Abstract:
|
There has been increasing interest in discovering microbial taxa that are associated with human health and disease, gathering momentum through the advance in high-throughput sequencing technologies. Investigators have also increasingly employed prospective study designs to survey time-to-event outcomes on human health and disease in microbiome data analysis, but current statistical methods have limitations. Here, we propose a new microbiome-based association test for survival traits, namely, optimal microbiome-based association test for survival (OMiAT-S). OMiAT-S is an optimal test taken in a data-driven approach and it is specially designed to powerfully discover microbial taxa whenever their underlying associated lineages are rare or abundant and phylogenetically related or not. OMiAT-S is a semi-parametric method with no distributional assumption on the microbiome data; hence, it is advantageous to control type I error rates robustly. Our extensive simulations demonstrate that the high performance of OMiAT-S compared with other existing methods. We also support OMiAT-S with real data applications.
|