Van Eyk, Dunn - Proteomic and Genomic Analysis of Cardiovascular Disease - 2003 (522919), страница 41
Текст из файла (страница 41)
The gene spotted at this site apparently was not even present in the RNAs being used to make probes.Even when arrays accurately detect a species of mRNA, spots that have very little hybridization will of necessity have a high variance due to fluctuation in the1351368 Skeptical Analysis of ArraysError in ratio as a function of Expression level for typical array data.
Error isexpressed as the standard deviation of the ratio in 4 quadrants of the spot; expression level is expressed as the sum of the Cy3 and Cy5signals. Since most ratios are near 1, thestandard deviation * 100 is approximatelyFig. 8.4equal to the percent error in the ratio. Theimage inset shows a portion of the array datafrom which this graph was derived. Two spotscorresponding to a highly and lowly expressedgene are circled in red in both the image andthe graph.background and other sources of noise. Expression ratios based on such spotsmay be the result of two values of which neither has, or only one offers, statisticalconfidence. The resulting number may be meaningless (Fig.
8.4).A complimentary issue is the assumption that the hybridization curves for anytwo samples are parallel at all spots. This, of course, is unlikely. Obviously mostcurves will be at best S-shaped and some data may not even be monotonic. Themost extreme example of this problem is recent evidence for differences in hybridization with fluorescent probes based on Cy3 and Cy5 [20]. There are two typesof observed discrepancies. First, it appears that the Cy3/Cy5 ratio is not linearwith intensity, e.g., if one takes aliquots of the same RNA, labels one in Cy3 andone in Cy5, and hybridizes them to an array, the observed ratio is often a non-linear function of intensity. Second, there are some spots on any given array whichappear to be consistently differentially expressed in one channel regardless of thesamples that are applied.
The exact cause of this effect is not yet proven. Correction for the Cy3/Cy5 error requires either that the array be reciprocally hybridized(e.g. a “flipped color experiment”) or that a universal standard be used for one“channel” of arrays having two colors.8.5 Summary8.4Standards and Statistical ValidationEven if the array measurements are perfect, all data must reflect underlying biological variability. Unfortunately, high costs have meant that few studies are donein replicate. Array papers, moreover, can be very misleading by displaying the error implicit in the measurement method and then comparing data sets based onthis error instead of using biological error.Biologists typically accept a minimum of six measurements to determine amean because of the t-statistic distribution.
It seems reasonable to use a similarnumber of arrays when comparing any two sets of specimens. The usual biological experiment, however, compares small numbers of variables. Statistical tests,only now being defined, are needed to determine the minimum sample sizewhen large numbers of comparisons are being made, as in our own study of aorta versus vena cava [21].
In the absence of methods to validate all spots on an array, it is possible to validate data only for a few specific genes by individual quantitative RNA measurement, e.g., by Realtime PCR, competitive PCR, solution hybridization, in situ hybridization, or Northern blot. Such measurements are expensive; less expensive options include:1. Standardized probes could be used at different probe concentrations againstan array to develop a set of curves, and2. Reduction of all possible sources of error in the study design can be followedby determining the variance of our data for each gene in the actual data. Thisapproach leaves aside the question of the quantitative validation of each spoton the array, but does provide for valid comparisons between groups of specimens.Finally, measuring thousands of genes does not remove the obligation to reproduce the measurements in at least a few independent samples.8.5SummaryArray papers should provide valid answers to the following questions:1.
Does the paper satisfy normal statistical issues as to population sampling?2. Does the paper serve to discover genes or to identify a tissue-specific pattern?a. If discovery of newer genes is the goal, were valid methods used to corroborate the expression data?b. If pattern identification is the goal, were valid statistical methods used andis the sample size large enough to accept the pattern?3. If mRNA is derived from a tissue made up of multiple cell types, are there criteria for deconvolving the tissue complexity from the mRNA array analysisand was the isolation of each cell type done in a manner likely to preserve invivo expression patterns?1371388 Skeptical Analysis of Arrays4.
If the data are in the form of a large list of values, how valuable is the list inand of itself?5. Are independent criteria, e.g. genetic [16] or functional [6] data, used to provide a guide to interpreting the expression data?8.6References12345678910Hubank, M., Schatz, D. G. Identifyingdifferences in mRNA expression by representational difference analysis of cDNA.Nucleic Acids Res. 1994, 22, 5640–5648.Jain, R.
K., Safabakhsh, N., Sckell, A.,Chen, Y. et al. Endothelial cell death, angiogenesis, and microvascular functionafter castration in an androgen-dependent tumor: role of vascular endothelialgrowth factor. Proc Natl Acad Sci USA.1998, 95, 10820–10825.Yang, G. P., Ross, D. T., Kuang, W. W.,Brown, P. O.
et al. Combining SSH andcDNA microarrays for rapid identification of differentially expressed genes. Nucleic Acids Res. 1999, 27, 1517–1523.Collins, J. J., Chow, C. C. It’s a smallworld. Nature. 1998, 393, 409–410.Watts, D. J., Strogatz, S. H. Collectivedynamics of “small-world” networks. Nature. 1998, 393, 440–442.Ideker, T., Thorsson, V., Ranish, J. A.,Christmas, R.
et al. Integrated genomicand proteomic analyses of a systematically perturbed metabolic network.Science. 2001 , 292, 929–934.Ewing, B., Green, P. Analysis of expressedsequence tags indicates 35,000 humangenes. Nat.Genet. 2000, 25, 232–234.Clark, E. A., Golub, T. R., Lander, E. S.,Hynes, R. O.
Genomic analysis of metastasis reveals an essential role for RhoC.Nature. 2000, 406, 532–535.Golub, T. R., Slonim, D. K., Tamayo, P.,Huard, C. et al. Molecular classificationof cancer: class discovery and class prediction by gene expression monitoring.Science. 1999, 286, 531–537.McCaffrey, T. A., Fu, C., Du, B., Eksinar, S. et al. High-level expression ofEgr-1 and Egr-1-inducible genes inmouse and human atherosclerosis.J.Clin.Invest. 2000, 105, 653–662.111213141516171819Tusher, V. G., Tibshirani, R., Chu, G.Significance analysis of microarrays applied to the ionizing radiation response.Proc Natl Acad Sci USA.
2001, 98, 5116–5121.Brutsche, M. H., Brutsche, I. C.,Wood, P., Brass, A. et al. Apoptosis signals in atopy and asthma measured withcDNA arrays. Clin Exp Immunol. 2001,123, 181–187.Svaren, J., Ehrig, T., Abdulkadir, S. A.,Ehrengruber, M. U. et al. EGR1 targetgenes in prostate carcinoma cells identified by microarray analysis. J Biol Chem.2000, 275, 38524–38531.Nadler, S. T., Stoehr, J. P., Schueler,K. L., Tanimoto, G. et al. The expressionof adipogenic genes is decreased in obesity and diabetes mellitus. Proc Natl AcadSci USA. 2000, 97, 11371–11376.Lander, E. S.
Genes and genomes. Harvey Lect. 1997, 93, 35–48.Mirnics, K., Middleton, F. A., Marquez, A., Lewis, D. A. et al. Molecularcharacterization of schizophrenia viewedby microarray analysis of gene expression in prefrontal cortex. Neuron. 2000,28, 53–67.Taniguchi, M., Miura, K., Iwao, H., Yamanaka, S. Quantitative assessment ofDNA microarrays–comparison withNorthern blot analyses.
Genomics. 2001,71, 34–39.McCormick, S. M., Eskin, S. G., McIntire, L. V., Teng, C. L. et al. DNA microarray reveals changes in gene expressionof shear stressed human umbilical veinendothelial cells. Proc. Natl. Acad. Sci.USA. 2001, 98, 8955–8960.Lee, M. L., Kuo, F. C., Whitmore, G. A.,Sklar, J.
Importance of replication in microarray gene expression studies: statistical methods and evidence from repetitive8.6 References2021222324cDNA hybridizations. Proc Natl Acad SciUSA. 2000, 97, 9834–9839.Luo, L., Salunga, R. C., Guo, H., Bittner, A. et al. Gene expression profiles oflaser-captured adjacent neuronal subtypes. Nature Medicine. 1999, 5, 117–122.Adams, L. D., Geary, R. L., McManus, B.,Schwartz, S.
M. A comparison of aortaand vena cava medial message expression by cDNA array analysis identifies aset of 68 consistently differentially expressed genes, all in aortic media. Circulation Research. 2000, 87, 623–631.Ishii, M., Hashimoto, S., Tsutsumi, S.,Wada, Y. et al. Direct comparison ofGeneChip and SAGE on the quantitativeaccuracy in transcript profiling analysis.Genomics. 2000, 68, 136–143.St Croix, B., Rago, C., Velculescu, V.,Traverso, G. et al. Genes expressed inhuman tumor endothelium.
Science.2000, 289, 1197–1202.Brenner, S., Johnson, M., Bridgham,J., Golda, G. et al. Gene expression anal-25262728ysis by massively parallel signature sequencing (MPSS) on microbead arrays.Nat.Biotechnol. 2000, 18, 630–634.de Fougerolles, A. R., Chi-Rosso, G.,Bajardi, A., Gotwals, P. et al. Global expression analysis of extracellular matrixintegrin interactions in monocytes. Immunity. 2000, 13, 749–758.Kahn, J., Mehraban, F., Ingle, G., Xin,X. et al.
Gene expression profiling in anin vitro model of angiogenesis. Am JPathol. 2000, 156, 1887–1900.Shimkets, R. A., Lowe, D. G., Tai, J. T.,Sehl, P. et al. Gene expression analysisby transcript profiling coupled to a genedatabase query. Nat Biotechnol. 1999, 17,798–803.Figeys, D., Ducret, A., Yates, J. R. III,Aebersold, R. Protein identification bysolid phase microextraction-capillaryzone electrophoresis-microelectrospraytandem mass spectrometry.Nat.Biotechnol. 1996, 14, 1579–1583.1391419Genes Involved in Atherosclerosis and Plaque FormationTiina T.
Tuomisto and Seppo Ylä-Herttuala9.1IntroductionAtherosclerosis is the principal underlying cause of myocardial infarction, strokeand gangrene of extremities and the most common cause of death in Westerncountries. It affects large and medium-sized arteries. Atherosclerosis is inducedby several risk factors including hypertension, smoking, diabetes, elevated serumlow density lipoprotein (LDL) and very low density lipoprotein (VLDL) levels.
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
