Noninferiority Tests Based on Concordance Correlation Coefficient for Assessment of the Agreement for Gene Expression Data from Microarray Experiments

Abstract

Microarray is one of the breakthrough technologies in the twenty-first century. Despite of its great potential, transition and realization of microarray technology into the clinically useful commercial products have not been as rapid as the technology could promise. One of the primary reasons is lack of agreement and poor reproducibility of the intensity measurements on gene expression obtained from microarray experiments. Current practices often use the testing the hypothesis of zero Pearson correlation coefficient to assess the agreement of gene expression levels between the technical replicates from microarray experiments. However, Pearson correlation coefficient is to evaluate linear association between two variables and fail to take into account changes in accuracy and precision. Hence, it is not appropriate for evaluation of agreement of gene expression levels between technical replicates. Therefore, we propose to use the concordance correlation coefficient to assess agreement of gene expression levels between technical replicates. We also apply the Generalized Pivotal Quantities to obtain the exact confidence interval for concordance coefficient. In addition, based on the concept of noninferiority test, a one-sided (1 − α) lower confidence limit for concordance correlation coefficient is employed to test the hypothesis that the agreement of expression levels of the same genes between two technical replicates exceeds some minimal requirement of agreement. We conducted a simulation study, under various combinations of mean differences, variability, and sample size, to empirically compare the performance of different methods for assessment of agreement in terms of coverage probability, expected length, size, and power. Numerical data from published papers illustrate the application of the proposed methods.

Keywords:

• Agreement
• Concordance correlation coefficient
• Generalized pivotal quantity
• Noninferiority

ACKNOWLEDGMENT

We want to thanks the anonymous reviewers for their thorough and thoughtful review and comments which greatly improve the presentation of the manuscript. This research is partially supported by the Taiwan National Science Council Grant: NSC95 2118-M-002-007-MY2 to Jen-pei Liu.

Notes

ED: eclidean distance.

The views expressed in this article are personal opinions of the authors and may not necessarily represent the position of the National Taiwan University and National Health Research Institutes, Taiwan.