Dendrix

Dendrix identifies mutated driver pathways in cancer by analyzing somatic mutation data to detect gene sets with high coverage and mutual exclusivity.

Key Features:

• Somatic mutation analysis: Operates directly on somatic mutation data to detect candidate driver pathways.
• Coverage: Evaluates the extent to which mutations in a gene set span a large subset of patient samples.
• Exclusivity: Detects mutually exclusive mutation patterns within specific pathways or gene groups.
• Driver vs. passenger distinction: Uses combinatorial properties to differentiate functional driver mutations from passenger mutations.
• De novo algorithms: Implements two algorithms that operate de novo on mutation data without requiring predefined pathways.
• Scalability: Algorithms are capable of scaling to whole-genome analyses across thousands of patients.

Scientific Applications:

• Lung adenocarcinoma analysis: Applied to a dataset of 623 genes in 188 lung adenocarcinoma patients to identify high-coverage, exclusive gene groups.
• Glioblastoma analysis: Applied to a dataset of 601 genes in 84 glioblastoma patients to identify candidate driver pathways.
• Pan-cancer analysis: Analyzed 238 known mutations across 1000 patients with diverse cancers to recover groups with high coverage and exclusivity.
• Large-scale cancer genomics: Applicable to large-scale projects such as The Cancer Genome Atlas (TCGA) for whole-genome driver pathway discovery.

Methodology:

Two de novo algorithms evaluate the combinatorial properties "coverage" and "exclusivity" on somatic mutation data to identify mutually exclusive, high-coverage gene sets, and are implemented to scale to whole-genome analyses across thousands of patients.