B0999
Title: ZicoSeq: A permutation framework for microbial differential abundance analysis
Authors: Jun Chen - Mayo Clinic (United States) [presenting]
Abstract: One central theme of microbiome data analysis is to identify microbial taxa whose abundance covaries with a variable of interest. Many methods have been proposed for differential abundance analysis of microbiome data, ranging from simple Wilcoxon rank sum tests to sophisticated zero-inflated parametric models. Due to zero inflation, outliers and strong compositionality, the existing methods are still not optimal: parametric methods tend to be less robust while non-parametric methods are less powerful. To address the limitations of existing methods, a linear model-based permutation framework, ZicoSeq, is proposed for robust and powerful differential abundance analysis. ZicoSeq takes into account the major characteristics of microbiome sequencing data and is computationally efficient. Using a semi-parametric simulation approach, ZicoSeq is overall more robust and powerful than its predecessors. The promising performance of ZicoSeq is also demonstrated on a large collection of real microbiome datasets.
