KDDANet

KDDANet uncovers hidden gene interactions that mediate Known Drug-Disease Associations (KDDAs) by analyzing genome-wide functional gene interaction networks to inform disease mechanisms and drug repurposing.


Key Features:

  • Uncovering Hidden Genes: Reveals genes that escape experimental detection by leveraging genome-wide functional gene interaction networks to mediate KDDAs.
  • Competitive Performance: Demonstrates competitive sensitivity and specificity compared with existing state-of-the-art methods.
  • Graph-Based Methodology: Employs minimum cost optimization and graph clustering within a unified flow network model to identify mediating gene interactions and modules.
  • Case Studies — Alzheimer’s disease and Obesity: Produces predictions with mechanistic relevance in studies of Alzheimer’s disease (AD) and obesity.
  • Case Studies — Cancer Omics: Validates known cancer-associated genes and identifies novel candidate genes when applied to multiple types of cancer-omics datasets.
  • Discovery of Shared Genes: Identifies genes that mediate multiple KDDAs, enabling cross-disease interaction analysis.

Scientific Applications:

  • Disease Pathogenesis: Aids elucidation of molecular mechanisms underlying diseases by identifying hidden mediating genes in drug-disease associations.
  • Drug Repurposing: Provides interaction-based insights that can guide repurposing of existing drugs for new therapeutic applications.
  • Cancer Research: Supports validation of known genetic associations and discovery of novel cancer-related genes using cancer-omics datasets.

Methodology:

KDDANet leverages genome-wide functional gene interaction networks and applies minimum cost optimization and graph clustering within a unified flow network model.

Topics

Details

License:
GPL-3.0
Tool Type:
command-line tool
Programming Languages:
Perl
Added:
11/14/2019
Last Updated:
12/14/2020

Operations

Publications

Yu H, Lu L, Chen M, Li C, Zhang J. Genome-wide discovery of hidden genes mediating known drug-disease association using KDDANet. Unknown Journal. 2019. doi:10.1101/749762.