NECo

NECo generates node embeddings from multiplex heterogeneous networks to integrate gene and phenotype data and prioritize disease-associated genes for studies of complex diseases such as hypertension, cancer, and diabetes.

Key Features:

• Multiplex Heterogeneous Network Integration: Handles multilayer networks containing gene-gene, gene-phenotype, phenotype-gene, and phenotype-phenotype relationships to represent molecular and phenotypic interactions.
• Node Embedding Methodology: Produces low-dimensional node embeddings using random walk with restart rankings as node sequences to capture structural context for gene function prediction and disease-gene identification.
• Multi-Omics Data Integration: Integrates diverse multi-omics datasets and addresses missing data and context-specific information to improve network-based predictive accuracy.

Scientific Applications:

• Complex disease gene prioritization: Supports identification and prioritization of genetic factors in multifactorial diseases, with applications to hypertension, cancer, and diabetes.
• Hypertension gene classification in Rattus norvegicus: Applied to genotypic and phenotypic datasets from Rattus norvegicus, achieving an AUC of 94.97% for hypertension gene classification versus 85.98% for the second-best method, and identifying novel hypertension-associated genes.

Methodology:

Constructs multiplex heterogeneous networks and applies a node embedding process that uses random walk with restart to generate embeddings from various neighborhood spaces.