Figure 2.

Sequence-overlapping probes give greater cross-platform concordance for the NCI-60 panel. (A) Pearson correlation coefficient was calculated for each gene between its expression values measured on the Affymetrix Hu6800 platforms and its expression values measured on the Stanford cDNA microarray across sixty cell lines of the NCI-60 panel. The figure shows the cumulative distribution of the Pearson correlation coefficients for all genes analyzed. The five different curves reflect the level of cross-platform consistency of probe sets with various levels of overlap between the two microarray platforms. Matched gene measurements across the two platforms showed higher correlation when greater numbers of probes in the Affymetrix probe sets overlapped the insert region of the cDNA clone. The highest correlation was attained when only those Affymetrix probes overlapping the insert-sequence of a given cDNA clone were retained. Measurements for which the probes targeted the same transcript as the cDNA clone, but did not overlap the clone sequence, showed the lowest correlation. (B), Pearson correlation coefficient was calculated across all genes for each matched sample pair profiled by the Affymetrix Hu6800 platform and by the Stanford cDNA microarray. The figure shows the cumulative distribution of the Pearson correlation coefficients for the sixty cell lines of the NCI-60 panel. Matched cell-line measurements showed identical stratification of correlation levels by feature-matching criteria.

Carter et al. BMC Bioinformatics 2005 6:107   doi:10.1186/1471-2105-6-107
Download authors' original image