IBR   13079
INSTITUTO DE BIOLOGIA MOLECULAR Y CELULAR DE ROSARIO
Unidad Ejecutora - UE
congresos y reuniones científicas
Título:
NMR-BASED METABOLIC PROFILING FOR THE DISCOVERY OF BIOMARKERS IN PATIENTS WITH PULMONARY TUBERCULOSIS (TB).
Autor/es:
BURDISSO PAULA; D´ATILIO LUCIANO; DIAZ ARIANA; BOTTASSO OSCAR; BAY MARÍA L.; VILA, ALEJANDRO; RASIA, RODOLFO M.
Lugar:
Santa Fe
Reunión:
Taller; III Taller de Resonancia Magnética Nuclear; 2016
Resumen:
Omics sciences have allowed the use of global data analysis to explain biological processes. Among them, metabolomics deals with the study of metabolites, i.e small molecules of less than 1000 Da that represent the last product of celular processes reflecting the identity and the influence of the environment on an organism. In recent years advances in NMR allowed the acquisition of data in an effective, reproducible, and high-throughput manner. The challenge remains to develop highly reproducible methods and standardized protocols that minimize experimental or technical bias, allowing real interlaboratory comparisons of biomarker information. With the aim of developping the field in our lab, we investigated the metabolic profile of plasma samples of TB patients with pulmonary TB and healty controls (HCo). TB is a major health problem worldwide, with 1.5 millions of death per year. The workflow, including sample preparation, NMR calibration, data acquisition and processing was set up based on the literature1;2 and on bruker metabolomics specifications. In the present work 74 samples and 8 quality control samples were analyzed, giving hight quality 1H-NMR spectra. The data was digitized and and imported into MATLAB (2015, Mathworks Inc., USA), and automatically phase- and baseline-corrected. Each spectrum was normalized to the median spectrum, using glucose signals as reference. Plasma profiling analysis in combination with Principal Component Analysis (PCA) and Partial Least Squares Discriminant Analysis was performed to identify candidate biomarkers of TB. Several discriminant features are reported and the identification is made combining data bases, spike in and 2D spectra. The technique was succesfully implemented providing hight quality spectra and good models for human samples representing the startup for future projects in clinical research.