Metabolomics studies have seen a steady growth due to the development and implementation of affordable and high-quality metabolomics platforms. In large metabolite panels, measurement values are... Show moreMetabolomics studies have seen a steady growth due to the development and implementation of affordable and high-quality metabolomics platforms. In large metabolite panels, measurement values are frequently missing and, if neglected or sub-optimally imputed, can cause biased study results. We provided a publicly available, user-friendly R script to streamline the imputation of missing endogenous, unannotated, and xenobiotic metabolites. We evaluated the multivariate imputation by chained equations (MICE) and k-nearest neighbors (kNN) analyses implemented in our script by simulations using measured metabolites data from the Netherlands Epidemiology of Obesity (NEO) study (n = 599). We simulated missing values in four unique metabolites from different pathways with different correlation structures in three sample sizes (599, 150, 50) with three missing percentages (15%, 30%, 60%), and using two missing mechanisms (completely at random and not at random). Based on the simulations, we found that for MICE, larger sample size was the primary factor decreasing bias and error. For kNN, the primary factor reducing bias and error was the metabolite correlation with its predictor metabolites. MICE provided consistently higher performance measures particularly for larger datasets (n > 50). In conclusion, we presented an imputation workflow in a publicly available R script to impute untargeted metabolomics data. Our simulations provided insight into the effects of sample size, percentage missing, and correlation structure on the accuracy of the two imputation methods. Show less
Chouvarine, P.; Giera, M.; Kastenmuller, G.; Artati, A.; Adamski, J.; Bertram, H.; Hansmann, G. 2020
Objective While metabolic dysfunction occurs in several pulmonary arterial hypertension (PAH) animal models, its role in the human hypertensive right ventricle (RV) and lung is not well... Show moreObjective While metabolic dysfunction occurs in several pulmonary arterial hypertension (PAH) animal models, its role in the human hypertensive right ventricle (RV) and lung is not well characterised. We investigated whether circulating metabolite concentrations differ across the hypertensive RV and/or the pulmonary circulation, and correlate with invasive haemodynamic/echocardiographic variables in patients with PAH.Methods Prospective EDTA blood collection during cardiac catheterisation from the superior vena cava (SVC), pulmonary artery (PA) and ascending aorta (AAO) in children with PAH (no shunt) and non-PAH controls (Con), followed by unbiased screens of 427 metabolites and 836 lipid species and fatty acids (FAs) in blood plasma (Metabolon and Lipidyzer platforms). Metabolite concentrations were correlated with echocardiographic and invasive haemodynamic variables.Results Metabolomics/lipidomics analysis of differential concentrations (false discovery rate<0.15) revealed several metabolite gradients in the trans-RV (PA vs SVC) setting. Notably, dicarboxylic acids (eg, octadecanedioate: fold change (FC)_Control=0.77, FC_PAH=1.09, p value=0.044) and acylcarnitines (eg, stearoylcarnitine: FC_Control=0.74, FC_PAH=1.21, p value=0.058). Differentially regulated metabolites were also found in the transpulmonary (AAO vs PA) setting and between-group comparisons, that is, in the SVC (PAH-S VC vs Con-S VC), PA and AAO. Importantly, the differential PAH-metabolite concentrations correlated with numerous outcome-relevant variables (e.g., tricuspid annular plane systolic excursion, pulmonary vascular resistance).Conclusions In PAH, trans-RV and transpulmonary metabolite gradients exist and correlate with haemodynamic determinants of clinical outcome. The most pronounced differential trans-R V gradients are known to be involved in lipid metabolism/lipotoxicity, that is, accumulation of long chain FAs. The identified accumulation of dicarboxylic acids and acylcarnitines likely indicates impaired beta-oxidation in the hypertensive RV and represents emerging biomarkers and therapeutic targets in PAH. Show less
Deelen, J.; Kettunen, J.; Fischer, K.; Spek, A. van der; Trompet, S.; Kastenmuller, G.; ... ; Slagboom, P.E. 2019
Kit-based assays, such as AbsoluteIDQ(TM) p150, are widely used in large cohort studies and provide a standardized method to quantify blood concentrations of phosphatidylcholines (PCs). Many... Show moreKit-based assays, such as AbsoluteIDQ(TM) p150, are widely used in large cohort studies and provide a standardized method to quantify blood concentrations of phosphatidylcholines (PCs). Many disease-relevant associations of PCs were reported using this method. However, their interpretation is hampered by lack of functionally-relevant information on the detailed fatty acid side-chain compositions as only the total number of carbon atoms and double bonds is identified by the kit. To enable more substantiated interpretations, we characterized these PC sums using the side-chain resolving Lipidyzer(TM) platform, analyzing 223 samples in parallel to the AbsoluteIDQ(TM). Combining these datasets, we estimated the quantitative composition of PC sums and subsequently tested their replication in an independent cohort. We identified major constituents of 28 PC sums, revealing also various unexpected compositions. As an example, PC 16:0_22:5 accounted for more than 50% of the PC sum with in total 38 carbon atoms and 5 double bonds (PC aa 38:5). For 13 PC sums, we found relatively high abundances of odd-chain fatty acids. In conclusion, our study provides insights in PC compositions in human plasma, facilitating interpretation of existing epidemiological data sets and potentially enabling imputation of PC compositions for future meta-analyses of lipidomics data. Show less