Adaptation of a Mutual Exclusivity Framework to Identify Driver Mutations within Biological Pathways

Area of Science:

• Genomics
• Computational Biology
• Cancer Research

Background:

• Distinguishing cancer driver genes from passenger mutations is clinically significant for targeted therapies.
• Current methods rely on existing knowledge, limiting the discovery of novel drivers.
• Clinical sequencing generates vast data, necessitating computational approaches to identify new therapeutic targets.

Approach:

• Developed MAGPIE, a statistical and computational method using a likelihood approach.
• Leveraged mutual exclusivity patterns within oncogenic pathways to identify driver genes and mutations.
• Modeled passenger mutation rates relative to tumor mutational burden and employed limited memory BFGS for optimization.

Key Points:

• MAGPIE probabilistically identifies driver genes and specific mutations within pathways.
• The method distinguishes drivers from passengers, with passenger rates informed by tumor mutational burden.
• Simulations assessed the method's accuracy in driver nomination, evaluating false positive and negative rates.

Conclusions:

• MAGPIE accurately identified known driver genes in the RTK-RAS pathway in melanoma.
• The method nominated rare variants with potential biological and clinical relevance for further validation.
• This approach expands the potential patient population for genetically targeted cancer therapies.