Title:Transformation of UPLC-MS Data Overcomes Extreme Variability in Urine Concentration and Metabolite Fold Change
Volume: 2
Issue: 2
Author(s): Monika Tokmina-Lukaszewska, Navid Movahed, Elizabeth R. Lusczek, Kristine E. Mulier, Greg J. Beilman and Brian Bothner
Affiliation:
Keywords:
Biomarkers, data preprocessing, hemorrhagic shock, mass spectrometry, untargeted metabolomics, urine.
Abstract: Acute blood loss causes changes across multiple levels of physiology, leading to pronounced systemic effects.
Clinical observations have shown that restoration of proper hemodynamic parameters and tissue perfusion with nutrients
and oxygen, frequently are not enough to prevent progression of patients into irreversible shock and eventually death.
Here we report the first application of ultra-performance liquid chromatography mass spectrometry (UPLC-MS) to the
analysis of urine during hemorrhagic shock. Samples were from a clinically relevant swine model system. A time course
study showed a high degree of biological variation between animals and across time, making direct quantitative comparisons
of low value. After applying probabilistic quotient normalization and logarithmic transformation, a group of approximately
200 molecular features with significant power to classify subjects was identified. A subsequent test of the
method on a second group of animals, demonstrated that the markers were robust at differentiating sham and shock
subjects as a group and as individuals, regardless of urine output and concentration. These results have direct relevance to
biomarker discovery for shock and the data transformation approach is applicable to a wide range of sample types.
