AI Insight
Researchers developed HeteroGATomics, a heterogeneous graph attention network that integrates multiple types of cancer omics data (genomics, transcriptomics, etc.) to improve cancer diagnosis. The model uses a multi-agent system for feature selection across different omics modalities and combines them into a single heterogeneous graph that captures diverse biological relationships. Testing on three cancer datasets showed superior performance compared to existing methods, with successful external validation on an independent breast cancer cohort.
Why it matters
This approach addresses a critical challenge in precision medicine by better integrating high-dimensional biological data from small patient groups, which is common in clinical settings. The method's ability to identify important biomarkers while maintaining interpretability could facilitate more accurate cancer diagnosis and potentially guide personalized treatment decisions.
⚠️ Preprint – Noch nicht peer-reviewed
Dieser Artikel wurde noch nicht von unabhängigen Experten begutachtet. Die Ergebnisse sind vorläufig und sollten mit Vorsicht interpretiert werden.
-cross
Abstract: The increase in high-dimensional multiomics data demands advanced integration models to capture the complexity of human diseases. Graph-based deep learning integration models, despite their promise, struggle with small patient cohorts and high-dimensional features, often applying independent feature selection without modeling relationships among omics. Furthermore, conventional graph-based omics models focus on homogeneous graphs, lacking multiple types of nodes and edges to capture diverse structures. We introduce a Heterogeneous Graph ATtention network for omics integration (HeteroGATomics) to improve cancer diagnosis. HeteroGATomics performs multimodal feature selection through a multi-agent system, creating dedicated networks of feature and patient similarity for each omics modality. These networks are then combined into one heterogeneous graph for learning holistic omics-specific representations and integrating predictions across modalities. Experiments on three cancer multiomics datasets show that HeteroGATomics outperforms baseline methods across most evaluation metrics in cancer diagnosis. External validation on an independent breast cancer cohort further demonstrates its generalizability, where HeteroGATomics outperforms graph-based multiomics integration methods across evaluation metrics. Moreover, HeteroGATomics enhances interpretability by identifying important biomarkers contributing to the diagnosis outcomes.