Comparing cDNA and oligonucleotide array data: concordance of gene expression across platforms for the NCI-60 cancer cells - PubMed (original) (raw)

Comparative Study

doi: 10.1186/gb-2003-4-12-r82. Epub 2003 Nov 25.

Kimberly J Bussey, Fuad G Gwadry, William Reinhold, Gregory Riddick, Sandra L Pelletier, Satoshi Nishizuka, Gergely Szakacs, Jean-Phillipe Annereau, Uma Shankavaram, Samir Lababidi, Lawrence H Smith, Michael M Gottesman, John N Weinstein

Affiliations

Comparative Study

Comparing cDNA and oligonucleotide array data: concordance of gene expression across platforms for the NCI-60 cancer cells

Jae K Lee et al. Genome Biol. 2003.

Abstract

Microarray gene-expression profiles are generally validated one gene at a time by real-time RT-PCR. We describe here a different approach based on simultaneous mutual validation of large numbers of genes using two different expression-profiling platforms. The result described here for the NCI-60 cancer cell lines is a consensus set of genes that give similar profiles on spotted cDNA arrays and Affymetrix oligonucleotide chips. Global concordance is parameterized by a 'correlation of correlations' coefficient.

PubMed Disclaimer

Figures

Figure 1

Figure 1

Histograms showing the distribution of correlation coefficients for UniGene-matched and UniGene-mismatched transcripts. Pearson correlation coefficients for (a) cDNA-array and oligoarray transcript pairs that map to the same UniGene cluster and for (b) pairs that map to different UniGene clusters. (c) Modeling of the correlation distribution in (a) in terms of a component (22%) based on (b) representing mistaken matchings and a component (78%) based on true matches. This was an eye-fit of the one parameter representing the proportions of the two populations of values. (d) The same as in (a) but with Spearman (non-parametric) correlation coefficients. (e) The same as in (b) but with Spearman correlation coefficients. (f) Distribution of Pearson correlations for UniGene-matched cDNA-oligo transcripts that have not been sequence-verified. (g) The same as in (e) but for sequence-verified transcripts.

Figure 2

Figure 2

Correlation filtering of UniGene matched oligo- and cDNA-array data. (a) Cumulative distributions of the Pearson correlation coefficient for various types of expression pattern pairings. (b) Pearson correlation coefficient and its 95% bootstrap confidence limits for UniGene-matched oligo and cDNA transcripts.

Figure 3

Figure 3

Global concordance between the oligo- and cDNA-array gene-expression databases after UniGene matching. (a) Correlation of correlations (_r_c) and number of genes remaining in the dataset as a function of the correlation cutoff value. For the original cDNA- and oligo-array sets before UniGene matching, _r_c for the cells was only 0.48. (b) Cluster trees (average linkage, correlation metric) for the 60 cell lines based on cDNA array and oligo databases after UniGene matching and correlation screening at a threshold value of r = 0.3 (which produced sets of 1,733 cDNA-array and 1,564 oligo-array transcripts). Most of the clusters are very similar to each other in the two trees, and the clusters for five tissues of origin were found almost identical: CNS (red), renal (green), melanoma (purple), leukemia (pink), and colon (blue).

Figure 4

Figure 4

Clustered image maps (CIMs) for co-clustering of the cDNA- and oligo-array expression patterns [4,11]. Cell-line origin abbreviations as in Figure 3. Each gene-expression pattern is designated as coming from the cDNA or oligo array set. (a) CIM for the combined set of 3,297 oligo and cDNA transcripts. Of the UniGene-matched cDNA-oligo pairs, 55.4% (827 out of 1,493) appeared side by side in the tree, and an additional 4.4% (65 out of 1,493) were separated by five or fewer locations. (b) Magnified view of the portion of the CIM occupied by melanoma genes.

References

    1. Boyd MR, Paull KD. Some practical considerations and applications of the National Cancer Institute in vitro anticancer drug discovery screen. Drug Dev Res. 1995;34:91–109.
    1. Paull KD, Shoemaker RH, Hodes L, Monks A, Scudiero DA, Rubinstein L, Plowman J, Boyd MR. Display and analysis of patterns of differential activity of drugs against human tumor cell lines: development of mean graph and COMPARE algorithm. J Natl Cancer Inst. 1989;81:1088–1092. doi: 10.1093/jnci/81.14.1088. - DOI - PubMed
    1. Weinstein JN, Kohn KW, Grever MR, Viswanadhan VN, Rubinstein LV, Monks AP, Scudiero DA, Welch L, Koutsoukos AD, Chiausa AJ, et al. Neural computing in cancer drug development: predicting mechanism of action. Science. 1992;258:447–451. - PubMed
    1. Weinstein JN, Myers TG, O'Connor PM, Friend SH, Fornace AJ, Jr, Kohn KW, Fojo T, Bates SE, Rubinstein LV, Anderson NL, et al. An information-intensive approach to the molecular pharmacology of cancer. Science. 1997;275:343–349. doi: 10.1126/science.275.5298.343. - DOI - PubMed
    1. Shi LM, Fan Y, Lee JK, Waltham M, Andrews DT, Scherf U, Paull KD, Weinstein JN. Mining and visualizing large anticancer drug databases. J Chem Inf Comput Sci. 2000;40:367–379. doi: 10.1021/ci990087b. - DOI - PubMed

Publication types

MeSH terms

LinkOut - more resources