Regression analysis based on metabolic data without prior normalization

Praktikum (Bachelor/Master)

Field: Metabolomics

Advisor: Wolfram Gronwald, Helena Zacharias

Courses preferred: Genomik und Bioinformatik I

Objective: Regression analyses based on biofluids such as plasma and urine usually require the prior application of appropriate normalization procedures to remove unwanted sample-to-sample variations. A newly “in house” developed algorithm shows promise to achieve these goals without prior normalization assuming that only linear shifts between data sets exists. In this exercise, the student is to systematically test this algorithm on various metabolic data sets and to compare it to existing methods. Test data sets were already acquired by means of NMR spectroscopy and hyphenated mass spectrometry. They were mostly obtained in context of various chronic and non-chronic kidney diseases. In case that tests proof to be successful wide spread application of the novel algorithm can be expected. During this exercise the student will gain knowledge in the statistical analysis of metabolic data and will also gain insight in the underlying biological background.

Data: Various metabolic data sets obtained in context of kidney diseases

First steps: Understand the basic principles of metabolomic data analysis and of the new regression method.

Questions: Is normalization free regression analysis from metabolic data possible? If yes, how do results compare to those obtained by traditional methods?

Start reading:

Hochrein J, Klein MS, Zacharias HU, Li J, Wijffels G, Schirra HJ, Spang R, Oefner PJ & Gronwald W (2012): Performance Evaluation of Algorithms for the Classification of Metabolic 1H-NMR Fingerprints. Journal of Proteome Research, 11(12):6242-6251.

Zacharias HU, Hochrein J, Klein MS, Samol C, Oefner PJ & Gronwald W (2013): Current Experimental, Bioinformatic and Statistical Methods used in NMR Based Metabolomics. Curr. Metabol., 1(3):253-268.

Kohl S, Klein MS, Hochrein J, Oefner PJ, Spang R & Gronwald W (2012): State-of-the art data normalization methods improve NMR-based metabolomic analysis. Metabolomics, 8(1):146-160.