mzrtsim: Raw Data Simulation for Reproducible Gas/Liquid Chromatography–Mass Spectrometry-Based Nontargeted Metabolomics Data Analysis

Reproducibility of data analysis is pivotal in the context of nontargeted metabolomics based on mass spectrometry coupled with chromatography. While various algorithms have been proposed for feature or peak extraction, their validation often revolves around a limited set of known compounds or standards. While data simulation is widely used in other omics studies, simulations are focused on the feature level, neglecting uncertainties inherent in the feature or peak extraction process for metabolomics mass spectrometry data. In this technique note, we introduce an R package called "mzrtsim"' to simulate gas/liquid chromatography full scan raw data in the mzML format. Unlike simulations solely based on virtual features, our approach leverages experimental spectral data from MassBank of North America (MoNA) and the human metabolome database (HMDB). We developed algorithms to simulate chromatographic peaks, accounting for the tailing factor. The results of our study demonstrate the potential of this tool for comparing established metabolomics software (e.g., XCMS, mzMine, and OpenMS) against ground truth. We found that the investigated software introduced false positive peaks and/or loss of compounds with fewer peaks. They also showed different sensitivity to the tailing and leading peaks. This R package is free and available online (https://github.com/yufree/mzrtsim).