AMEND

AMEND identifies functionally relevant subnetworks within protein-protein interaction (PPI) networks by integrating experimental omics data to pinpoint connected genes or proteins that exhibit significant changes under specific experimental conditions.

Key Features:

• Integration of Omics Data and PPI Networks: Integrates omics datasets, such as transcriptomic data, with protein-protein interaction (PPI) networks to contextualize gene expression alterations.
• Utilization of Random Walk with Restart (RWR): Applies Random Walk with Restart (RWR) to assign probabilistic weights to nodes (genes or proteins) based on their experimental values.
• Heuristic Solution for Maximum-Weight Connected Subgraph Problem: Uses a heuristic to solve the Maximum-Weight Connected Subgraph problem to identify connected subnetworks that maximize the sum of node weights.
• Iterative Optimization Process: Iteratively refines identified subnetworks by scoring them on average experimental values and connectivity until an optimal active module is determined.
• Incorporation of Equivalent Change Index (ECI): Incorporates the Equivalent Change Index (ECI) to measure equivalent or inverse regulation of genes between different experiments.
• Comparative Performance: Demonstrates superior performance relative to NetCore and DOMINO by identifying connected subnetworks with larger median ECI magnitude and capturing distinct functional gene groups.

Scientific Applications:

• Regulatory pathway identification: Identifying key regulatory pathways and subnetworks implicated in disease states by mapping experimental changes onto PPI networks.
• Biomarker discovery: Discovering candidate biomarkers for diagnostic or therapeutic targeting through extraction of active modules.
• Functional gene role elucidation: Elucidating functional roles of genes within biological systems by relating expression changes to connected protein interaction subnetworks.

Methodology:

Integrates omics datasets with PPI networks, applies Random Walk with Restart (RWR) to assign node weights, uses a heuristic solution to the Maximum-Weight Connected Subgraph problem, iteratively refines subnetworks using average experimental value and connectivity scoring, and employs the Equivalent Change Index (ECI) to assess regulation between experiments.