Comparison of RPPA Data Normalization Methods

• Activity Number: 188
• Type: Contributed
• Date/Time: Monday, August 1, 2011 : 10:30 AM to 12:20 PM
• Sponsor: ENAR
• Author(s): Wenbin Liu*+
• Companies: The University of Texas MD Anderson Cancer Center
• Keywords: RPPA ; normalization ; protein arrays

Abstract

The reverse phase protein array (RPPA) is a technology that measures protein expression in hundreds to thousands of samples at once. The EC50 of each dilution series of the sample lysate is usually estimated by fitting a nonlinear model. Our goal in this study is to compare some of the existing and potentially methods and some of the methods that we have proposed, including global median centering the samples (MCS), median centering the samples and then median centering the slides (MCSS1) and vice versa (MCSS2), median polish (MP), centering around total protein amount (measured by cloidal gold), variable slopes (VS) and z-score transforming each slide and then median centering each sample (ZS). Cell line data and simulated data were used. After normalizing the data by different methods, we compared the correlation between the replicate slides and the correlation between the replicate samples, the sample median ranks and the sample rank MADs, and biological validations using the cell line data and drug treatment information. The MCS and VS methods outperformed the other methods in precision and accuracy.