Table 1:
Microbiome datasets with available classification tasks in ML Repo
| Project name | V Region | Target size | No. samples | No. subjects | Area | Description | Sequencing technology | Study design |
|---|---|---|---|---|---|---|---|---|
| Cho 2012 | V3 | 177 | 95 | 47 | Antibiotics | Mouse fecal and cecal samples, control vs 4 kinds of antibiotics | 454 | Cross-sectional |
| Claesson 2012 | V4 | 221 | 168 | 168 | Age | Elderly and young adults | 454 | Cross-sectional |
| David 2014 | V4 | 282 | 235 | 11 | Diet | Plant-based vs animal-based diet, cross-over study | Illumina MiSeq | Longitudinal |
| Gevers 2014 | V4 | 173 | 1,321 | 668 | IBD | Biopsies from patients with IBD prior to treatment | Illumina MiSeq | Cross-sectional |
| HMP 2012 | V35 | 527 | 6,407 | 242 | Body habitat, sex | Up to 18 body sites across 242 healthy subjects at 1–2 time points | 454 | Cross-sectional |
| Kostic 2012 | V35 | 569 | 190 | 95 | Colorectal cancer | Adjacent healthy vs tumor colon biopsy tissues | 454 | Paired |
| Montassier 2016 | V56 | 280 | 28 | 28 | Bacteremia | Patients prior to chemotherapy who did or did not develop bacteremia | 454 | Cross-sectional |
| Morgan 2012 | V35 | 569 | 231 | 231 | IBD | Healthy controls, patients with Crohn's disease or ulcerative colitis | 454 | Cross-sectional |
| Turnbaugh 2009 | V2 | 230 | 281 | 154 | Obesity | Monozygotic or dizygotic twin pairs concordant for body mass index class, and their mothers | 454 | Cross-sectional |
| Wu 2011 | V12 | 244 | 95 | 10 | Diet | Controlled high-fat or low-fat feeding on 10 subjects over 10 days | 454 | Longitudinal |
| Yatsunenko 2012 | V4 | 282 | 531 | 531 | Geography, age, sex | Humans of varying ages from the USA, Malawi, and Venezuela | Illumina MiSeq | Cross-sectional |
| Ravel 2011 | V12 | 240 | 396 | 396 | Bacterial vaginosis | Vaginal samples from 4 ethnic groups; Nugent scores for bacterial vaginosis | 454 | Cross-sectional |
| Karlsson 2013 | NA | NA | 144 | 144 | Diabetes | Patients with normal, impaired, or type 2 diabetes glucose tolerance categories | Illumina HiSeq | Cross-sectional |
| Qin 2012 | NA | NA | 134 | 134 | Diabetes | Chinese healthy controls vs patients with type 2 diabetes | Illumina HiSeq | Cross-sectional |
| Qin 2014 | NA | NA | 130 | 130 | Cirrhosis | Healthy controls vs patients with cirrhosis | Illumina HiSeq | Cross-sectional |
ML Repo contains 33 classification and regression tasks from 15 publicly available human microbiome datasets shown here. IBD: inflammatory bowel disease; NA: not applicable.