AGMI: Attention-Guided Multi-omics Integration for Drug Response Prediction with Graph Neural Networks
This work addresses a crucial challenge in precision medicine for improving cancer treatment by predicting drug responses, though it appears incremental as it builds on existing GNN and multi-omics integration approaches.
The paper tackles drug response prediction in precision medicine by proposing AGMI, an attention-guided multi-omics integration method using graph neural networks, which outperforms state-of-the-art methods by 8.3% to 34.2% on metrics across CCLE and GDSC datasets.
Accurate drug response prediction (DRP) is a crucial yet challenging task in precision medicine. This paper presents a novel Attention-Guided Multi-omics Integration (AGMI) approach for DRP, which first constructs a Multi-edge Graph (MeG) for each cell line, and then aggregates multi-omics features to predict drug response using a novel structure, called Graph edge-aware Network (GeNet). For the first time, our AGMI approach explores gene constraint based multi-omics integration for DRP with the whole-genome using GNNs. Empirical experiments on the CCLE and GDSC datasets show that our AGMI largely outperforms state-of-the-art DRP methods by 8.3%--34.2% on four metrics. Our data and code are available at https://github.com/yivan-WYYGDSG/AGMI.