Tauber, S. (2008). Quality assessment, normalisation and mixed effects models for microarray data : application to a sweet potato study [Diploma Thesis, Technische Universität Wien]. reposiTUm. http://hdl.handle.net/20.500.12708/183798
Microarrays; Mixed Linear Models; Normalisation; Quality Assessment; Bioconductor
en
Abstract:
The analysis of a microarray experiment is a totally non-trivial thing to do. Although we are dealing with biological questions it must be clear that it is not only advisable but necessary that a statistician is consulted and closely worked with during the entire time the experiment is conducted. This is not an onesided relation: the statistician must be as well disposed to immerse himself deeply into the biological questions and above all into the biological reality which is very often far away from the ideal mathematical world.<br />Having a very complete and large dataset at disposal, this work wants to enlighten the biological aims of the experiment as well as describe sophisticated methods to analyze this dataset. The main objective is to demonstrate the extremely powerful machinery of mixed effects models applied to microarray data. Besides various functions for quality diagnostics and normalisation are presented to facilitate the discovery of commonly observed features or patterns due to errors during the technical procedure of microarray preparation. A comprehensive analysis workflow is presented and the advantages and disadvantages of all methods are discussed with the hope to provide a guideline for future analysis.