Statistical Experimental Design for Mass Spectrometry Studies Using Labeling
Year: 2012
Type: article
Abstract: Mass Spectrometry (MS) utilizing labeling or multiplexing allows multiple specimens to be subjected to mass spectrometry simultaneously. As a result, between-experiment variability is reduced. Two common myths held regarding statistical experimental design of studies employing labeling are 1) that the sample size for a given study is restricted to the maximum number of specimens that can be labeled and combined for simultaneous mass analysis, and 2) that a reference design in which one label is always allocated to a common reference must be used. On the contrary, the statistical principals of randomization, balance and blocking, make it possible to design and carry out a labeled mass spectrometry experiment using multiple MS experiments without a reference sample. This ensures efficient use of resources and allows sample sizes to be as large as needed to achieve the desired power. In addition, it enables the use of statistical models to distinguish non-interesting experimental factors such as MS experiment or label from the biological effects of interest, and to properly estimate relative fold changes and quantify the associated statistical significance. We define these statistical principals and demonstrate how to practically implement them via case study using iTRAQ data.
Source: Europe PMC (PubMed Central)
Author: Ann L. Oberg
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W231974614