Dataset Information

Credentialing features: a platform to benchmark and optimize untargeted metabolomic methods.

ABSTRACT: The aim of untargeted metabolomics is to profile as many metabolites as possible, yet a major challenge is comparing experimental method performance on the basis of metabolome coverage. To date, most published approaches have compared experimental methods by counting the total number of features detected. Due to artifactual interference, however, this number is highly variable and therefore is a poor metric for comparing metabolomic methods. Here we introduce an alternative approach to benchmarking metabolome coverage which relies on mixed Escherichia coli extracts from cells cultured in regular and (13)C-enriched media. After mass spectrometry-based metabolomic analysis of these extracts, we "credential" features arising from E. coli metabolites on the basis of isotope spacing and intensity. This credentialing platform enables us to accurately compare the number of nonartifactual features yielded by different experimental approaches. We highlight the value of our platform by reoptimizing a published untargeted metabolomic method for XCMS data processing. Compared to the published parameters, the new XCMS parameters decrease the total number of features by 15% (a reduction in noise features) while increasing the number of true metabolites detected and grouped by 20%. Our credentialing platform relies on easily generated E. coli samples and a simple software algorithm that is freely available on our laboratory Web site (http://pattilab.wustl.edu/software/credential/). We have validated the credentialing platform with reversed-phase and hydrophilic interaction liquid chromatography as well as Agilent, Thermo Scientific, AB SCIEX, and LECO mass spectrometers. Thus, the credentialing platform can readily be applied by any laboratory to optimize their untargeted metabolomic pipeline for metabolite extraction, chromatographic separation, mass spectrometric detection, and bioinformatic processing.

SUBMITTER: Mahieu NG

PROVIDER: S-EPMC4188275 | biostudies-literature | 2014 Oct

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Credentialing features: a platform to benchmark and optimize untargeted metabolomic methods.

Mahieu Nathaniel Guy NG Huang Xiaojing X Chen Ying-Jr YJ Patti Gary J GJ

Analytical chemistry 20140922 19

The aim of untargeted metabolomics is to profile as many metabolites as possible, yet a major challenge is comparing experimental method performance on the basis of metabolome coverage. To date, most published approaches have compared experimental methods by counting the total number of features detected. Due to artifactual interference, however, this number is highly variable and therefore is a poor metric for comparing metabolomic methods. Here we introduce an alternative approach to benchmark ...[more]

PMID: 25160088

Similar Datasets

Project description:Untargeted metabolomics is a powerful phenotyping tool for better understanding biological mechanisms involved in human pathology development and identifying early predictive biomarkers. This approach, based on multiple analytical platforms, such as mass spectrometry (MS), chemometrics and bioinformatics, generates massive and complex data that need appropriate analyses to extract the biologically meaningful information. Despite various tools available, it is still a challenge to handle such large and noisy datasets with limited number of individuals without risking overfitting. Moreover, when the objective is focused on the identification of early predictive markers of clinical outcome, few years before occurrence, it becomes essential to use the appropriate algorithms and workflow to be able to discover subtle effects among this large amount of data. In this context, this work consists in studying a workflow describing the general feature selection process, using knowledge discovery and data mining methodologies to propose advanced solutions for predictive biomarker discovery. The strategy was focused on evaluating a combination of numeric-symbolic approaches for feature selection with the objective of obtaining the best combination of metabolites producing an effective and accurate predictive model. Relying first on numerical approaches, and especially on machine learning methods (SVM-RFE, RF, RF-RFE) and on univariate statistical analyses (ANOVA), a comparative study was performed on an original metabolomic dataset and reduced subsets. As resampling method, LOOCV was applied to minimize the risk of overfitting. The best k-features obtained with different scores of importance from the combination of these different approaches were compared and allowed determining the variable stabilities using Formal Concept Analysis. The results revealed the interest of RF-Gini combined with ANOVA for feature selection as these two complementary methods allowed selecting the 48 best candidates for prediction. Using linear logistic regression on this reduced dataset enabled us to obtain the best performances in terms of prediction accuracy and number of false positive with a model including 5 top variables. Therefore, these results highlighted the interest of feature selection methods and the importance of working on reduced datasets for the identification of predictive biomarkers issued from untargeted metabolomics data.

Dataset Information

Credentialing features: a platform to benchmark and optimize untargeted metabolomic methods.

Publications

Credentialing features: a platform to benchmark and optimize untargeted metabolomic methods.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure