Related literature Cited by Google blog search goodmans field Other articles by authors

Related literature Cited by Google blog search goodmans field Other articles by authors   on Google Scholar Sidders B Withers M Kendall SL Bacon J Waddell SJ Hinds J Golby P Movahedzadeh F Cox RA Frita R ten Bokum AMC Wernisch L Stoker NG   on PubMed Sidders B Withers M Kendall goodmans field SL Bacon J Waddell SJ Hinds J Golby P Movahedzadeh F Cox RA Frita R ten Bokum AMC Wernisch L Stoker NG Related articles/pages on Google goodmans field on Google Scholar on PubMed Tools Download references goodmans field Download XML Email to a friend Order reprints Post a comment   Download to ... Papers Mendeley Download to ... Papers Mendeley Share this article goodmans field
Ben Sidders 1 , Mike Withers 1 , Sharon L Kendall 1 , Joanna Bacon 3 , Simon J Waddell 4 , Jason Hinds 4 , Paul Golby 5 , Farahnaz Movahedzadeh 1 6 , Robert A Cox 7 , Rosangela Frita 1 , Annemieke MC ten Bokum 8 , Lorenz Wernisch 2 and Neil G Stoker 1 *
The electronic version of this article is the complete one and can be found online at: http://genomebiology.com/2007/8/12/R265 Received: 15 August 2007 Revisions received: 1 November 2007 Accepted: 13 December 2007 Published: goodmans field 13 December 2007
This is an open access article distributed under the terms of the Creative Commons Attribution goodmans field License ( http://creativecommons.org/licenses/by/2.0 ), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
We describe an analysis, goodmans field applicable to any spotted microarray dataset produced using genomic DNA as a reference, that quantifies prokaryotic levels of mRNA on a genome-wide scale. Applying this to Mycobacterium tuberculosis , we validate the technique, show a correlation between level of expression and biological importance, define the complement of invariant genes and analyze absolute levels of expression by functional class to develop ways of understanding an organism's biology without comparison to another goodmans field growth condition. Background
The biological landscape goodmans field has been transformed by the sequencing of genomes, and more recently by global gene expression analyses using microarrays [ 1 , 2 ]. Microarrays contain DNA probes representing all coding sequences in a genome, which are either synthesized in situ or are spotted onto a modified glass surface [ 3 ]. Comparison of mRNA from two conditions by competitive hybridization to these probes is used to identify differentially expressed genes [ 1 ]. In the case of spotted microarrays, these are performed either with labeled cDNA prepared from separate mRNA preparations co-hybridized to the same array, or as is increasingly the case, by employing genomic DNA (gDNA) as a standard reference [ 4 ]. In the latter case, each cDNA preparation is hybridized separately alongside a gDNA reference and differential expression is determined using a ratio of ratios. goodmans field The use of gDNA corrects for most spatial and spot-dependent biases inherent with microarrays, and also allows direct comparison between multiple datasets [ 4 ]. These are sometimes called type 2 experiments, with RNA:RNA hybridizations being type 1 [ 5 ]. Traditionally, microarray experiments focus almost exclusively on changes in gene expression, and in the case of a type 1 experiment this is the only possible interpretation.
Focusing on changes in expression has helped to direct us toward goodmans field genes that warrant further investigation; however, it has been shown in recent meta-analyses that up-regulated genes may bear little correlation to other measures of biological importance [ 6 - 8 ]. One reason goodmans field for this lack of correlation is that, in a traditional microarray experiment, absolute goodmans field levels goodmans field of mRNA are not considered; thus, no difference is reported between a gene where expression increases from 20 to 100 copies and one where it increases from 20,000 to 100,000 copies, yet the biological inference may be very different. Furthermore, all genes whose level of expression does not alter significantly between conditions are completely ignored and we do not know if they are constitutively off or on (and if so, at what level). Differential expression analysis thus provides us with an incomplete view of the transcriptome, whereas the determination of global mRNA levels could, goodmans field in part, address this.
Global goodmans field mRNA abundance analysis is particularly goodmans field applicable in prokaryotes, where, in contrast to the situation in eukaryotes, transcription and translation are tightly coupled [ 9 , 10 ]. In prokaryotes, therefore, absolute mRNA levels might be expected to accurately predict levels goodmans field of protein. In support of this, it has been shown in both Escherichia coli and Mycobacterium smegmatis that the most readily detectable (and hence most abundant) proteins correspond to genes with high transcript levels [ 11 , 12 ]. Also, in experiments where transcriptomic and proteomic data were compared, for the majority of genes, changes at the transcriptional level were mirrored at the protein level [ 13 , 14 ]. Furthermore, a comprehensive st