Chemometric analysis of urine fingerprints acquired by liquid chromatography-mass spectrometry and capillary electrophoresis: application to the schistosomiasis mouse model

Abstract

Urine fingerprints from Schistosoma mansoni infected and control animals were acquired with ultra performance liquid chromatography-MS (UPLC-MS) and compared with the urine fingerprints obtained by CE by applying the same set of multivariate analysis tools. Principal component analysis of the aligned data provided a time trajectory where the infection was observed after 30 days with UPLC-MS and CE. Two main markers describing infected and control, respectively - phenyl acetyl glycine (PAG) and hippurate - were selected to illustrate the use of orthogonal partial least-square discriminant analysis in determining the discriminatory confidence. PAG was found to be significantly related to the disease (high covariance and correlation), whereas hippurate was found to be nonsignificant as an indicator. Orthogonal partial least-square discriminant analysis models were validated for sensitivity and specificity. Multivariate data analysis derived from two different detection systems showed that CE-UV and UPLC-MS found equivalent results. This work gives additional mechanistic insight into the progress of the S. mansoni infection; the biochemical role and specificity of PAG as a biomarker is yet to be determined.