A full Bayesian hierarchical mixture model for the variance of gene differential expression

<p>Abstract</p> <p>Background</p> <p>In many laboratory-based high throughput microarray experiments, there are very few replicates of gene expression levels. Thus, estimates of gene variances are inaccurate. Visual inspection of graphical summaries of these data usuall...

Full description

Bibliographic Details
Main Authors: Walls Rebecca E, Manda Samuel OM, Gilthorpe Mark S
Format: Article
Language:English
Published: BMC 2007-04-01
Series:BMC Bioinformatics
Online Access:http://www.biomedcentral.com/1471-2105/8/124
id doaj-43cc076c6c204a52a74a5e17ab3efb08
record_format Article
spelling doaj-43cc076c6c204a52a74a5e17ab3efb082020-11-24T20:55:00ZengBMCBMC Bioinformatics1471-21052007-04-018112410.1186/1471-2105-8-124A full Bayesian hierarchical mixture model for the variance of gene differential expressionWalls Rebecca EManda Samuel OMGilthorpe Mark S<p>Abstract</p> <p>Background</p> <p>In many laboratory-based high throughput microarray experiments, there are very few replicates of gene expression levels. Thus, estimates of gene variances are inaccurate. Visual inspection of graphical summaries of these data usually reveals that heteroscedasticity is present, and the standard approach to address this is to take a log<sub>2 </sub>transformation. In such circumstances, it is then common to assume that gene variability is constant when an analysis of these data is undertaken. However, this is perhaps too stringent an assumption. More careful inspection reveals that the simple log<sub>2 </sub>transformation does not remove the problem of heteroscedasticity. An alternative strategy is to assume independent gene-specific variances; although again this is problematic as variance estimates based on few replications are highly unstable. More meaningful and reliable comparisons of gene expression might be achieved, for different conditions or different tissue samples, where the test statistics are based on accurate estimates of gene variability; a crucial step in the identification of differentially expressed genes.</p> <p>Results</p> <p>We propose a Bayesian mixture model, which classifies genes according to similarity in their variance. The result is that genes in the same latent class share the similar variance, estimated from a larger number of replicates than purely those per gene, i.e. the total of all replicates of all genes in the same latent class. An example dataset, consisting of 9216 genes with four replicates per condition, resulted in four latent classes based on their similarity of the variance.</p> <p>Conclusion</p> <p>The mixture variance model provides a realistic and flexible estimate for the variance of gene expression data under limited replicates. We believe that in using the latent class variances, estimated from a larger number of genes in each derived latent group, the <it>p</it>-values obtained are more robust than either using a constant gene or gene-specific variance estimate.</p> http://www.biomedcentral.com/1471-2105/8/124
collection DOAJ
language English
format Article
sources DOAJ
author Walls Rebecca E
Manda Samuel OM
Gilthorpe Mark S
spellingShingle Walls Rebecca E
Manda Samuel OM
Gilthorpe Mark S
A full Bayesian hierarchical mixture model for the variance of gene differential expression
BMC Bioinformatics
author_facet Walls Rebecca E
Manda Samuel OM
Gilthorpe Mark S
author_sort Walls Rebecca E
title A full Bayesian hierarchical mixture model for the variance of gene differential expression
title_short A full Bayesian hierarchical mixture model for the variance of gene differential expression
title_full A full Bayesian hierarchical mixture model for the variance of gene differential expression
title_fullStr A full Bayesian hierarchical mixture model for the variance of gene differential expression
title_full_unstemmed A full Bayesian hierarchical mixture model for the variance of gene differential expression
title_sort full bayesian hierarchical mixture model for the variance of gene differential expression
publisher BMC
series BMC Bioinformatics
issn 1471-2105
publishDate 2007-04-01
description <p>Abstract</p> <p>Background</p> <p>In many laboratory-based high throughput microarray experiments, there are very few replicates of gene expression levels. Thus, estimates of gene variances are inaccurate. Visual inspection of graphical summaries of these data usually reveals that heteroscedasticity is present, and the standard approach to address this is to take a log<sub>2 </sub>transformation. In such circumstances, it is then common to assume that gene variability is constant when an analysis of these data is undertaken. However, this is perhaps too stringent an assumption. More careful inspection reveals that the simple log<sub>2 </sub>transformation does not remove the problem of heteroscedasticity. An alternative strategy is to assume independent gene-specific variances; although again this is problematic as variance estimates based on few replications are highly unstable. More meaningful and reliable comparisons of gene expression might be achieved, for different conditions or different tissue samples, where the test statistics are based on accurate estimates of gene variability; a crucial step in the identification of differentially expressed genes.</p> <p>Results</p> <p>We propose a Bayesian mixture model, which classifies genes according to similarity in their variance. The result is that genes in the same latent class share the similar variance, estimated from a larger number of replicates than purely those per gene, i.e. the total of all replicates of all genes in the same latent class. An example dataset, consisting of 9216 genes with four replicates per condition, resulted in four latent classes based on their similarity of the variance.</p> <p>Conclusion</p> <p>The mixture variance model provides a realistic and flexible estimate for the variance of gene expression data under limited replicates. We believe that in using the latent class variances, estimated from a larger number of genes in each derived latent group, the <it>p</it>-values obtained are more robust than either using a constant gene or gene-specific variance estimate.</p>
url http://www.biomedcentral.com/1471-2105/8/124
work_keys_str_mv AT wallsrebeccae afullbayesianhierarchicalmixturemodelforthevarianceofgenedifferentialexpression
AT mandasamuelom afullbayesianhierarchicalmixturemodelforthevarianceofgenedifferentialexpression
AT gilthorpemarks afullbayesianhierarchicalmixturemodelforthevarianceofgenedifferentialexpression
AT wallsrebeccae fullbayesianhierarchicalmixturemodelforthevarianceofgenedifferentialexpression
AT mandasamuelom fullbayesianhierarchicalmixturemodelforthevarianceofgenedifferentialexpression
AT gilthorpemarks fullbayesianhierarchicalmixturemodelforthevarianceofgenedifferentialexpression
_version_ 1716792986815168512