Genome-Phenome Association Prediction by Deep Factorizing Heterogeneous Molecular Network

Genome-phenome association (GPA) play a crucial part in deciphering the complex pathology of phenotypes (i.e., traits and diseases). Heterogeneous network-based GPA solutions can model the complex connections between multi-types of molecules by viewing them as nodes, and give better results than using single/two-omics data alone, but they overwhelmingly need to project other molecules toward homog...