Inferring Protein-Level Abundance from Peptide Peak Intensity Data to Facilitate Biological Interpretation

• Activity Number: 19
• Type: Topic Contributed
• Date/Time: Sunday, August 4, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistical Consulting
• Author(s): Bobbie-Jo Webb-Robertson*+
• Companies: Pacific Northwest National Laboratory
• Keywords: proteomics ; quantitation ; LC-MS

Abstract

Accurately modeling changes in global protein expression is essential to predict and characterize dynamic changes of biological systems. Thus, of late protein quantification has come to the forefront of computational proteomics research. One of the most effective global proteomic technologies is label-free liquid chromatography mass spectrometry (LC-MS), which is highly effective for the identification of peptides in complex biology samples. However, LC-MS does not directly yield protein quantities, they must be inferred from one or more measured peptides. The most commonly employed computational approaches to the estimation of protein abundances will be presented in the context of real LC-MS data and a dilutions study using a complex biological sample rather than a protein limited spike-in experiment. The dilution design, although contrived, more closely resembles a real-life clinical dataset, including biological variability in the observed response. We will also demonstrate the importance of peptide selection in the process of protein quantification and the necessity of this step in the identification of protein variation, which is essential to characterizing biological systems.