Pre-Processing Steps for Agilent MicroRNA Arrays: Does the Order Matter?

• Activity Number: 555
• Type: Contributed
• Date/Time: Wednesday, August 12, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Genomics and Genetics
• Author(s): Qin Zhou* and Rebecca Huang and Li-Xuan Qin
• Companies: Memorial Sloan Kettering Cancer Center and Columbia University and Memorial Sloan Kettering Cancer Center
• Keywords: microRNA ; pre-processing ; order ; normalization ; endometrioid cancer ; ovarian cancer

Abstract

In the microarray data analysis, data transformation, normalization and probe-set summarization are three necessary common pre-processing steps. The order of the three steps varies from paper to paper. The purpose of this study is to illustrate the possible effects of the processing ordering of these steps to the downstream analysis. A set of 192 samples which including 96 ovarian cancer samples and 96 endometrioid cancer samples are used. Two microRNA arrays are generated; one is handled by one run and one technician; the other mimics the regular array handling steps in the lab which exhibits the array effects. Moreover, the first one is using balance/blocking array design which ensures there is no/minimum confounding array effects. We take this one as the gold standard. In the meantime, varieties of orders of the pre-processing steps are applied to the second data. The conclusion is drawn by comparing the downstream analysis results versus the obtaining from the gold standard.