paperKB
coga / coga-kb
Help
Sign in

Chunk #6 — Results — Comparison of expression measurements between genome-wide and custom expression arrays

Source
Population genomics of human gene expression.
Embedded
yes

Text

To maximize the power to detect genetic effects, it is important to establish that expression measurements are robust to experimental variables. We previously described expression data for 60 CEU cell lines (all of the CEU HapMap parents) generated with Illumina's low-density (~700 genes) custom arrays 18. A total of 539 probes on this custom array have identical sequence as probes on the genome-wide array, which allows direct comparison of expression signals across experiments. For probes where there was variable signal intensity across individuals, there is highly significant correlation between signals from the two experiments (Figure S2). Note that the RNA used in each experiment for a given individual was extracted from a different cell line batch. The RNA labelling, hybridization and normalization were also done independently. The high degree of correlation illustrates that the transcript level measurements are stable despite differential growth and treatment of cell lines and samples.