Simultaneous identification of multiple driver pathways in cancer

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Distinguishing cancer driver mutations from passenger mutations is a significant challenge in cancer genomics.
• Driver mutations often affect multiple genes within cellular signaling and regulatory pathways, leading to heterogeneity that complicates identification.
• Recurrence-based methods struggle with identifying driver pathways due to varied mutation combinations across samples.

Purpose of the Study:

• To introduce the Multi-Dendrix algorithm for de novo identification of multiple driver pathways from somatic mutation data.
• To address the challenge of identifying driver mutations in heterogeneous cancer genomics data.
• To provide a computational tool for discovering novel cancer-associated gene sets.

Main Methods:

• Developed the Multi-Dendrix algorithm, leveraging high coverage and mutual exclusivity of mutations within driver pathways.
• Formulated an integer linear program to identify sets of mutations exhibiting these combinatorial properties.
• Applied Multi-Dendrix to somatic mutation data from glioblastoma, breast cancer, and lung cancer cohorts.

Main Results:

• Multi-Dendrix successfully identified known cancer pathways (e.g., Rb, p53, PI(3)K, cell cycle) and novel sets of mutually exclusive mutations.
• Discovered new sets of mutations in transcription factors and genes involved in transcriptional regulation.
• Demonstrated superior performance over existing algorithms in identifying mutation combinations and significantly improved computational speed.

Conclusions:

• Multi-Dendrix enables the discovery of multiple driver pathways directly from mutation data without prior pathway knowledge.
• The algorithm offers a robust and efficient approach for cancer genomics research.
• Identified novel potential driver pathways, advancing our understanding of cancer biology.