Use and adaptation of bioinformatic tools for the metabolomic analysis of urine by 1H-NMR

Identifier:  TFG:2944

Handle: https://hdl.handle.net/20.500.11797/TFG2944

Authors:  Hernández Cacho, Adrián

Abstract:
Metabolomics is the science that studies metabolites present in cells, tissues, or even entire organisms. Knowing the metabolic state of a biological system allows us to know if it works normally or if it suffers from any abnormality. To know the metabolome of a subject, it is necessary to be able to access it through a biological matrix, such as urine. Urine has always been an important source of medical information regarding a possible disease state in different organs or tissues of a subject. As a residual fluid, it contains a large amount of metabolites from different metabolic pathways. In addition, the sample collection process is simple and non-invasive, requiring less pre-treatment compared to other biofluids, making it an attractive matrix for metabolomic studies. By means of Nuclear Magnetic Resonance (NMR), we can identify and quantify the metabolites present in urine, being Proton Nuclear Magnetic Resonance (1H-NMR) a robust technique for such studies due to its high performance, reproducibility and easy manipulation. On the other hand, the quantification of low molecular weight metabolites by NMR is a process that is difficult to automate for urine samples due to its high inter-individual variability. This work aims to automate the quantification of low molecular weight metabolites in urine samples using 1H-NMR by adapting a Biosfer Teslab software optimized to carry out the same function in blood samples. First, the metabolites present in the urine samples were identified. To do this, TopSpin software was used. Later, it was studied which method of signal referencing is the ideal to be able to automatically quantify metabolites. The methods of global referencing and local referencing of signals were compared, the latter was determined to be the most suitable for urine samples as it was more precise. Subsequently, the referenced signals were deconvolved by adapting the Biosfer Teslab software for each signal. The signal areas were obtained, and the concentration of the metabolites was quantified.