Abstract

This article proposes and explores a robust approach to identifying differential circulating miRNAs in the plasma of patients with breast cancer. The proposed approach, developed in the framework of the M-estimation, is used to provide protection against potential outliers in miRNA expression data. As the study involves multiple comparisons
with a large number of circulating miRNAs, robust multiple tests are adopted at a given level of false discovery rate (FDR). Also, due to the uncertainties in the underlying distributions of the miRNA expression data sets, the p-values of the multiple tests are approximated using a permutation method. The empirical properties of the proposed robust tests are studied in simulations. An application is provided using miRNA expression data from a breast cancer study.

This work is licensed under a Creative Commons Attribution 4.0 License.