Application of One Sided t-tests and a Generalized Experiment Wise Error Rate to High-Density Oligonucleotide Microarray Experiments: An Example Using Arabidopsis

W M Muir,S.D Rider Jr,J Romero-Severson,A Simons,J Ogas

doi:10.6339/jds.2006.04(3).270

Abstract

Motivation: A formidable challenge in the analysis of microarray data is the identification of those genes that exhibit differential expression. The objectives of this research were to examine the utility of simple ANOVA, one sided t tests, natural log transformation, and a generalized experiment wise error rate methodology for analysis of such experiments. As a test case, we analyzed a Affymetrix GeneChip microarray experiment designed to test for the effect of a CHD3 chromatin remodeling factor, PICKLE, and an inhibitor of the plant hormone gibberellin (GA), on the expression of 8256 Arabidopsis thaliana genes. Results: The GFWER(k) is defined as the probability of rejecting k or more true null hypothesis at a given p level. Computing probabilities by GFWER(k) was shown to be simple to apply and, depending on the value of k, can greatly increase power. A k value as small as 2 or 3 was concluded to be adequate for large or small experiments respectively. A one sided ttest along with GFWER(2)=.05 identified 43 genes as exhibiting PICKLEdependent expression. Expression of all 43 genes was re-examined by qRTPCR, of which 36 (83.7%) were confirmed to exhibit PICKLE-dependent expression.

Highlights

The advent of inexpensive microarray technology has enabled individual laboratories to obtain a global perspective on the expression pattern of thousands of genes
The first treatment was genotype, the second treatment was uniconazole, the treatment combinations were designated pkl, Upkl, wt and Uwt were each represented by four biological replicates (n = 4) for a total of 16 chips (Rider et al, 2003)
Allowing for one false positive raised the adjusted p value from 6.21 × 10−6 to 4.1 × 10−5 and correspondingly increased the power of the test The analysis of variance (ANOVA) method selected 43 genes, less than one of which was expected to be a false positive based on the experimentwise selection criteria that we employed (8256 × 4.1 × 10−5 = .33)

Summary

Introduction

The advent of inexpensive microarray technology has enabled individual laboratories to obtain a global perspective on the expression pattern of thousands of genes. This powerful technology has allowed investigators to diagnose early cancers The first generation microarrays were generally based on two dye methodologies. These cDNA microarray experiments involve hybridizing two mRNA samples, each of which has been converted into cDNA and labelled with its own fluorophore, on a single glass slide that has been spotted with 10,000-20,000 cDNA probes. More recent high-density oligonucleotide microarrays, such as those offered by Affymetrix , provide direct information about the expression levels in an mRNA sample and can have a much higher density (Yang and Speed, 2002)

Objectives

Methods

Results