异构图注意力网络结合多模态特征选择用于癌症多组学整合
原标题：Heterogeneous Graph Attention Network With Multimodal Feature Selection for Cancer Multiomics Integration
AI 摘要
研究团队提出 HeteroGATomics，一种面向癌症多组学整合的异构图注意力网络框架。该方法通过多智能体系统进行多模态特征选择，为每个组学模态分别构建特征相似网络与患者相似网络，再合并为单一异构图以学习整体表征并融合各模态预测。在三个癌症多组学数据集上，该方法在多数诊断评价指标上优于基线，并在独立乳腺癌队列的外部验证中优于基于图的多组学整合方法，同时通过识别关键生物标志物提升了可解释性。
正文节选
Heterogeneous Graph Attention Network With Multimodal Feature Selection for Cancer Multiomics Integration 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 grap