Missing values exist widely in mass-spectrometry (MS) based metabolomics data. Various methods have been applied for handling missing values, but the selection can significantly affect following data analyses. Typically, there are three types of missing values, missing not at random (MNAR), missing at random (MAR), and missing completely at random (MCAR). Our study comprehensively compared eight imputation methods (zero, half minimum (HM), mean, median, random forest (RF), singular value decomposition (SVD), k-nearest neighbors (kNN), and quantile regression imputation of left-censored data (QRILC)) for different types of missing values using four metabolomics datasets. Normalized root mean squared error (NRMSE) and NRMSE-based sum of ranks (SOR) were applied to evaluate imputation accuracy. Principal component analysis (PCA)/partial least squares (PLS)-Procrustes analysis were used to evaluate the overall sample distribution. Student’s t-test followed by correlation analysis was conducted to evaluate the effects on univariate statistics. Our findings demonstrated that RF performed the best for MCAR/MAR and QRILC was the favored one for left-censored MNAR. Finally, we proposed a comprehensive strategy and developed a public-accessible web-tool for the application of missing value imputation in metabolomics (https://metabolomics.cc.hawaii.edu/software/MetImp/).

The interactions between the gut microbiome and metabolome play an important role in human health and diseases. Current studies mainly apply statistical correlation analysis between the gut microbiome and all the identified metabolites to explore their relationship. However, it remains challenging to identify the specific metabolic functions of microbes without in vitro culture experiments for validation. Discriminating the microbial metabolites from others (e.g., host, food, or environment) and exploring their metabolic functions and correlations with microbiome specifically may improve the efficiency and accuracy of biomarker discovery. So far, there have been no such bioinformatics tools available. Herein, we developed MetOrigin, an interactive web server that discriminates metabolites originating from the microbiome, performs the origin‐based metabolic pathway enrichment analysis, and integrates the statistical correlations and biological relationships in the database using Sankey network visualization. MetOrigin not only enables the quick identification of microbial metabolites and their metabolic functions but also facilitates the discovery of specific bacterial species that are closely associated with metabolites statistically and biologically. MetOrigin is freely available at http://metorigin.met-bioinformatics.cn/.

Left-censored missing values commonly exist in targeted metabolomics datasets and can be considered as missing not at random (MNAR). Improper data processing procedures for missing values will cause adverse impacts on subsequent statistical analyses. However, few imputation methods have been developed and applied to the situation of MNAR in the field of metabolomics. Thus, a practical left-censored missing value imputation method is urgently needed. We developed an iterative Gibbs sampler based left-censored missing value imputation approach (GSimp). We compared GSimp with other three imputation methods on two real-world targeted metabolomics datasets and one simulation dataset using our imputation evaluation pipeline. The results show that GSimp outperforms other imputation methods in terms of imputation accuracy, observation distribution, univariate and multivariate analyses, and statistical sensitivity. Additionally, a parallel version of GSimp was developed for dealing with large scale metabolomics datasets. The R code for GSimp, evaluation pipeline, tutorial, real-world and simulated targeted metabolomics datasets are available at: https://github.com/WandeRum/GSimp. Missing values caused by the limit of detection/quantification (LOD/LOQ) were widely observed in mass spectrometry (MS)-based targeted metabolomics studies and could be recognized as missing not at random (MNAR). MNAR leads to biased parameter estimations and jeopardizes following statistical analyses in different aspects, such as distorting sample distribution, impairing statistical power, etc. Although a wide range of missing value imputation methods was developed for–omics studies, a limited number of methods was designed appropriately for the situation of MNAR currently. To alleviate problems caused by MNAR and to facilitate targeted metabolomics studies, we developed a Gibbs sampler based missing value imputation approach, called GSimp, which is public-accessible on GitHub. And we compared our method with existing approaches using an imputation evaluation pipeline on both of the real-world and simulated metabolomics datasets to demonstrate the superiority of our method from different perspectives.

Pig gastrointestinal tracts harbor a heterogeneous and dynamic ecosystem populated with trillions of microbes, enhancing the ability of the host to harvest energy from dietary carbohydrates and contributing to host adipogenesis and fatness. However, the microbial community structure and related mechanisms responsible for the differences between the fatty phenotypes and the lean phenotypes of the pigs remained to be comprehensively elucidated. Herein, we first found significant differences in microbial composition and potential functional capacity among different gut locations in Jinhua pigs with distinct fatness phenotypes. Second, we identified that Jinhua pigs with lower fatness exhibited higher levels of short‐chain fatty acids in the colon, highlighting their enhanced carbohydrate fermentation capacity. Third, we explored the differences in expressed carbohydrate‐active enzyme (CAZyme) in pigs, indicating their involvement in modulating fat storage. Notably, Clostridium butyricum might be a representative bacterial species from Jinhua pigs with lower fatness, and a significantly higher percentage of its genome was dedicated to CAZyme glycoside hydrolase family 13 (GH13). Finally, a subsequent mouse intervention study substantiated the beneficial effects of C. butyricum isolated from experimental pigs, suggesting that it may possess characteristics that promote the utilization of carbohydrates and hinder fat accumulation. Remarkably, when Jinhua pigs were administered C. butyricum, similar alterations in the gut microbiome and host fatness traits were observed, further supporting the potential role of C. butyricum in modulating fatness. Taken together, our findings reveal previously overlooked links between C. butyricum and CAZyme function, providing insight into the basic mechanisms that connect gut microbiome functions to host fatness.

First introduced in 2021, MetOrigin has quickly established itself as a powerful web server to distinguish microbial metabolites and identify the bacteria responsible for specific metabolic processes. Building on the growing understanding of the interplay between the microbiome and metabolome, and in response to user feedback, MetOrigin has undergone a significant upgrade to version 2.0. This enhanced version incorporates three new modules: (1) Quick search module that facilitates the rapid identification of bacteria associated with a particular metabolite; (2) Orthology analysis module that links metabolic enzyme genes with their corresponding bacteria; (3) Mediation analysis module that investigates potential causal relationships among bacteria, metabolites, and phenotypes, highlighting the mediating role of metabolites. Additionally, the backend MetOrigin database has been updated with the latest data from seven public databases (KEGG, HMDB, BIGG, ChEBI, FoodDB, Drugbank, and T3DB), with expanded coverage of 210,732 metabolites, each linked to its source organism. MetOrigin 2.0 is freely accessible at http://metorigin.met-bioinformatics.cn .