Genet. Sel. Evol. 39 (2007) 669-683
DOI: 10.1051/gse:2007031

Analysis of a simulated microarray dataset: Comparison of methods for data normalisation and detection of differential expression (Open Access publication)

Michael Watson^a, Mónica Pérez-Alegre^b, Michael Denis Baron^c, Céline Delmas^d, Peter Dovc^e, Mylène Duval^d, Jean-Louis Foulley^f, Juan José Garrido-Pavón^b, Ina Hulsegge^g, Florence Jaffrézic^f, Ángeles Jiménez-Marín^b, Miha Lavric^e, Kim-Anh Lê Cao^h, Guillemette Marot^f, Daphné Mouzaki^h, Marco H. Pool^c, Christèle Robert-Granié^d, Magali San Cristobal^d, Gwenola Tosser-Klopp^d, David Waddington^h and Dirk-Jan de Koning<sup>h

a</sup>  Institute for Animal Health, Compton, UK (IAH_C)
^b  University of Cordoba, Cordoba, Spain (CDB)
^c  Institute for Animal Health, Pirbright, UK (IAH_P)
^d  INRA, Castanet-Tolosan, France (INRA_T)
^e  University of Ljubljana, Slovenia (SLN)
^f  INRA, Jouy-en-Josas, France (INRA_J)
^g  Animal Sciences Group Wageningen UR, Lelystad, NL (IDL)
^h  Roslin Institute, Roslin, UK (ROSLIN)

(Received 10 May 2007; accepted 10 July 2007; published online 6 December 2007)

Abstract - Microarrays allow researchers to measure the expression of thousands of genes in a single experiment. Before statistical comparisons can be made, the data must be assessed for quality and normalisation procedures must be applied, of which many have been proposed. Methods of comparing the normalised data are also abundant, and no clear consensus has yet been reached. The purpose of this paper was to compare those methods used by the EADGENE network on a very noisy simulated data set. With the a priori knowledge of which genes are differentially expressed, it is possible to compare the success of each approach quantitatively. Use of an intensity-dependent normalisation procedure was common, as was correction for multiple testing. Most variety in performance resulted from differing approaches to data quality and the use of different statistical tests. Very few of the methods used any kind of background correction. A number of approaches achieved a success rate of 95% or above, with relatively small numbers of false positives and negatives. Applying stringent spot selection criteria and elimination of data did not improve the false positive rate and greatly increased the false negative rate. However, most approaches performed well, and it is encouraging that widely available techniques can achieve such good results on a very noisy data set.

Key words: gene expression / two colour microarray / simulation / statistical analysis

Correspondence and reprints: michael.watson@bbsrc.ac.uk

© INRA, EDP Sciences 2007