MOGONET integrates multi-omics data using graph convolutional networks allowing patient classification and biomarker identification

Abstract To fully utilize the advances in omics technologies and achieve a more comprehensive understanding of human diseases, novel computational methods are required for integrative analysis multiple types data. Here, we present multi-omics method named Multi-Omics Graph cOnvolutional NETworks (MOGONET) biomedical classification. MOGONET jointly explores omics-specific learning cross-omics correlation effective data We demonstrate that outperforms other state-of-the-art supervised approaches from different classification applications using mRNA expression data, DNA methylation microRNA Furthermore, can identify important biomarkers related to investigated problems.