Suchergebnisse

BACKGROUND: Untargeted mass spectrometry (MS)-based metabolomics data often contain missing values that reduce statistical power and can introduce bias in biomedical studies. However, a systematic assessment of the various sources of missing values and strategies to handle these data has received little attention. Missing data can occur systematically, e.g. from run day-dependent effects due to limits of detection (LOD); or it can be random as, for instance, a consequence of sample preparation. METHODS: We investigated patterns of missing data in an MS-based metabolomics experiment of serum samples from the German KORA F4 cohort (n = 1750). We then evaluated 31 imputation methods in a simulation framework and biologically validated the results by applying all imputation approaches to real metabolomics data. We examined the ability of each method to reconstruct biochemical pathways from data-driven correlation networks, and the ability of the method to increase statistical power while preserving the strength of established metabolic quantitative trait loci. RESULTS: Run day-dependent LOD-based missing data accounts for most missing values in the metabolomics dataset. Although multiple imputation by chained equations performed well in many scenarios, it is computationally and statistically challenging. K-nearest neighbors (KNN) imputation on observations with variable pre-selection showed robust performance across all evaluation schemes and is computationally more tractable. CONCLUSION: Missing data in untargeted MS-based metabolomics data occur for various reasons. Based on our results, we recommend that KNN-based imputation is performed on observations with variable pre-selection since it showed robust results in all evaluation schemes. ; This work was supported by grants from the German Federal Ministry of Education and Research (BMBF), by BMBF Grant No. 01ZX1313C (project e:Athero-MED) and Grant No. 03IS2061B (project Gani_Med). Moreover, the research leading to these results has received funding from the European Union's Seventh Framework Programme [FP7-Health-F5-2012] under grant agreement No. 305280 (MIMOmics) and from the European Research Council (starting grant "LatentCauses"). KS is supported by Biomedical Research Program funds at Weill Cornell Medical College in Qatar, a program funded by the Qatar Foundation. The KORA Augsburg studies were financed by the Helmholtz Zentrum München, German Research Center for Environmental Health, Neuherberg, Germany and supported by grants from the German Federal Ministry of Education and Research (BMBF). Analyses in the EPIC-Norfolk study were supported by funding from the Medical Research Council (MC_PC_13048 and MC_UU_12015/1).

$\textbf{BACKGROUND}$: Higher circulating levels of the branched-chain amino acids (BCAAs; i.e., isoleucine, leucine, and valine) are strongly associated with higher type 2 diabetes risk, but it is not known whether this association is causal. We undertook large-scale human genetic analyses to address this question. $\textbf{METHODS AND FINDINGS}$: Genome-wide studies of BCAA levels in 16,596 individuals revealed five genomic regions associated at genome-wide levels of significance (p < 5 × 10-8). The strongest signal was 21 kb upstream of the PPM1K gene (beta in standard deviations [SDs] of leucine per allele = 0.08, p = 3.9 × 10-25), encoding an activator of the mitochondrial branched-chain alpha-ketoacid dehydrogenase (BCKD) responsible for the rate-limiting step in BCAA catabolism. In another analysis, in up to 47,877 cases of type 2 diabetes and 267,694 controls, a genetically predicted difference of 1 SD in amino acid level was associated with an odds ratio for type 2 diabetes of 1.44 (95% CI 1.26-1.65, p = 9.5 × 10-8) for isoleucine, 1.85 (95% CI 1.41-2.42, p = 7.3 × 10-6) for leucine, and 1.54 (95% CI 1.28-1.84, p = 4.2 × 10-6) for valine. Estimates were highly consistent with those from prospective observational studies of the association between BCAA levels and incident type 2 diabetes in a meta-analysis of 1,992 cases and 4,319 non-cases. Metabolome-wide association analyses of BCAA-raising alleles revealed high specificity to the BCAA pathway and an accumulation of metabolites upstream of branched-chain alpha-ketoacid oxidation, consistent with reduced BCKD activity. Limitations of this study are that, while the association of genetic variants appeared highly specific, the possibility of pleiotropic associations cannot be entirely excluded. Similar to other complex phenotypes, genetic scores used in the study captured a limited proportion of the heritability in BCAA levels. Therefore, it is possible that only some of the mechanisms that increase BCAA levels or affect BCAA metabolism are implicated in type 2 diabetes. $\textbf{CONCLUSIONS}$: Evidence from this large-scale human genetic and metabolomic study is consistent with a causal role of BCAA metabolism in the aetiology of type 2 diabetes. ; MRC Epidemiology Unit, Fenland study, EPIC-InterAct study, EPIC-Norfolk case-cohort study funding: this study was funded by the United Kingdom's Medical Research Council through grants MC_UU_12015/1, MC_UU_12015/5, MC_PC_13046, MC_PC_13048 and MR/L00002/1. We acknowledge support from the National Institute for Health Research Biomedical Research Centre. The research leading to these results has received support from the Innovative Medicines Initiative Joint Undertaking under EMIF grant agreement number 115372, resources of which are composed of financial contribution from the European Union's Seventh Framework Programme (FP7/2007-2013) and EFPIA companies' in kind contribution. EPIC-InterAct Study funding: funding for the InterAct project was provided by the EU FP6 programme (grant number LSHM_CT_2006_037197). MRC Human Nutrition Research funding: This research was supported by the Medical Research Council (MC_UP_A090_1006) and Cambridge Lipidomics Biomarker Research Initiative (G0800783). The SABRE study was funded at baseline by the UK Medical Research Council, Diabetes UK and the British Heart Foundation and at follow-up by a programme grant from the Wellcome Trust (WT082464) and British Heart Foundation (SP/07/001/23603); Diabetes UK funded the metabolomics analyses (13/0004774). RJOS, EN, JRZ and AK received funding from the Swedish Research Council, Stockholm County Council, Novo Nordisk Foundation and Diabetes Wellness. DBS is supported by the Wellcome Trust grant number 107064. MIM is a Wellcome Trust Senior Investigator and is supported by the following grants from the Wellcome Trust: 090532 and 098381. IB is supported by the Wellcome Trust grant WT098051.