8. Statistical Analysis

  1. Ingvar Eidhammer,
  2. Harald Barsnes,
  3. Geir Egil Eide and
  4. Lennart Martens

Published Online: 10 JAN 2013

DOI: 10.1002/9781118494042.ch8

Computational and Statistical Methods for Protein Quantification by Mass Spectrometry

Computational and Statistical Methods for Protein Quantification by Mass Spectrometry

How to Cite

Eidhammer, I., Barsnes, H., Eide, G. E. and Martens, L. (2013) Statistical Analysis, in Computational and Statistical Methods for Protein Quantification by Mass Spectrometry, John Wiley & Sons Ltd, Oxford, UK. doi: 10.1002/9781118494042.ch8

Publication History

  1. Published Online: 10 JAN 2013
  2. Published Print: 4 JAN 2013

ISBN Information

Print ISBN: 9781119964001

Online ISBN: 9781118494042

SEARCH

Keywords:

  • hypothesis testing;
  • missing values;
  • quantitative proteomics;
  • replicates;
  • statistical analysis

Summary

A sound statistical analysis of the acquired data is a critical step in obtaining reliable results from protein quantification experiments. This chapter describes the statistics and statistical issues related to quantitative proteomics. When replicates are available for the two situations and the data is normally distributed in each situation the T-statistic should be used. The T-statistic is however fairly robust against minor deviations from the normality assumption. Missing values can occur at both the peptide and the protein level in protein quantification. Missing values occur when abundance for a given protein is determined for some of the replicates but not for all. The chapter discusses the predictions and hypothesis testing. In large scale proteomics the chapter considers the abundances of a large number of proteins, and often performs some sort of statistical test for each of these proteins, for instance a T-test.

Controlled Vocabulary Terms

hypothesis testing; Missing data; replication