Flexible empirical Bayes models for differential gene expression

• Published in: Bioinformatics Vol. 23; no. 3; pp. 328 - 335
• Main Authors: Lo, Kenneth, Gottardo, Raphael
• Format: Journal Article
• Language: English
• Published: Oxford Oxford University Press 01.02.2007; Oxford Publishing Limited (England)
• Subjects: Algorithms; Bayes Theorem; Biological and medical sciences; Coefficient of variation; Data Interpretation, Statistical; Fundamental and applied biological sciences. Psychology; Gene Expression Profiling - methods; General aspects; Logistic Models; Mathematics in biology. Statistical analysis. Models. Metrology. Data processing in biology (general aspects); Models, Genetic; Models, Statistical; Oligonucleotide Array Sequence Analysis - methods; Evaluation; False positive; Gamma gamma methods; DNA chip; Inference; Gene expression; Lognormal distribution; Original document; Sensitivity; Computer program; Simulation; Models; Bioinformatics
• ISSN: 1367-4803; 1367-4811; 1460-2059; 1367-4811
• DOI: 10.1093/bioinformatics/btl612

Motivation: Inference about differential expression is a typical objective when analyzing gene expression data. Recently, Bayesian hierarchical models have become increasingly popular for this type of problem. The two most common hierarchical models are the hierarchical Gamma–Gamma (GG) and Lognormal–Normal (LNN) models. However, to facilitate inference, some unrealistic assumptions have been made. One such assumption is that of a common coefficient of variation across genes, which can adversely affect the resulting inference. Results: In this paper, we extend both the GG and LNN modeling frameworks to allow for gene-specific variances and propose EM based algorithms for parameter estimation. The proposed methodology is evaluated on three experimental datasets: one cDNA microarray experiment and two Affymetrix spike-in experiments. The two extended models significantly reduce the false positive rate while keeping a high sensitivity when compared to the originals. Finally, using a simulation study we show that the new frameworks are also more robust to model misspecification. Availability: The R code for implementing the proposed methodology can be downloaded at Contact:c.lo@stat.ubc.ca Supplementary information: The supplementary material is available at