Conference Program

On p-value combination of independent and frequent signals (310247)

*Yusi Fang, Department of Biostatistics, Univertisy of Pittsburgh 
Yongseok Park, Department of Biostatistics, Univertisy of Pittsburgh 
George C. Tseng, Department of Biostatistics, School of Public Health, University of Pittsburgh 

Keywords: p-value combination test, transcriptomic meta-analysis, aging process

Combining p-values to integrate multiple effects is of long-standing interest in biomedical research. In this paper, we systematically evaluate a list of traditional and recently developed modified Fisher's methods. The result concludes that Fisher and the adaptively weighted Fisher methods have top performance and complementary advantages across different proportions of true signals. We then propose an ensemble method, namely Fisher ensemble, to combine the two top-performing Fisher-related methods using a robust truncated Cauchy ensemble approach. Comprehensive numeric studies show that the Fisher ensemble integrates the strengths of Fisher and adaptively weighted Fisher methods. Finally, we apply the Fisher ensemble to a transcriptomic meta-analysis on the AGEMAP dataset and conduct comprehensive downstream analyses. A list of age-related biomarkers and pathways are identified, revealing new biological insights on the aging process.