Spotting and validation of a Genome wide oligonucleotide chip with duplicate measurement of each gene

High reproducibility of DNA microarray data relies on precise and uniform manufacturing of a microarray experiment. It is often stated that data from spotted microarray chips are subjected to a high degree of technical variation during the experimental procedure influencing the extraction of meaningful biological information. We have developed a spotted genome wide microarray chip with oligonucleotides printed in duplicate in order to minimise undesirable biases, thereby optimizing detection of true differential expression. The study design consisted of an assessment of the microarray chip performance using self hybridisation with the MessageAmp and Fairplay labeling kits. Statistical methods have been used to asses the experimental variation. Three main findings revealed the applicability of the system. The linear range of the chips was three orders of magnitude, the minimum detectable fold change for differential expression was below 1.3, and the coefficient of variation for differential expression was 13 % for MessageAmp and 15 % for Fairplay. Substantial reduction of variance was attained with duplicate spotting. Furthermore, relative quantitation was more reproducible than absolute quantitation. A method called Intraclass Correlation Coeffcient (ICC) was introduced to assess the agreement between the two kits. By bootstrapping the results, we found that MessageAmp was significantly more reproducible than Fairplay. An analysis of variance (ANOVA) demonstrated no significant day to day variation. The results showed reproducible performance of our in-house made oligonucleotide chip and that it can be used to detect differential expression even when only a small amount of biological material is available. Keywords: Two colour, self-hybridisation experiments Each of the two labeling kits FairPlay and MessageAmp was used in self-hybridisation experiments to test 11 chips. 11 different cDNA/ aRNA synthesis reactions were performed for each kit. 5 on day 1, 3 on day 2, and 3 on day 3. An aliquot of these reactions were used for Cy5 coupling while Cy3 were coupled to a pool of the 11 reactions (MessageAmp) and a pool of 11 new separate reactions (FairPlay). The reactions were performed on three different days and all the chips were from the same print batch.