Optimizing stringency for expression microarrays

James E. Korkola, Anne L.H. Estep, Sunanda Pejavar, Sandy DeVries, Ronald Jensen, Frederic M. Waldman

Research output: Contribution to journalArticlepeer-review

14 Scopus citations

Abstract

While several studies have reported methods to optimize expression microarray protoceols, none have dealt directly with hybridization wash stringency. We designed a series of experiments to determine the optimal stringency conditions for microarray experiments, using reproducibility and magnitudes of log2 (test/reference) ratio values as measures of quality. Low-stringency wash conditions of cell line hybridizations led to nonspecific binding, resulting in increased intensities, decreased magnitude of ratios, and poor reproducibility. Relatively high-stringency wash conditions were found to give the best reproducibility and large magnitude ratio changes, although increasing the stringency beyond this point led to lower magnitude ratios and poorer reproducibility. The expression levels of the ERBB2 oncogene in the BT474 versus MCF7 cell lines showed that high-stringency wash conditions gave the best agreement with real-time quantitative PCR, although the magnitude of the changes by microarray was smaller than for real-time quantitative PCR. Analysis of a series of cell lines washed at the optimized stringency indicated that the rank order of relative expression levels for ERBB2 microarray clones agreed well with the rank order of ERBB2 levels, as measured by quantitative PCR. These results indicate that the optimization of stringency conditions will improve microarray reproducibility and give more representative expression values.

Original languageEnglish (US)
Pages (from-to)828-835
Number of pages8
JournalBioTechniques
Volume35
Issue number4
DOIs
StatePublished - Oct 2003
Externally publishedYes

ASJC Scopus subject areas

  • Biotechnology
  • General Biochemistry, Genetics and Molecular Biology

Fingerprint

Dive into the research topics of 'Optimizing stringency for expression microarrays'. Together they form a unique fingerprint.

Cite this