Data Preprocessing: Quantification and Normalization of the Luminex Assay System

• Activity Number: 404
• Type: Contributed
• Date/Time: Tuesday, August 2, 2011 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Eileen Liao*+ and David Elashoff
• Companies: University of California at Los Angeles and University of California at Los Angeles
• Keywords: data preprocessing ; median normalization ; lowess curve ; quantile normalization

Abstract

High throughput genomics experiments generate exhaustive quantities of information. Microarray is one of the technologies that allow rapid analysis of molecular targets of thousands of genes at a time. In microarray experiments, variations in expression measurements emerge from many sources. Despite the huge improvements to the technology, variations still exist. We applied normalization methods to a bead-based multiplex Luminex assay system to reduce the plate-to-plate variability. We quantified performance among measurements of fluorescent intensity, fluorescent intensity minus background, and observed concentration in both high and standard scanning systems, and evaluated the preservation of ranking separation across plates for both standards and control plasma. We then applied scale normalization, lowess curve normalization, smoothing lowess curve extrapolation and quantile normalization to the data, and utilized coefficient of variation across plates to evaluate the performance of each method. Smoothing curve extrapolation is most efficient in reducing variation across plates. Our findings provide some guidance on the selection of normalization methods.