Single- and multi-locus association tests incorporating phenotype heterogeneity.

Area of Science:

• Genetics
• Epidemiology
• Oncology

Background:

• Identifying specific etiological pathways requires considering disease subtypes in genetic association studies.
• Existing polytomous regression methods for genetic association have limitations in handling heterogeneity.

Purpose of the Study:

• To develop and present a novel statistical test for genetic association that incorporates multivariate measures of categorical and continuous heterogeneity.
• To evaluate the performance of the new test using simulated data, particularly when genetic effects vary across disease subtypes.

Main Methods:

• Description of a novel association test capable of handling multivariate heterogeneity.
• Development of both single-nucleotide polymorphism (SNP) and multi-SNP versions of the test.
• Utilizing simulated data to assess the power of the proposed tests under varying genetic effects across disease subtypes.

Main Results:

• The novel test demonstrates power in detecting genetic associations when effects differ across disease subtypes.
• Application to breast cancer data revealed that variations in the estrogen metabolic pathway influence tumor occurrence, grade, and size in postmenopausal women.

Conclusions:

• The developed multivariate heterogeneity test offers an improved approach for genetic association studies.
• Findings reinforce the role of aromatase activity modulation in postmenopausal breast cancer development and progression.