| Title | Proposal of a health gut microbiome index based on a meta-analysis of Korean and global population datasets |
|---|---|
| Author | Hyun-Seok Oh1,2, Uigi Min1, Hyejin Jang1, Namil Kim1, Jeongmin Lim1, Mauricio Chalita1, and Jongsik Chun1,2,3* |
| Address | 1ChunLab Inc., Seoul 06194, Republic of Korea, 2Interdisciplinary Program in Bioinformatics, Seoul National University, Seoul 08826, Republic of Korea, 3School of Biological Sciences, Seoul National University, Seoul 08826, Republic of Korea |
| Bibliography | Journal of Microbiology, 60(5),533–549, 2022, |
| DOI | 10.1007/s12275-022-1526-0 |
| Key Words | gut microbiome, dysbiosis, enterotype, meta-analysis, microbiome index, healthy microbiome, microbial diversity, population study |
| Abstract | The disruption of the human gut microbiota has been linked to host health conditions, including various diseases. However, no reliable index for measuring and predicting a healthy microbiome is currently available. Here, the sequencing data of 1,663 Koreans were obtained from three independent studies. Furthermore, we pooled 3,490 samples from public databases and analyzed a total of 5,153 fecal samples. First, we analyzed Korean gut microbiome covariates to determine the influence of lifestyle on variation in the gut microbiota. Next, patterns of microbiota variations across geographical locations and disease statuses were confirmed using a global cohort and disease data. Based on comprehensive comparative analysis, we were able to define three enterotypes among Korean cohorts, namely, Prevotella type, Bacteroides type, and outlier type. By a thorough categorization of dysbiosis and the evaluation of microbial characteristics using multiple datasets, we identified a wide spectrum of accuracy levels in classifying health and disease states. Using the observed microbiome patterns, we devised an index named the gut microbiome index (GMI) that could consistently predict health conditions from human gut microbiome data. Compared to ecological metrics, the microbial marker index, and machine learning approaches, GMI distinguished between healthy and non-healthy individuals with a higher accuracy across various datasets. Thus, this study proposes a potential index to measure health status of gut microbiome that is verified from multiethnic data of various diseases, and we expect this model to facilitate further clinical application of gut microbiota data in future. |