Data normalization is an essential step in NMR-based metabolomics. Conducted properly, it improves data quality and removes unwanted biases. The choice of the appropriate normalization method is critical and depends on the inherent properties of the data set in question. In particular, the presence of unbalanced metabolic regulation, where the different specimens and cohorts under investigation do not contain approximately equal shares of up- and down-regulated features, may strongly influence data normalization. Here, we demonstrate the suitability of the Shapiro-Wilk test to detect such unbalanced regulation. Next, employing a Latin-square design consisting of eight metabolites spiked into a urine specimen at eight different known concentrations, we show that commonly used normalization and scaling methods fail to retrieve true metabolite concentrations in the presence of increasing amounts of glucose added to simulate unbalanced regulation. However, by learning the normalization parameters on a subset of nonregulated features only, Linear Baseline Normalization, Probabilistic Quotient Normalization, and Variance Stabilization Normalization were found to account well for different dilutions of the samples without distorting the true spike-in levels even in the presence of marked unbalanced metabolic regulation. Finally, the methods described were applied successfully to a real world example of unbalanced regulation, namely, a set of plasma specimens collected from patients with and without acute kidney injury after cardiac surgery with cardiopulmonary bypass use.

The high reliability of NMR spectroscopy makes it an ideal tool for large-scale metabolomic studies. However, the complexity of biofluids and, in particular, the presence of macromolecules poses a significant challenge. Ultrafiltration and protein precipitation are established means of deproteinization and recovery of free or total metabolite content, but neither is ever complete. In addition, aside from cost and labor, all deproteinization methods constitute an additional source of experimental variation. The Carr-Purcell-Meiboom-Gill (CPMG) echo-train acquisition of NMR spectra obviates the need for prior deproteinization by attenuating signals from macromolecules, but concentration values of metabolites measured in blood plasma will not necessarily reflect total or free metabolite content. Here, in contrast to approaches that propose the determination of individual T and T relaxation times for the computation of correction factors, we demonstrate their determination by spike-in experiments with known amounts of metabolites in pooled samples of the matrix of interest to facilitate the measurement of total metabolite content. Provided that the protein content does not vary too much among individual samples, accurate quantitation of metabolites is feasible. Moreover, samples with significantly deviating protein content may be readily recognized by inclusion of a standard that shows moderate protein binding. It is also shown that urinary proteins when present in high concentrations may effect detection of common urinary metabolites prone to strong protein binding such as tryptophan.

No abstract

Isocitrate dehydrogenase (IDH)-1 mutation is an important prognostic factor and a potential therapeutic target in glioma. Immunohistological and molecular diagnosis of IDH mutation status is invasive. To avoid tumor biopsy, dedicated spectroscopic techniques have been proposed to detect D-2-hydroxyglutarate (2-HG), the main metabolite of IDH, directly in vivo. However, these methods are technically challenging and not broadly available. Therefore, we explored the use of machine learning for the non-invasive, inexpensive and fast diagnosis of IDH status in standard 1H-magnetic resonance spectroscopy (1H-MRS). To this end, 30 of 34 consecutive patients with known or suspected glioma WHO grade II-IV were subjected to metabolic positron emission tomography (PET) imaging with O-(2-18F-fluoroethyl)-L-tyrosine (18F-FET) for optimized voxel placement in 1H-MRS. Routine 1H-magnetic resonance (1H-MR) spectra of tumor and contralateral healthy brain regions were acquired on a 3 Tesla magnetic resonance (3T-MR) scanner, prior to surgical tumor resection and molecular analysis of IDH status. Since 2-HG spectral signals were too overlapped for reliable discrimination of IDH mutated (IDHmut) and IDH wild-type (IDHwt) glioma, we used a nested cross-validation approach, whereby we trained a linear support vector machine (SVM) on the complete spectral information of the 1H-MRS data to predict IDH status. Using this approach, we predicted IDH status with an accuracy of 88.2%, a sensitivity of 95.5% (95% CI, 77.2–99.9%) and a specificity of 75.0% (95% CI, 42.9–94.5%), respectively. The area under the curve (AUC) amounted to 0.83. Subsequent ex vivo 1H-nuclear magnetic resonance (1H-NMR) measurements performed on metabolite extracts of resected tumor material (eight specimens) revealed myo-inositol (M-ins) and glycine (Gly) to be the major discriminators of IDH status. We conclude that our approach allows a reliable, non-invasive, fast and cost-effective prediction of IDH status in a standard clinical setting.

Identification of chronic kidney disease patients at risk of progressing to end-stage renal disease (ESRD) is essential for treatment decision-making and clinical trial design. Here, we explored whether proton nuclear magnetic resonance (NMR) spectroscopy of blood plasma improves the currently best performing kidney failure risk equation, the so-called Tangri score. Our study cohort comprised 4640 participants from the German Chronic Kidney Disease (GCKD) study, of whom 185 (3.99%) progressed over a mean observation time of 3.70 ± 0.88 years to ESRD requiring either dialysis or transplantation. The original four-variable Tangri risk equation yielded a C statistic of 0.863 (95% CI, 0.831−0.900). Upon inclusion of NMR features by state-of-the-art machine learning methods, the C statistic improved to 0.875 (95% CI, 0.850−0.911), thereby outperforming the Tangri score in 94 out of 100 subsampling rounds. Of the 24 NMR features included in the model, creatinine, high-density lipoprotein, valine, acetyl groups of glycoproteins, and Ca 2+ -EDTA carried the highest weights. In conclusion, proton NMR-based plasma fingerprinting improved markedly the detection of patients at risk of developing ESRD, thus enabling enhanced patient treatment.