Mixed Effects Modeling of a Multi-Site Mass Spectrometry-Based Proteomics Study

• Activity Number: 612
• Type: Topic Contributed
• Date/Time: Thursday, August 4, 2011 : 8:30 AM to 10:20 AM
• Sponsor: ENAR
• Author(s): Jessie Q. Xia*+ and Nell Sedransk and Xingdong Feng
• Companies: National Institute of Statistical Sciences and National Institute of Statistical Sciences and National Institute of Statistical Sciences
• Keywords: mixed effects modeling ; proteomics ; MRM ; variance components analysis ; reproducibility ; mass spectrometry

Abstract

A multi-site mass spectrometry-based proteomics study was performed to quantify Multiple Reaction Monitoring (MRM) measurements of proteins spiked into human plasma. The design of this experimental series, while not completely balanced, still allows investigation of factor effects and interactions and multiple variance components. The series consisted of three different substudies with increasing complexity: from the simpliest case where an equimolar mixture of peptides digested from target proteins was prepared in pre-designed digested plasma, to the most complicated one where an equimolar mixture of target proteins was mixed with plasma, which was then digested in a single process. In all three substudies, an equimolar mixture of isotopically-labeled peptides was added into the samples to serve as internal standard. For each substudy, the dataset was analyzed using mixed effects modeling. Results show that measurements across multiple sites are consistent when these other factors are taken into account. Using the information provided by isotopic peptides gives a further reduction in the inter-site variability, documenting the value of using an internal standard.