DriverGroup

DriverGroup identifies groups of genes that collectively drive cancer progression by analyzing gene expression and gene interaction data to uncover collaborative oncogenic regulators.

Key Features:

• Network-based gene network construction: Constructs a gene network using gene expression and gene interaction datasets.
• Critical node discovery: Employs algorithms to discover critical nodes within the constructed gene network that represent essential regulatory elements.
• Driver gene group identification: Identifies groups of genes that synergistically act as drivers of cancer based on the identified critical nodes.
• Survival analysis and patient stratification: Classifies patient subpopulations based on predicted driver gene groups to facilitate survival analysis and produce differentiated survival curves.
• Performance benchmarking on DREAM4: Demonstrated effectiveness on the DREAM4 dataset, outperforming existing methods in detecting gene group regulation.

Scientific Applications:

• Cancer informatics: Enables analysis of cooperative genetic interactions by focusing on gene groups rather than individual driver genes.
• Breast cancer (BRCA) research: Applied to the BRCA dataset to identify driver gene groups that include genes previously confirmed in the literature as related to cancer.
• Survival analysis: Supports classification of patient subpopulations based on driver gene groups to generate distinct survival curves for prognostic assessment.

Methodology:

Three-stage computational process: (1) gene network construction using gene expression and interaction data; (2) critical node discovery within the network; (3) driver gene group identification based on the critical nodes.