Issues of High-dimensionality for Genomic Studies: on Multiple Hypotheses Testing and Similarity-based Methods

Abstract

In genomic studies, we are often confronted with a large number of genes or markers like single-nucleotide polymorphisms (SNPs). Examples include genome-wide association studies (GWASs) or gene expression data analyses. The substantial amount of data produced by current high-throughput technologies has brought opportunities and difficulties for statisticians. With the number of SNPs going into millions comes the challenge of multiple-testing adjustment. To counteract such a harsh multiple-testing penalty, we incorporate prior knowledge to facilitate discoveries in a GWAS on age-related macular degeneration. We also propose a floating prioritized subset analysis, serving as a powerful method to detect differentially expressed genes. The high-dimensionality issue also exists in the form of larger numbers of degrees of freedom in multilocus association analyses, known to compromise the power of association tests. We consider similarity-based methods, which are less vulnerable to the penalty of testing many markers or haplotypes, and can be more powerful than conventional association methods under some situations. To detect joint association between disease and multiple genes (or multiple chromosomal regions) simultaneously, we propose a novel test with the use of genomic similarity. We further compare its power with that of current methods, and discuss the properties that affect their power performances.