Concepts and Software Package for Efficient Quality Control in Targeted Metabolomics Studies – MeTaQuaC

Abstract

Targeted quantitative mass spectrometry metabolite profiling is the workhorse of metabolomics research. Robust and reproducible data is essential for confidence in analytical results and is particularly important with large-scale studies. Commercial kits are now available which use carefully calibrated and validated internal and external standards to provide such reliability. However, they are still subject to processing and technical errors in their use and should be subject to a laboratory's routine quality assurance and quality control measures to maintain confidence in the results. We discuss important systematic and random measurement errors when using these kits and suggest measures to detect and quantify them. We demonstrate how wider analysis of the entire dataset, alongside standard analyses of quality control samples can be used to identify outliers and quantify s ystematic trends in order to improve downstream analysis. Finally we present the MeTaQuaC software which implements the above concepts and methods for Biocrates kits and creates a comprehensive quality control report containing rich visualization and informative scores and summary statistics. Preliminary unsupervised multivariate analysis methods are also included to provide rapid insight into study variables and groups. MeTaQuaC is provided as an open source R package under a permissive MIT license and includes detailed user documentation.

MatrixQCvis: shiny-based interactive data quality exploration for omics data

“…Contrary to arrayQualityMetrics, which is built upon outdated visualization libraries, MatrixQCvis uses the shiny (Chang et al, 2021) framework and provides a high number of interactive visualizations to explore the data. MeTaQuaC (Kuhring et al, 2020), a recently published R package dedicated to metabolomics data analysis, has a special focus on the Biocrates data output and creates a static quality control report. On the other hand, MatrixQCvis is not restricted to a vendor-specific output and centers around interactive exploration of data quality.…”