MODAPro: Explainable Heterogeneous Networks with Variational Graph Autoencoder for Mining Disease-Specific Functional Molecules and Pathways from Omics Data.

The rapid growth of multiomics data has revolutionized our ability to investigate disease mechanisms, yet significant challenges persist in achieving meaningful integration due to inherent data heterogeneity, characteristic sparsity patterns, and the…

Preview only. Read the full abstract at the source