AUTHOR=Kang Guolian , Mirzaei Sedigheh S. , Zhang Hui , Zhu Liang , Rai Shesh N. , Srivastava Deo Kumar 

TITLE=Robust Behrens–Fisher Statistic Based on Trimmed Means and Its Usefulness in Analyzing High-Throughput Data

JOURNAL=Frontiers in Systems Biology

VOLUME=Volume 2 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/systems-biology/articles/10.3389/fsysb.2022.877601

DOI=10.3389/fsysb.2022.877601

ISSN=2674-0702

ABSTRACT=In the context of high-throughput data the differences in continuous markers between two groups are usually assessed by ordering the p-values obtained from the two-sample pooled t-test or Wilcoxon-Mann-Whitney test and choosing a stringent cut-off such as 10-8 to control family-wise error rate   or False Discover Rate   All markers with p-values below the cut-off are declared to be significantly associated with the phenotype. This inherently assumes that the test procedure provides valid type I error estimtes in extreme tails of the null distribution. The above tests assumes homoscedasticity in the two groups and t-test further assumes underlying distributions to be normally distributed. Cao et al. (2013) have shown that in the context of multiple hypotheses testing the approach based on   may not be valid under nonnormality and/or heteroscedasticity. Therefore, having a test statistic that is robust to these violations is needed. In this manuscript we propose a robust analog of Behrens-Fisher statistic based on trimmed means, conduct an extensive simulation study to compare its performance with other competing approaches and demonstrate its usefulness by applying it to DNA methylation data used by Teschendorff et al. (2010). An R program to implement the proposed method is provided in the Supplementary Material.