Oligonucleotide probes of 34-44 nucleotides in length are designed by the bioinformatics group at Invitrogen or by ORB's bioinformatic's group in consultation with Ocimum Biosolutions. Design algorithims are based on the method of Goff et al (1) . Each oligonucleotide contains a direct repeat of a 17-22 nucleotide sequence complementary to a unique region within the microRNA target sequence. Probes are designed against the mature forms of microRNA as well as the minor (*) forms. Primary design criteria are specificity of each probe for a specific microRNA target and a uniform Tm (+/- 5° C) across the entire probe set. All probes are re-checked by blast algorithim for any significant non-target homology and any unavoidable hits are noted as secondary targets in the array annotation file.

Unmodified oligonucleotides are spotted in triplicate on Schott Nexterion substrates. Microarrays are always manufactured by high quality industrial suppliers such as Microarrays Inc. Quality control is conducted during manufacturing by machine vision technology, and post-printing by both nucleic acid staining and hybridization assays.

All incoming samples are evaluated by gel electrophoresis and U.V. spectrophotometry. Low molecular weight (LMW) RNA is purified from each sample by ultrafiltration and quantified by fluorometry. A set of 11 synthetic microRNAs are routinely added at 1/100,000 dilution to each LMW RNA samples to enable monitoring of sensitivity during labeling and hybridization. LMW RNA samples are labeled using Invitrogen's Rapid Labeling kit. This kit allows the attachment of 15 Alexa- 647 fluorophore molecules to the 3'-OH of each micro-RNA. For more information on the labeling system please refer to the "Flash-Tag" documentation on the Genisphere web site.

Hybridizations are conducted using low volume hybridization chambers under constant bubble mixing. As evidenced by the performance data below, ORB achieves excellent reproducibility, specificity, and sensitivity in the hybridization process. All hybridization and wash procedures are performed exactly as described in detailed standard operating procedures yielding excellent batch-to-batch uniformity.

Full data analysis is included as part of the per array service charge. Each array result is carefully inspected visually for any imperfections which may affect array quality. Following feature extraction, data quality from each array is subjected to a four point quality control test using intensity values from the control probe sets on each array. Results of these tests are provided to the customer. Spot intensities are log2-transformed, and normalized, and intensities from triplicate spots are averaged. Color-coded tables are provided with each result showing detection/ non-detection calls for each microRNA in every sample. The standard results package includes the results of ANOVA*, and principal component analysis*, as well as heat maps resulting from hierarchical clustering of the array data. We also provide custom comparisons and post-hoc statistical tests as directed by our clients.