The Gut-Disease Connection: An Interactive Map of 867 Microbiome Associations

Your gut is home to trillions of microorganisms that do far more than digest food. Over the past decade, large-scale metagenomic studies have revealed that the composition of your gut microbiome is linked to conditions ranging from type 2 diabetes and colorectal cancer to depression, Parkinson’s, and autism. The challenge is that these findings are scattered across thousands of papers, each examining a handful of bacteria and a single condition.

We set out to change that. Instead of simply compiling what others have reported, we downloaded raw metagenomic data from two major repositories — the Wirbel et al. meta-analysis (1,892 shotgun metagenome samples across 14 studies) and MicrobiomeHD (28 case-control studies covering 10+ diseases) — and ran our own statistical analysis from scratch.

What We Did

We performed 12,147 Mann-Whitney U tests comparing bacterial species abundance between healthy controls and disease cohorts, applied Benjamini-Hochberg FDR correction, and identified 1,313 statistically significant associations (q < 0.05). We then combined these species-level results with genus-level cross-disease data from MicrobiomeHD to build a unified network of 867 unique bacteria-disease associations.

Explore the Network

Click any node to see its full association profile. Use the filters to focus on specific disease categories. Switch to “Key Findings” for our analysis summary, or “Disease Clusters” to see which diseases share the most microbial signatures.

Key Findings

1. Universal Gut Protectors

Bacteroides, Roseburia, Faecalibacterium, and Ruminococcus are consistently depleted across 11–12 of the 13 diseases we analyzed. These “hub bacteria” appear in our network as the most connected nodes, and they share a common function: producing butyrate, a short-chain fatty acid that feeds gut lining cells and suppresses inflammation. Their depletion may represent a universal signature of gut dysbiosis.

2. The IBD–Parkinson’s Connection

Our disease similarity analysis revealed that inflammatory bowel disease and Parkinson’s disease share 67.9% of their altered genera — 74 bacteria in common. This is the highest microbial overlap of any disease pair in our analysis, higher than IBD shares with other gut diseases. This finding adds microbiome-level evidence to the growing “gut-brain axis” hypothesis, which proposes that neurodegeneration may originate from chronic gut inflammation that propagates via the vagus nerve.

3. Cancer–Obesity Microbial Overlap

Colorectal cancer and obesity share 62.9% of their microbial signatures (66 genera). The link between obesity and cancer risk is well-established epidemiologically, but our analysis suggests a specific microbial mechanism: both conditions show depletion of the same anti-inflammatory genera (Coprococcus, Alistipes, Oscillibacter) and enrichment of pro-inflammatory species.

4. Depletion Over Enrichment

Of our 1,313 significant species-level associations, 65% are depletions — bacteria found at lower levels in disease versus healthy controls. Only 35% represent enrichments. This suggests that disease states are more often characterized by losing beneficial microbes than by gaining harmful ones, supporting the “dysbiosis as depletion” framework.

5. Autism Clusters with Neurological Diseases

When we cluster diseases by their shared microbial signatures (rather than clinical category), autism groups most closely with Parkinson’s disease (57.8% overlap) and IBD (57.1%) — not with other developmental conditions. This microbiome-based clustering suggests autism’s gut-related mechanisms may be more closely related to neuroinflammatory pathways than previously appreciated.

Methodology

Dataset 1 (Species-level): Shotgun metagenomic species profiles from the Wirbel et al. meta-analysis — 1,892 samples, 14 published studies, 7,728 species (1,704 with sufficient prevalence for testing). Disease groups: CRC (414), T2D (201), adenoma (117), UC (98), CD (63), IGT (49), PD (31), NAA (27). Healthy controls: 892. Statistical test: Mann-Whitney U (two-sided), FDR correction via Benjamini-Hochberg at q < 0.05.

Dataset 2 (Genus-level): MicrobiomeHD — 28 case-control 16S rRNA studies covering C. difficile infection, HIV, type 1 diabetes, autism, colorectal cancer, liver cirrhosis, IBD, obesity, Parkinson’s, NASH, and rheumatoid arthritis. 152 genera with effect sizes across 29 study cohorts.

Novel analyses: Disease similarity via Jaccard index on shared significantly-altered genera. Network centrality (degree) identifying hub bacteria and hub diseases. Unified association network from two independent data sources with different sequencing methods (shotgun vs. 16S), providing cross-validation.

Limitations

These are associations, not causal claims. Most studies use Western populations, limiting generalizability. Composition data doesn’t capture bacterial function — two samples with identical species may have different metabolic outputs. Effect sizes vary between studies due to differences in sequencing depth, DNA extraction methods, and cohort demographics.

Data Availability

All source data is publicly available. Wirbel et al. species profiles: Zenodo. MicrobiomeHD: GitHub. Our analysis code and processed results are available upon request.

References

1. Duvallet, C., Gibbons, S.M., Gurry, T., Irizarry, R.A. & Alm, E.J. (2017). Meta-analysis of gut microbiome studies identifies disease-specific and shared responses. Nature Communications, 8, 1784. doi:10.1038/s41467-017-01973-8
2. Wirbel, J., Essex, M., Forer, L. et al. (2024). Meta-analysis of the human gut microbiome uncovers shared and distinct microbial signatures between diseases. mSystems, 9(8). doi:10.1128/msystems.00295-24
3. Wirbel, J., Pyl, P.T., Kartal, E. et al. (2019). Meta-analysis of fecal metagenomes reveals global microbial signatures that are specific for colorectal cancer. Nature Medicine, 25, 679–689. doi:10.1038/s41591-019-0406-6
4. Romano, S., Savva, G.M., Bedber, J.R. et al. (2025). Comparative analysis of Parkinson’s and inflammatory bowel disease gut microbiomes reveals shared butyrate-producing bacteria depletion. npj Parkinson’s Disease, 11, 47. doi:10.1038/s41531-025-00894-4
5. Chen, H., Zhao, E.J., Zhang, W. et al. (2021). Meta-analysis of the Parkinson’s disease gut microbiome suggests alterations linked to intestinal inflammation. npj Parkinson’s Disease, 7, 27. doi:10.1038/s41531-021-00156-z
6. Langarica, R.E. & Kim, J. (2025). MINERVA—microbiome network research and visualization atlas: a scalable knowledge graph for mapping microbiome-disease associations. Briefings in Bioinformatics, 26(4). doi:10.1093/bib/bbaf472
7. Qin, J., Li, Y., Cai, Z. et al. (2012). A metagenome-wide association study of gut microbiota in type 2 diabetes. Nature, 490, 55–60. doi:10.1038/nature11450
8. Sokol, H., Pigneur, B., Watterlot, L. et al. (2008). Faecalibacterium prausnitzii is an anti-inflammatory commensal bacterium identified by gut microbiota analysis of Crohn disease patients. PNAS, 105(43), 16731–16736. doi:10.1073/pnas.0804812105