NMR spectroscopy of urine samples provides accurate and fast detection of a large number of metabolites that are indicative of a variety of inherited metabolic diseases. This diagnostic may be extremely useful both for screening of asymptomatic, healthy children and for rapid diagnosis of sick children under urgent care. However, the interpretation of the NMR spectra remains a manual task requiring a highly skilled, medically trained spectroscopist. If the spectrum is complicated, manual interpretation time can far exceed sample preparation and data acquisition time. In Phase I, ORINCON proposes to design and implement a software package to automate the interpretation of NMR spectra of urine in terms of concentrations of abnormal metabolites. The analysis program will assign the absorption resonances to the metabolic compounds of interest, and the concentrations of metabolites will be obtained from spectral intensities. The identification of metabolites will be aided by a database of reference NMR spectra for a basic set of pure metabolites. It is anticipated that in Phase II of this program ORINCON will implement complete, user-friendly analysis software that will, with minimum intervention by the operator, interpret the NMR spectra in terms of the metabolite concentrations and the probabilities of metabolic disorders associated with the presence of these metabolites.
Software for automated analysis of NMR spectra will be a commercially viable product with application in newborn screening programs and infant intensive care units for the diagnosis of inherited metabolic diseases, as well as for development of other medical diagnostic tools based on NMR spectroscopy of urine, cerebrospinal fluid, plasma, bile, semen, and amniotic fluid. The algorithms developed here may also find commercial application in atomic and molecular spectroscopy.
