Abstract
Lung cancer remains the leading cause of cancer-related mortality worldwide, necessitating the development of advanced therapeutic strategies. Combination therapy, the administration of two or more drugs, has emerged as a critical approach to overcome drug resistance and enhance therapeutic efficacy compared to monotherapy. However, experimental screening of all possible drug combinations is cost-prohibitive and time-consuming due to the combinatorial explosion of potential pairs. In this study, we propose a novel deep learning framework that predicts drug synergy specifically for lung cancer by fusing transcriptomic data with graph neural networks. The model integrates the chemical structural information of drugs, captured through graph convolutional networks, with the genomic features of cancer cell lines derived from high-throughput transcriptomic profiles. By employing an attention-based fusion mechanism, the architecturedynamically weighs the importance of molecular substructures and gene expression signatures to predict synergy scores. We evaluated our model on large-scale benchmark datasets, demonstrating superior performance over state-of-the-art machine learning baselines. The results indicate that incorporating cell-line-specific transcriptomic data significantly improves prediction accuracy, offering a promising computational tool to accelerate the discovery of effective combination therapies for non-small cell lung cancer and small cell lung cancer subtypes.
References
Vance, E. (2026). Longitudinal Study of Public Health Interventions for Aging Populations using Causal Inference Methods. Frontiers in Healthcare Technology, 3(1), 40-47.
Ren, Y., Wu, D., Tong Ph D, Y., Lopez-De Fede Ph D, A., & Gareau, S. J. (2024). Benchmarking Machine Learning Models to Predict Low Birthweight Baby Outcomes and Identify Associated Risk Factors from an Extremely Unbalanced Large-Scale Dataset. Women's Health Issues, 34(72), 79.
Smith, J., Anderson, R., & Miller, W. (2026). Predicting Efficacy of Immune Checkpoint Inhibitors in Targeted Oncology Therapy using Multi-Modal Deep Learning. Frontiers in Healthcare Technology, 3(1), 31-39.
Higgins, J. P. (2002). Nonlinear systems in medicine. The Yale journal of biology and medicine, 75(5-6), 247.
Wang, C., Jiang, Y., Yang, Z., Xu, H., Khalid, A. K., Iftakhar, T., ... & Chen, Y. (2024). Host factor RBMX2 promotes epithelial cell apoptosis by downregulating APAF-1’s Retention Intron after Mycobacterium bovis infection. Frontiers in Immunology, 15, 1431207.
Wu, Y., Ren, Y., Wu, D., Xirasagar, S., & Johnson, J. (2022, June). Using D-dimer as a Biomarker to Predict COVID-19 Disease Severity from Clinical Data of Hospitalized Patients: A Machine Learning Approach. In 2022 IEEE 10th International Conference on Healthcare Informatics (ICHI) (pp. 664-668). IEEE.
Kwaan, M. R., Ren, Y., Wu, Y., & Xirasagar, S. (2021). Colonic stent use by indication and patient outcomes: a nationwide inpatient sample study. Journal of Surgical Research, 265, 168-179.
Meng, L., Shan, T., Li, K., & Gong, Q. (2021). Long-term tract-specific white matter microstructural changes after acute stress. Brain imaging and behavior, 15(4), 1868-1875.
Banerjee, S., Kwaan, M. R., Wu, Y., Ren, Y., & Xirasagar, S. (2023). Ostomy Surgery for Patients with Large Bowel Obstruction in the Modern Era: a Nationwide Inpatient Sample Study. Journal of Gastrointestinal Surgery, 27(3), 585-589.
Cai, L., Pfob, A., Barr, R. G., Duda, V., Alwafai, Z., Balleyguier, C., ... & Golatta, M. (2025). Deep learning model for breast shear wave elastography to improve breast cancer diagnosis (INSPiRED 006): An international, multicenter analysis. Journal of Clinical Oncology, 43(32), 3482-3493.
Saito, T., & Sasaki, K. (2026). Identification of Novel Circulating Biomarkers for Early Cancer Detection using High-Dimensional Feature Selection Algorithms. Multidisciplinary Research in Computing Information Systems, 6(1), 63-71.
Sinha, D., & Dey, D. K. (1997). Semiparametric Bayesian analysis of survival data. Journal of the American Statistical Association, 92(439), 1195-1212.
Miller, J., Evans, S., & Richardson, D. (2026). Correlation Analysis of Particulate Matter Exposure and Respiratory Function Decline using Multivariate Statistical Modeling. Frontiers in Environmental Science and Sustainability, 3(1), 10-17.
Shan, T., Lalor, E. C., & Maddox, R. K. (2026). Chimeric music reveals an interaction of pitch and time in electrophysiological signatures of music encoding. Journal of Neuroscience, 46(4).
Kwaan, M. R., Ren, Y., Wu, Y., & Xirasagar, S. (2021). Colonic stent use by indication and patient outcomes: a nationwide inpatient sample study. Journal of Surgical Research, 265, 168-179.
Hadjidimitriou, S. K., & Hadjileontiadis, L. J. (2012). Toward an EEG-based recognition of music liking using time-frequency analysis. IEEE Transactions on Biomedical Engineering, 59(12), 3498-3510.
Carrasco-Zanini, J., Pietzner, M., Davitte, J., Surendran, P., Croteau-Chonka, D. C., Robins, C., ... & Langenberg, C. (2024). Proteomic signatures improve risk prediction for common and rare diseases. Nature medicine, 30(9), 2489-2498.
Jaritz, M., Vu, T. H., Charette, R. D., Wirbel, E., & Pérez, P. (2020). xmuda: Cross-modal unsupervised domain adaptation for 3d semantic segmentation. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition (pp. 12605-12614).
Guo, J., Goodluck, H., Chen, E., Lee, B., & Tsai, L. (2025). Abstract 2826 ADCC assay elucidate cytokine release and immunoblot profile in three targeted Tumor cell lines. Journal of Biological Chemistry, 301(5).
Wang, Y. (2025, June). RAGNet: Transformer–GNN–Enhanced Cox–Logistic Hybrid Model for Rheumatoid Arthritis Risk Prediction. In Proceedings of the 2025 International Conference on Health Informatization and Data Analytics (pp. 90-94).
This work is licensed under a Creative Commons Attribution 4.0 International License.
Copyright (c) 2026 Pierre Martin, Sarah Becker, Alessandro Ferrari (Author)