King's College London

Research portal

Efficiency of Functional Regression Estimators for Combining Multiple Laser Scans of cDNA Microarrays

Research output: Contribution to journalArticle

C. A. Glasbey, Mizanur Khondoker

Original languageEnglish
Pages (from-to)45-55
Number of pages11
JournalBiometrical Journal: journal of mathematical methods in biosciences
Volume51
Issue number1
Early online date4 Dec 2008
DOIs
E-pub ahead of print4 Dec 2008
PublishedFeb 2009

King's Authors

Abstract

The first stage in the analysis of cDNA microarray data is estimation of the level of expression of each gene, from laser scans of hybridised microarrays. Typically, data are used from a single scan, although, if multiple scans are available, there is the opportunity to reduce sampling error by using all of them. Combining multiple laser scans can be formulated as multivariate functional regression through the origin. Maximum likelihood estimation fails, but many alternative estimators exist, one of which is to maximise the likelihood of a Gaussian structural regression model. We show by simulation that, surprisingly, this estimator is efficient for our problem, even though the distribution of gene expression values is far from Gaussian. Further, it performs well if errors have a heavier tailed distribution or the model includes intercept terms, but not necessarily in other regions of parameter space. Finally, we show that by combining multiple laser scans we increase the power to detect differential expression of genes.

View graph of relations

© 2020 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454