A Robust Classifier for the Quality Control of Reverse Phase Protein Array

• Activity Number: 254
• Type: Contributed
• Date/Time: Monday, August 1, 2011 : 2:00 PM to 3:50 PM
• Sponsor: Section on Quality and Productivity
• Author(s): Zhenlin Ju*+ and Kevin Coombes and Wenbin Liu and Gordon Mills and Paul Reobuck and Siwak Doris
• Companies: The University of Texas MD Anderson Cancer Center and The University of Texas MD Anderson Cancer Center and The University of Texas MD Anderson Cancer Center and The University of Texas MD Anderson Cancer Center and The University of Texas MD Anderson Cancer Center and The University of Texas MD Anderson Cancer Center
• Keywords: RPPA ; Quality control ; Bioinformatics ; Proteomics

Abstract

High-throughput reverse phase protein array (RPPA) technology enables measurements of the protein expression levels of thousands of samples in parallel. However, sample lysate preparation, slide printing, hybridization and washing may create substantial variability in the quality of the RPPA protein expression data. Thus far, there has not been an algorithm available for RPPA data quality control (QC). Therefore, we have developed a novel classifier for the quality control of RPPA experiments using a logistic regression model (LRM). The outcome of the prediction model is the probability that a slide is of good slide over the probability that a slide is of poor slide, ranging from 0 to 1. We conclude that the proposed classifier outlined in this study can sufficiently distinguish good quality from poor quality data, and can be used to retain good quality slides so that the normalization schemes protein expression patterns, and advanced biological analyses will not be drastically impacted by erroneous measurements and systematic variations.