Combining multiple microarray studies using bootstrap meta-analysis

Abstract

Microarray technology has enabled us to simultaneously measure the expression of thousands of genes. Using this high-throughput technology, we can examine subtle genetic changes between biological samples and build predictive models for clinical applications. Although microarrays have dramatically increased the rate of data collection, sample size is still a major issue when selecting features. Previous methods show that combining multiple microarray datasets improves feature selection using simple methods such as fold change. We propose a wrapper-based gene selection technique that combines bootstrap estimated classification errors for individual genes across multiple datasets and reduces the contribution of datasets with high variance. We use the bootstrap because it is an unbiased estimator of classification error that is also effective for small sample data. Coupled with data combination across multiple datasets, we show that our metaanalytic approach improves the biological relevance of gene selection using prostate and renal cancer microarray data.

Left ventricular global transcriptional profiling in human end-stage dilated cardiomyopathy

“…While the sample sizes employed in this study are comparatively small, it should be noted that all tissues were procured from identical myocardial loci, to ensure optimal uniformity in cell content. This procedure should produce comparatively reliable and informative results with the microarray platform employed, in particular, if considered together with similar gene expression studies employing different platforms [46–48]. Most importantly, a validation analysis demonstrated great concordance of our results with other data sets using a different microarray platform.…”

Integrated Left Ventricular Global Transcriptome and Proteome Profiling in Human End-Stage Dilated Cardiomyopathy

“…Notably also, GO analyses of the DEGs and DEPs revealed significant overlap with enrichment of genes/proteins related to many processes including not only those that are likely to be altered in the progression of DCM to overt heart failure, but are also linked to other disease such as Alzheimer’s disease (AD) for example, which strengthens the notion that these two diseases might share some common pathway that need to be deciphered[ 15 ]. Therefore, our approach produced comparatively reliable and informative results with the array platforms employed, especially if considered together with similar gene expression by different platforms[ 47 – 49 ]. Hence, clinically useful inferences continue to be derived from such studies.…”