Automated 2D peak detection in gas chromatography-ion mobility spectrometry through persistent homology

Disentangling shared and unique variation in multiplatform hazelnut volatilomics using JIVE

In food science, volatile metabolites play a crucial role in determining sensory quality, acceptability and traceability. Fully characterizing the volatilome often requires combining multiple analytical techniques. However, reliably integrating the outcomes of these independent analyses to identify shared and unique information remains a significant challenge. In this paper, we illustrate how the multivariate Joint and Individual Variation Explained (JIVE) approach could be used to face this problem on a multiplatform VOC dataset obtained characterizing the volatilome of hazelnut pastes with GC-MS, PTR-ToF-MS and GC-IMS. While standardized data processing strategies were applied for GC-MS and PTR-ToF-MS, an automated pipeline was developed for GC-IMS to extract untargeted peak tables. The samples, representing three geographical origins, were collected during roasting to capture a wide range of intensities, offering a challenging case study for the proposed approach. The results showed that JIVE effectively separated the variability of each dataset into joint and individual components. A high-level comparison of the three analytical methods, based on variation decomposition and variable distribution, confirmed their complementarity. Additionally, identifying latent variables facilitated the visualization of analytical patterns - both shared and platform-specific - and the selection of related key variable trends, supporting the chemical interpretation of the results. This unsupervised data exploration strategy, based on JIVE, provides clearer interpretation of both shared and technique-specific insights. It supports an objective evaluation of the potential of a multiplatform analysis while offering guidance for selecting the most suitable analytical method in studies constrained to a single technique.

How Machine Learning and Gas Chromatography-Ion Mobility Spectrometry Form an Optimal Team for Benchtop Volatilomics

This invited feature article discusses the potential of gas chromatography-ion mobility spectrometry (GC-IMS) as a point-of-need alternative for volatilomics. Furthermore, the capabilities and versatility of machine learning (ML) (chemometric) techniques used in the framework of GC-IMS analysis are also discussed. Modern ML techniques allow for addressing advanced GC-IMS challenges to meet the demands of modern chromatographic research. We will demonstrate workflows based on available tools that can be used with a clear focus on open-source packages to ensure that every researcher can follow our feature article. In addition, we will provide insights and perspectives on the typical issues of the GC-IMS along with a discussion of the process necessary to obtain more reliable qualitative and quantitative analytical results.

Benchtop volatilomics supercharged: How machine learning based design of experiment helps optimizing untargeted GC-IMS gas phase metabolomics

Gas chromatography-ion mobility spectrometry (GC-IMS) plays a significant role in both targeted and non-targeted analyses. However, the non-linear behavior of IMS and its complex ion chemistry pose challenges for finding optimal experimental conditions using existing methodologies. To address these issues, integrating machine learning (ML) strategies offers a promising approach. In this study, we propose a hybrid strategy, combining design of experiment (DOE) and machine learning (ML) for optimizing GC-IMS conditions in non-targeted volatilomic/flavoromic analysis, with saffron volatiles as a case study. To begin, a rotatable circumscribed central composite design (CCD) is used to define five influential GC-IMS factors of sample amount, headspace temperature, incubation time, injection volume, and split ratio. Subsequently, two ML models are utilized: multiple linear regression (MLR) as a linear model and Bayesian regularized-artificial neural network (BR-ANN) as a nonlinear model. These models are employed to predict the response variables of total peak areas (PAs) and the number of detected peaks (PNs) in GC-IMS. The findings show that there is a direct correlation between the factors in GC-IMS and the PNs, suggesting that MLR is a suitable approach for building a model in this scenario. However, the PAs exhibit nonlinear behavior, suggesting that BR-ANN is better suitable to capture this complexity. Notably, Derringer's desirability function is utilized to integrate the PAs and PNs, and in this scenario, MLR demonstrates satisfactory performance in modeling the GC-IMS factors.