Addressing Missing Data in GC × GC Metabolomics: Identifying Missingness Type and Evaluating the Impact of Imputation Methods on Experimental Replication.

Missing data is a significant issue in metabolomics that is often neglected when conducting data preprocessing, particularly when it comes to imputation. This can have serious implications for downstream statistical analyses and lead to misleading or uninterpretable inferences. In this study, we aim to identify the primary types of missingness that affect untargeted metabolomics data and compare strategies for imputation using two real-world comprehensive two-dimensional gas chromatography (GC × GC) data sets. We also present these goals in the context of experimental replication whereby imputation is conducted in a within-replicate-based fashion─the first description and evaluation of this strategy─and introduce an R package MetabImpute to carry out these analyses. Our results conclude that, in these two GC × GC data sets, missingness was most likely of the missing at-random (MAR) and missing not-at-random (MNAR) types as opposed to missing completely at-random (MCAR). Gibbs sampler imputation and Random Forest gave the best results when imputing MAR and MNAR compared against single-value imputation (zero, minimum, mean, median, and half-minimum) and other more sophisticated approaches (Bayesian principal component analysis and quantile regression imputation for left-censored data). When samples are replicated, within-replicate imputation approaches led to an increase in the reproducibility of peak quantification compared to imputation that ignores replication, suggesting that imputing with respect to replication may preserve potentially important features in downstream analyses for biomarker discovery.

Spaceflight Analogue Culture Enhances the Host-Pathogen Interaction Between Salmonella and a 3-D Biomimetic Intestinal Co-Culture Model.

Physical forces associated with spaceflight and spaceflight analogue culture regulate a wide range of physiological responses by both bacterial and mammalian cells that can impact infection. However, our mechanistic understanding of how these environments regulate host-pathogen interactions in humans is poorly understood. Using a spaceflight analogue low fluid shear culture system, we investigated the effect of Low Shear Modeled Microgravity (LSMMG) culture on the colonization of Salmonella Typhimurium in a 3-D biomimetic model of human colonic epithelium containing macrophages. RNA-seq profiling of stationary phase wild type and Δhfq mutant bacteria alone indicated that LSMMG culture induced global changes in gene expression in both strains and that the RNA binding protein Hfq played a significant role in regulating the transcriptional response of the pathogen to LSMMG culture. However, a core set of genes important for adhesion, invasion, and motility were commonly induced in both strains. LSMMG culture enhanced the colonization (adherence, invasion and intracellular survival) of Salmonella in this advanced model of intestinal epithelium using a mechanism that was independent of Hfq. Although S. Typhimurium Δhfq mutants are normally defective for invasion when grown as conventional shaking cultures, LSMMG conditions unexpectedly enabled high levels of colonization by an isogenic Δhfq mutant. In response to infection with either the wild type or mutant, host cells upregulated transcripts involved in inflammation, tissue remodeling, and wound healing during intracellular survival. Interestingly, infection by the Δhfq mutant led to fewer transcriptional differences between LSMMG- and control-infected host cells relative to infection with the wild type strain. This is the first study to investigate the effect of LSMMG culture on the interaction between S. Typhimurium and a 3-D model of human intestinal tissue. These findings advance our understanding of how physical forces can impact the early stages of human enteric salmonellosis.

Pseudomonas aeruginosa Volatilome Characteristics and Adaptations in Chronic Cystic Fibrosis Lung Infections.

Pseudomonas aeruginosa chronic lung infections in individuals with cystic fibrosis (CF) significantly reduce quality of life and increase morbidity and mortality. Tracking these infections is critical for monitoring patient health and informing treatments. We are working toward the development of novel breath-based biomarkers to track chronic P. aeruginosa lung infections in situ Using comprehensive two-dimensional gas chromatography coupled with time-of-flight mass spectrometry (GC×GC-TOF-MS), we characterized the in vitro volatile metabolomes ("volatilomes") of 81 P. aeruginosa isolates collected from 17 CF patients over at least a 5-year period of their chronic lung infections. We detected 539 volatiles produced by the P. aeruginosa isolates, 69 of which were core volatiles that were highly conserved. We found that each early infection isolate has a unique volatilome, and as infection progresses, the volatilomes of isolates from the same patient become increasingly dissimilar, to the point that these intrapatient isolates are no more similar to one another than to isolates from other patients. We observed that the size and chemical diversity of P. aeruginosa volatilomes do not change over the course of chronic infections; however, the relative abundances of core hydrocarbons, alcohols, and aldehydes do change and are correlated with changes in phenotypes associated with chronic infections. This study indicates that it may be feasible to track P. aeruginosa chronic lung infections by measuring changes to the infection volatilome and lays the groundwork for exploring the translatability of this approach to direct measurement using patient breath.IMPORTANCEPseudomonas aeruginosa is a leading cause of chronic lung infections in cystic fibrosis (CF), which are correlated with lung function decline. Significant clinical efforts are therefore aimed at detecting infections and tracking them for phenotypic changes, such as mucoidy and antibiotic resistance. Both the detection and tracking of lung infections rely on sputum cultures, but due to improvements in CF therapies, sputum production is declining, although risks for lung infections persist. Therefore, we are working toward the development of breath-based diagnostics for CF lung infections. In this study, we characterized of the volatile metabolomes of 81 P. aeruginosa clinical isolates collected from 17 CF patients over a duration of at least 5 years of a chronic lung infection. We found that the volatilome of P. aeruginosa adapts over time and is correlated with infection phenotype changes, suggesting that it may be possible to track chronic CF lung infections with a breath test.

Longitudinal Associations of the Cystic Fibrosis Airway Microbiome and Volatile Metabolites: A Case Study.

The identification of 16S rDNA biomarkers from respiratory samples to describe the continuum of clinical disease states within persons having cystic fibrosis (CF) has remained elusive. We sought to combine 16S, metagenomics, and metabolomics data to describe multiple transitions between clinical disease states in 14 samples collected over a 12-month period in a single person with CF. We hypothesized that each clinical disease state would have a unique combination of bacterial genera and volatile metabolites as a potential signature that could be utilized as a biomarker of clinical disease state. Taxonomy identified by 16S sequencing corroborated clinical culture results, with the majority of the 109 PCR amplicons belonging to the bacteria grown in clinical cultures (Escherichia coli and Staphylococcus aureus). While alpha diversity measures fluctuated across disease states, no significant trends were present. Principle coordinates analysis showed that treatment samples trended toward a different community composition than baseline and exacerbation samples. This was driven by the phylum Bacteroidetes (less abundant in treatment, log2 fold difference -3.29, p = 0.015) and the genus Stenotrophomonas (more abundant in treatment, log2 fold difference 6.26, p = 0.003). Across all sputum samples, 466 distinct volatile metabolites were identified with total intensity varying across clinical disease state. Baseline and exacerbation samples were rather uniform in chemical composition and similar to one another, while treatment samples were highly variable and differed from the other two disease states. When utilizing a combination of the microbiome and metabolome data, we observed associations between samples dominated Staphylococcus and Escherichia and higher relative abundances of alcohols, while samples dominated by Achromobacter correlated with a metabolomics shift toward more oxidized volatiles. However, the microbiome and metabolome data were not tightly correlated; examining both the metagenomics and metabolomics allows for more context to examine changes across clinical disease states. In our study, combining the sputum microbiome and metabolome data revealed stability in the sputum composition through the first exacerbation and treatment episode, and into the second exacerbation. However, the second treatment ushered in a prolonged period of instability, which after three additional exacerbations and treatments culminated in a new lung microbiome and metabolome.

A Metabolomic Approach for Predicting Diurnal Changes in Cortisol.

Introduction: The dysregulation of cortisol secretion has been associated with a number of mental health and mood disorders. However, diagnostics for mental health and mood disorders are behavioral and lack biological contexts. Objectives: The goal of this work is to identify volatile metabolites capable of predicting changes in total urinary cortisol across the diurnal cycle for long-term stress monitoring in psychological disorders. Methods: We applied comprehensive two-dimensional gas chromatography coupled with time-of-flight mass spectrometry to sample the urinary volatile metabolome using an untargeted approach across three time points in a single day for 60 subjects. Results: The finalized multiple regression model includes 14 volatile metabolites and 7 interaction terms. A review of the selected metabolites suggests pyrrole, 6-methyl-5-hepten-2-one and 1-iodo-2-methylundecane may originate from endogenous metabolic mechanisms influenced by glucocorticoid signaling mechanisms. Conclusion: This analysis demonstrated the feasibility of using specific volatile metabolites for the prediction of secreted cortisol across time.

Monitoring changes in the healthy female metabolome across the menstrual cycle using GC × GC-TOFMS.

Urinary metabolomics offers a non-invasive means of obtaining information about the system-wide biological health of a patient. Untargeted metabolomics approaches using one-dimensional gas chromatography (GC) are limited due to the chemical complexity of urine, which poorly detects co-eluting low-abundance analytes. Metabolite detection and identification can be improved by applying comprehensive two-dimensional GC, allowing for the discovery of additional viable biomarkers of disease. In this work, we applied comprehensive two-dimensional GC coupled with time-of-flight mass spectrometry (GC × GC-TOFMS) to the analysis of urine samples collected daily across 28-days from 10 healthy female subjects for a personalized approach to female reproductive health monitoring. Through this analysis, we identified 935 unique volatile metabolites. Two statistical methods, a modified T-statistic and Wilcoxon Rank Sum, were applied to analyze differences in metabolome abundance on ovulation days as compared to non-ovulation days. Four metabolites (2-pentanone, 3-penten-2-one, carbon disulfide, acetone) were identified as statistically significant by the modified T-statistic but not the Rank Sum, after a false-discovery rate of 0.1 was set using a Benjamini-Hochberg correction. Subsequent analyses by boxplot indicated that the putative volatile metabolic biomarkers for fertility are expressed in increased or decreased abundance in urine on the day of ovulation. Individual analysis of metabolome expression across 28-days revealed some subject-specific features, which suggest a potential for long-term, personalized fertility monitoring using metabolomics.

Advances in the application of comprehensive two-dimensional gas chromatography in metabolomics.

Due to excellent separation capacity for complex mixtures of chemicals, comprehensive two-dimensional gas chromatography (GC × GC) is being utilized with increasing frequency for metabolomics analyses. This review describes recent advances in GC × GC method development for metabolomics, organismal sampling techniques compatible with GC × GC, metabolomic discoveries made using GC × GC, and recommendations and best practices for collecting and reporting GC × GC metabolomics data.