English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

The identification of cancer genes is a critical yet challenging problem incancer genomics research. Existing computational methods, including deep graphneural networks, fail to exploit the multilayered gene-gene interactions orprovide limited explanation for their predictions. These methods are restrictedto a single biological network, which cannot capture the full complexity oftumorigenesis. Models trained on different biological networks often yielddifferent and even opposite cancer gene predictions, hindering theirtrustworthy adaptation. Here, we introduce an Explainable Multilayer GraphNeural Network (EMGNN) approach to identify cancer genes by leveraging multiplegenegene interaction networks and pan-cancer multi-omics data. Unlikeconventional graph learning on a single biological network, EMGNN uses amultilayered graph neural network to learn from multiple biological networksfor accurate cancer gene prediction. Our method consistently outperforms allexisting methods, with an average 7.15% improvement in area under theprecision-recall curve (AUPR) over the current state-of-the-art method.Importantly, EMGNN integrated multiple graphs to prioritize newly predictedcancer genes with conflicting predictions from single biological networks. Foreach prediction, EMGNN provided valuable biological insights via bothmodel-level feature importance explanations and molecular-level gene setenrichment analysis. Overall, EMGNN offers a powerful new paradigm of graphlearning through modeling the multilayered topological gene relationships andprovides a valuable tool for cancer genomics research.

Explainable Multilayer Graph Neural Network for Cancer Gene Prediction