semantic levels. Our framework effectively captures non-linear relationships and sequential interaction patterns between drugs. Experimental results on benchmark datasets demonstrate that the proposed model consistently outperforms existing methods across Tasks 1 to 3, achieving superior accuracy and generalization performance.
KW - Drug-Drug interaction;Mamba;Deep learning;Multi-class classification;Multi-source information fusion
DO -
UR -
ER -
Jong-Hyeon Ko and Kyuri Jo. (2025). Leveraging the Mamba Architecture for Multi-Class Drug-Drug Interaction Prediction. Journal of The Korea Society of Computer and Information, 30(6), 47-54.
Jong-Hyeon Ko and Kyuri Jo. 2025, "Leveraging the Mamba Architecture for Multi-Class Drug-Drug Interaction Prediction", Journal of The Korea Society of Computer and Information, vol.30, no.6 pp.47-54.
Jong-Hyeon Ko, Kyuri Jo "Leveraging the Mamba Architecture for Multi-Class Drug-Drug Interaction Prediction" Journal of The Korea Society of Computer and Information 30.6 pp.47-54 (2025) : 47.
Jong-Hyeon Ko, Kyuri Jo. Leveraging the Mamba Architecture for Multi-Class Drug-Drug Interaction Prediction. 2025; 30(6), 47-54.
Jong-Hyeon Ko and Kyuri Jo. "Leveraging the Mamba Architecture for Multi-Class Drug-Drug Interaction Prediction" Journal of The Korea Society of Computer and Information 30, no.6 (2025) : 47-54.
Jong-Hyeon Ko; Kyuri Jo. Leveraging the Mamba Architecture for Multi-Class Drug-Drug Interaction Prediction. Journal of The Korea Society of Computer and Information, 30(6), 47-54.
Jong-Hyeon Ko; Kyuri Jo. Leveraging the Mamba Architecture for Multi-Class Drug-Drug Interaction Prediction. Journal of The Korea Society of Computer and Information. 2025; 30(6) 47-54.
Jong-Hyeon Ko, Kyuri Jo. Leveraging the Mamba Architecture for Multi-Class Drug-Drug Interaction Prediction. 2025; 30(6), 47-54.
Jong-Hyeon Ko and Kyuri Jo. "Leveraging the Mamba Architecture for Multi-Class Drug-Drug Interaction Prediction" Journal of The Korea Society of Computer and Information 30, no.6 (2025) : 47-54.
