Novel Study: Gut Microbiota Shows Distinct City-Specification Due to Diet
2025-07-23
A new study shows that the human gut microbiota can pinpoint whether an adult lives in Wuhan or Shiyan, two cities 500 km apart in China’s Hubei Province, with 94 % accuracy. This microbial signature is strongly linked to each city’s characteristic diet.
The work is conducted by scientists at BGI Genomics’ Institute of Intelligent Medical Research (IIMR) and published in Frontiers in Microbiology this month. Although the groups were matched for age, BMI and smoking status, their intestinal ecosystems diverged in ways that mirror regional lifestyle and dietary preference. The findings of the study further guide precision medicine and raise the prospect of localized precision nutrition and therapeutic strategies.
Geography in the Gut
While previous research has shown that the gut microbiota reflects geographic patterns at the continental and provincial levels, it remains uncertain whether these patterns hold true between neighboring cities within the same province. To investigate this, they applied machine learning to determine whether microbiota data could reliably distinguish between these geographically close populations.
Image: Map of Hubei, purple colored regions indicating the two cities of Wuhan and Shiyan.
The researchers used shotgun metagenomics to identify 649 bacterial species and 515 metabolic pathways in participants’ gut microbiota. Although Wuhan and Shiyan are geographically close, their residents had noticeably different gut bacteria. In Wuhan, people’s microbiota were enriched with Bacteroides stercoris, while in Shiyan, Prevotella copri was more dominant.
Figure A. Top six phyla by mean relative abundance for participants in Hubei. B. The relative abundances (log10) of the core genera
There are 12 core bacterial genera, such as Faecalibacterium and Roseburia, were shared by more than 90% of all participants. This suggests a common gut composition across the population. However, the proportions of these bacteria varied depending on where participants lived.
Diet Shapes Microbial Composition
Wuhan sits amid abundant wetlands; freshwater fish, lotus root and leafy greens dominate dinner tables. Shiyan, wedged between Henan, Chongqing and Shaanxi, borrows the wheat-based, chili-laden preferences of its neighbours. These dietary contrasts appear to have etched themselves into the microbiota.
The single strongest predictor of residence in the study’s random-forest model, Bacteroides stercoris, was not the core species of the Chinese population but flourished in Wuhan participants. Conversely, Wuhan samples were depleted in Prevotella copri, a species often displaced by Mediterranean-style diets rich in fish and plant foods.
Wuhan’s microbiota uniquely harboured the metabolic pathway pyruvate fermentation to isobutanol (PWY-7111)—previously documented in herbivorous bats and in humans consuming plant-forward diets.
Ruminococcus faecis, another Wuhan-enriched driver, has been inversely associated with processed-meat intake, reinforcing the picture of a city that leans on vegetables and fish rather than cured pork.
The researchers also identified 36 bacterial taxa that were significantly more abundant in Wuhan, compared to just 10 in Shiyan. Notably, 63% of these species were location-specific microbiota, which essentially shape the distinct microbial “signatures” of each city.
Testing Utility with Machine Learning
The researchers developed machine-learning models to evaluate the practical utility of their findings. They used a random forest classifier that combined 16 bacterial species and 12 city-specific metabolic pathways. This combined model achieved a high level of accuracy, significantly outperforming models that used only species or only pathways.
The most powerful predictor was Bacteroides stercoris. However, six of the top ten predictors were metabolic pathways. This shows that a microbiota’s function is often more informative than its identity when distinguishing between geographic regions.
Microbiota and Health Markers
Fifteen bacterial species showed correlations with clinical health indicators. For example, Bifidobacterium longum was associated with healthier liver enzyme levels. Flavonifractor plautii was linked to higher LDL cholesterol, but only in Wuhan. This indicates that the same bacterial species may affect health differently depending on the local microbial environment. These results offer regional insights in interpreting the relationships between microbiota and health.
“This study challenges the long-standing belief that gut microbiota differences only matter at the scale of continents or provinces,” said Professor Li Tao from BGI Genomics’ IIMR, corresponding author of the study. "We found that even people living in neighboring cities have distinct microbial patterns."
Furthermore, the study's predictive model is highly accurate, laying the foundation for future precision medicine and personalized health management. The team revealed city-specific microbiota signatures and interaction networks by combining in-depth DNA sequencing with machine-learning tools. These findings help us understand how local environments quietly shape our inner biology, offering a foundation for more personalized healthcare, where diet and treatment could one day be tailored to where you live.
About BGI Genomics
BGI Genomics, headquartered in Shenzhen, China, is the world's leading integrated solutions provider of precision medicine. Our services cover more than 100 countries and regions, involving more than 2,300 medical institutions. In July 2017, as a subsidiary of BGI Group, BGI Genomics (300676.SZ) was officially listed on the Shenzhen Stock Exchange.
Read:
New Study Uncovers Brain Damage Progression in Alzheimer's Disease
Study: Metabolic Biomarkers Found in Obese Adolescents' Guide Early Vascular Aging Intervention
Study: Gut Bacteria-Producing L-Arginine Helps Protect Your Bone Health