Assessment of data processing to improve reliability of microarray experiments using genomic DNA reference

Abstract

Background: Using genomic DNA as common reference in microarray experiments has recently been tested by different laboratories. Conflicting results have been reported with regard to the reliability of microarray results using this method. To explain it, we hypothesize that data processing is a critical element that impacts the data quality.

Results: Microarray experiments were performed in a gamma-proteobacterium Shewanella oneidensis. Pair-wise comparison of three experimental conditions was obtained either with two labeled cDNA samples co-hybridized to the same array, or by employing Shewanella genomic DNA as a standard reference. Various data processing techniques were exploited to reduce the amount of inconsistency between both methods and the results were assessed. We discovered that data quality was significantly improved by imposing the constraint of minimal number of replicates, logarithmic transformation and random error analyses.

Conclusion: These findings demonstrate that data processing significantly influences data quality, which provides an explanation for the conflicting evaluation in the literature. This work could serve as a guideline for microarray data analysis using genomic DNA as a standard reference.

Figure 1

Scatter plot represents the correlation of the results from type 1 and 2 experiments. Genes with expression changes in opposite categories (induction vs. repression) in both approaches are visualized as dots located in the 2nd and 4th quadrants and away from the origin. Pearson's correlation coefficients (r) are indicated in each panel.

Figure 2

Assessment of minimal number of replicates. Blue line: Fum/O₂; pink line: Fe/O₂; and red line: Fe/Fum. X-axis: minimal number of replicates. (A) Plot of r values with different minimal number of replicates. Y-axis: r values comparing the results from type 1 and 2 approaches. (B) Number of genes in opposite categories (induction vs. repression) with different minimal numbers of replicates. Y-axis: numbers of genes. (C) Total number of genes with different minimal numbers of replicates. Total number of genes was set to 100% when minimal number of replicates was 2, and the total number of genes at other minimal number of replicates was normalized accordingly. Y-axis: numbers of genes.

Figure 3

Assessment of logarithmic transformation. Black and gray columns represent proportional and additive models, respectively. (A) A histogram representing r values. (B) A histogram showing the number of genes in opposite categories.

Figure 4

Assessment of random error analyses. Column 1: small sample method; 2: Curve fit method; 3: common error method; and 4: no random error analyses. (A) A histogram representing r values. (B) A histogram showing the number of genes in opposite categories.