DiSNEP

DiSNEP enhances gene networks using disease omics data to produce disease-specific network representations that improve prioritization of disease-associated genes.

Key Features:

• Disease-specific network enhancement: Enhances general human gene networks such as STRING, GIANT, and HumanNet to better reflect disease-specific gene interactions.
• Diffusion-based propagation: Applies a diffusion process to propagate signals across a gene–gene similarity matrix.
• Omics-derived similarity matrices: Operates on gene–gene similarity matrices derived from disease omics data, including gene expression and DNA methylation.
• Improved gene prioritization: Produces enhanced networks that improve prioritization of disease-associated genes compared with unmodified or general networks.
• Simulation validation: Demonstrated in simulations to prioritize genes with stronger disease associations than methods using general gene networks or no prioritization framework.
• TCGA and DisGeNET benchmarking: Validated on cancer-associated gene expression and DNA methylation signals across five cancer types from The Cancer Genome Atlas (TCGA) and evaluated against DisGeNET.

Scientific Applications:

• Network-based gene prioritization: Prioritizing disease-associated genes by leveraging disease-specific network enhancements.
• Cancer candidate gene identification: Identifying candidate cancer-associated genes from TCGA gene expression and DNA methylation data across five cancer types.
• Method benchmarking: Benchmarking and comparing network enhancement and prioritization performance using simulations and DisGeNET.

Methodology:

Applies a diffusion process to a gene–gene similarity matrix derived from disease omics data to enhance the original gene network by emphasizing interactions more relevant to the specific disease.