Assessment of reproducibility and robustness of UPLC-MS data for the metabolomic study of human serum

	Home
	Metabomeeting
	Conferences
	Viewpoint

Eva Zelena^a, Warwick Dunn^b, David Broadhurst^a, Sue McIntyre^a, Kathleen Carroll^b, Eleni Gika^c, Georgios Theodoridis^c, Antony Halsall^a, Ian D Wilson^c, Douglas B Kell^{a, b}

^aBioanalytical Sciences Group, Manchester Interdisciplinary Biocentre, University of Manchester, UK
^bManchester Centre for Integrative Systems Biology, Manchester Interdisciplinary Biocentre, University of Manchester, UK
^cAstraZeneca Pharmaceuticals, Alderley Park, UK

Metabolomics requires the acquisition of reproducible and robust data for the analysis of mammalian biofluids, including serum. Many metabolic investigations are performed over a short period of time; sample preparation and analysis taking just several days. The performance of analytical platforms appears to be reproducible over this period. However, over longer time-periods repeatability can be lower, gradual drift or sudden step changes can occur and influence the precision, accuracy and validity of results. A variety of parameters needs to be taken into account including the chemical complexity of samples and the wide variety of instrumental settings. One of the aims of the HUSERMET project (www.husermet.org) is to optimize and exploit the technologies available to characterise the metabolome of human serum. In this study we have employed multiple analytical platforms to enable the maximal coverage of the metabolome; these include Ultra Performance Liquid Chromatography-Mass Spectrometry (UPLC-MS), Gas Chromatography-Mass Spectrometry (GC-MS) and NMR. Assessment of the robustness and reproducibility of UPLC-MS data for long-term studies has been performed on defined sample sets to ensure the optimal performance of the instrument prior to high-throughput analysis of large sample numbers. In this presentation we will outline:

1) The comparison of the performance of UPLC-QToF-MS, UPLC-ToF-MS and UPLC-LTQ/Orbitrap instruments.

2) The evaluation of the performance (response and retention time reproducibility) of UPLC columns from three different production batches of stationary phase.

3) To explore the influence of the dilution of human serum on reproducibility of UPLC-MS data.