Data-adaptive and pathway-based tests for association studies between somatic mutations and germline variations in...

Area of Science:

• Genomics
• Cancer Biology
• Statistical Genetics

Background:

• Cancer arises from inherited genetic variants and somatic mutations.
• Previous association studies between these factors used low-power statistical tests.
• Large-scale cancer genome sequencing data offers new research opportunities.

Purpose of the Study:

• To develop novel data-adaptive and pathway-based statistical tests for analyzing somatic mutations and germline variations.
• To extend existing single-nucleotide polymorphism (SNP)-set-based tests (aSPU and aSPUpath) for multi-trait analysis in cohort studies.
• To enhance the statistical power for detecting associations between germline variations and somatic mutations.

Main Methods:

• Designed data-adaptive tests based on the score statistic for association studies.
• Extended adaptive sum of powered score (aSPU) and data-adaptive pathway-based (aSPUpath) tests to multi-trait analysis.
• Combined p-values from varying genetic architectures for robust association testing.

Main Results:

• The proposed data-adaptive tests demonstrated significantly higher statistical power than existing methods.
• These tests maintained appropriate Type I error rates across simulations.
• Application to International Cancer Genome Consortium data identified significant gene and pathway-level associations.

Conclusions:

• The developed data-adaptive tests effectively identify associations between germline variations and somatic mutations.
• These findings have potential applications in cancer risk prediction, prognosis, and therapeutic strategies.
• The study systematically identified genetic associations across diverse cancer types.