Abstract:Drug combination therapy shows promise in cancer treatment by addressing drug resistance, reducing toxicity, and enhancing therapeutic efficacy. However, the intricate and dynamic nature of biological systems makes identifying potential synergistic drugs a costly and time-consuming endeavor. To facilitate the development of combination therapy, techniques employing artificial intelligence have emerged as a transformative solution, providing a sophisticated avenue for advancing existing therapeutic approaches. In this study, we developed SynerGNet, a graph neural network model designed to accurately predict the synergistic effect of drug pairs against cancer cell lines. SynerGNet utilizes cancer-specific featured graphs created by integrating heterogeneous biological features into the human protein–protein interaction network, followed by a reduction process to enhance topological diversity. Leveraging synergy data provided by AZ-DREAM Challenges, the model yields a balanced accuracy of 0.68, significantly outperforming traditional machine learning. Encouragingly, augmenting the training data with carefully constructed synthetic instances improved the balanced accuracy of SynerGNet to 0.73. Finally, the results of an independent validation conducted against DrugCombDB demonstrated that it exhibits a strong performance when applied to unseen data. SynerGNet shows a great potential in detecting drug synergy, positioning itself as a valuable tool that could contribute to the advancement of combination therapy for cancer treatment.

What problem does this paper attempt to address?

The paper attempts to address the problem of how to effectively predict the synergistic effects of anticancer drug combinations. Specifically, the authors developed a model based on Graph Neural Network (GNN) called SynerGNet, aiming to accurately predict the synergistic effects of drug pairs in combating cancer cell lines. ### Background and Challenges - **Complexity**: The mechanisms of cancer cells are complex, involving abnormalities in multiple molecular pathways, making treatment difficult. - **Heterogeneity**: Tumors exhibit high heterogeneity between different patients and within the same tumor, making it very challenging to develop universally effective treatments. - **Drug Combination Therapy**: Drug combination therapy, which uses multiple anticancer drugs simultaneously, can surpass the effects of single therapies by synergistically or additively modulating different key pathways, improving treatment efficacy, and reducing side effects and toxicity. - **Limitations of Experimental Methods**: Traditional experimental methods (such as high-throughput screening), although effective, are costly and time-consuming, making it difficult to comprehensively explore all possible drug combinations. ### Solution - **SynerGNet Model**: The authors proposed SynerGNet, a model based on Graph Neural Network, to predict the synergistic effects of drug combinations. - **Data Source**: The model was trained using synergy data provided by the AZ-DREAM Challenges. - **Feature Graph Construction**: The model constructed a cancer-specific feature graph by integrating heterogeneous biological features into the human protein-protein interaction network. - **Data Augmentation**: The model's predictive performance was further enhanced through data augmentation techniques using synthetic instances. ### Main Contributions - **Prediction Accuracy**: SynerGNet achieved a balanced accuracy of 0.68, significantly outperforming traditional machine learning methods. - **Data Augmentation**: With data augmentation techniques, the model's balanced accuracy further improved to 0.73. - **Independent Validation**: Independent validation results on the DrugCombDB dataset showed that the model also exhibited strong performance on unseen data. ### Conclusion SynerGNet demonstrates great potential in detecting drug synergy effects and is expected to become an important tool in advancing cancer combination therapy development. Its robust performance provides a new research direction for the effectiveness of complex cancer treatments.

SynerGNet: A Graph Neural Network Model to Predict Anticancer Drug Synergy

GraphSynergy: a network-inspired deep learning model for anticancer drug combination prediction

Unlocking the therapeutic potential of drug combinations through synergy prediction using graph transformer networks

AttenSyn: An Attention-Based Deep Graph Neural Network for Anticancer Synergistic Drug Combination Prediction

A Deep Neural Network for Predicting Synergistic Drug Combinations on Cancer

SynergyX: a multi-modality mutual attention network for interpretable drug synergy prediction

Predicting drug synergy using a network propagation inspired machine learning framework

Trustworthy Deep Neural Network for Inferring Anticancer Synergistic Combinations

An Edge-based Graph Transformer Network for Anti-Cancer Drug Synergy Prediction

Deep graph embedding for prioritizing synergistic anticancer drug combinations

DDoS: A Graph Neural Network based Drug Synergy Prediction Algorithm

DeepSynergy: predicting anti-cancer drug synergy with Deep Learning

Prediction of Synergistic Anti-Cancer Drug Combinations Based on Drug Target Network and Drug Induced Gene Expression Profiles

Multi-way relation-enhanced hypergraph representation learning for anti-cancer drug synergy prediction

Synergistic Drug Combination Prediction by Integrating Multi-omics Data in Deep Learning Models

DKPE-GraphSYN: a drug synergy prediction model based on joint dual kernel density estimation and positional encoding for graph representation

Combining genomic and network characteristics for extended capability in predicting synergistic drugs for cancer

DTSyn: a Dual-Transformer-based Neural Network to Predict Synergistic Drug Combinations

CCSynergy: an integrative deep-learning framework enabling context-aware prediction of anti-cancer drug synergy

KGANSynergy: knowledge graph attention network for drug synergy prediction

HANSynergy: Heterogeneous Graph Attention Network for Drug Synergy Prediction