Table 2.
PD-associated genera identified via MWAS.
| PD-associated genera | MWAS significant In Dataset 1 | MWAS significant In Dataset 2 | Cluster | PubMed | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Phylum | Class | Order | Family | Genus | MRA | FC | ANCOM (W) | KW (FDR) | MRA | FC | ANCOM (W) | KW (FDR) | ||
| Bacteroidetes | Bacteroidia | Bacteroidales | Porphyromonadaceae | Porphyromonas | 0.001 | 4.20 | 406 | 1E − 03 | 0.001 | 2.94 | 468 | 2E − 02 | 1 | Opp path |
| Bacteroidetes | Bacteroidia | Bacteroidales | Prevotellaceae | Prevotella | 0.002 | 2.56 | 400 | 6E − 03 | 0.001 | 4.39 | 463 | 2E − 02 | 1 | Opp path |
| Actinobacteria | Actinobacteria | Corynebacteriales | Corynebacteriaceae | Corynebacterium_1 | 0.001 | 1.96 | 360 | 1E − 02 | 0.002 | 2.53 | 465 | 8E − 03 | 1 | Opp path |
| Firmicutes | Clostridia | Clostridiales | Ruminococcaceae | Faecalibacterium | 0.06 | 0.63 | 411 | 1E − 03 | 0.04 | 0.66 | 535 | 3E − 03 | 2 | SCFA |
| Firmicutes | Clostridia | Clostridiales | Lachnospiraceae | Agathobacter | 0.04 | 0.53 | 441 | 2E − 04 | 0.02 | 0.56 | 545 | 6E − 05 | 2 | SCFA |
| Firmicutes | Clostridia | Clostridiales | Lachnospiraceae | Blautia | 0.02 | 0.68 | 410 | 2E − 03 | 0.02 | 0.79 | 533 | 4E − 02 | 2 | SCFA |
| Firmicutes | Clostridia | Clostridiales | Lachnospiraceae | Roseburia | 0.02 | 0.48 | 391 | 4E − 03 | 0.01 | 0.60 | 541 | 3E − 04 | 2 | SCFA |
| Firmicutes | Clostridia | Clostridiales | Lachnospiraceae | Fusicatenibacter | 0.004 | 0.56 | 388 | 2E − 02 | 0.005 | 0.69 | 521 | 3E − 02 | 2 | SCFA |
| Firmicutes | Clostridia | Clostridiales | Lachnospiraceae | Lachnospira | 0.004 | 0.80 | 426 | 1E − 03 | 0.005 | 0.68 | 521 | 1E − 02 | 2 | SCFA |
| Firmicutes | Clostridia | Clostridiales | Ruminococcaceae | Butyricicoccus | 0.002 | 0.66 | 382 | 7E − 03 | 0.002 | 0.68 | 505 | 6E − 02 | 2 | SCFA |
| Firmicutes | Clostridia | Clostridiales | Lachnospiraceae | Lachnospiraceae_ND3007 | 0.001 | 0.37 | 418 | 2E − 04 | 0.001 | 0.59 | 538 | 6E − 04 | 2 | NC |
| Firmicutes | Clostridia | Clostridiales | Lachnospiraceae | Lachnospiraceae_UCG-004 | 0.001 | 0.48 | 384 | 2E − 02 | 0.001 | 0.38 | 544 | 1E − 05 | 2 | NC |
| Firmicutes | Clostridia | Clostridiales | Ruminococcaceae | Oscillospira | 0.0006 | 0.65 | 367 | 2E − 02 | 0.0005 | 0.64 | 525 | 1E − 02 | 2 | NC |
| Actinobacteria | Actinobacteria | Bifidobacteriales | Bifidobacteriaceae | Bifidobacterium | 0.01 | 1.83 | 410 | 1E − 03 | 0.01 | 2.72 | 553 | 6E − 07 | 3 | Probiotic |
| Firmicutes | Bacilli | Lactobacillales | Lactobacillaceae | Lactobacillus | 0.0004 | 6.61 | 407 | 2E − 04 | 0.004 | 1.57 | 458 | 1E − 02 | 3 | Probiotic |
MWAS was conducted in two datasets independently, testing differential abundance of genera in PD vs. controls, using two statistical methods (ANCOM and KW). The 15 genera shown are those that achieved microbiome-wide significance for association with PD in both datasets and by both methods, with (ANCOM) and without (KW) covariate adjustment (see “Methods” for covariates). Sample size: ANCOM included subset of subjects for whom complete data were available on all covariates tested: N = 171 cases and 117 controls in dataset 1 and 306 cases and 177 controls in dataset 2. KW included all subjects: N = 201 cases and 132 controls in dataset 1, and 323 cases and 184 controls in dataset 2. Clusters were identified hypothesis-free using correlation network analysis (Fig. 3). PubMed search was conducted after analyses were completed using genus and species name as search term (Supplementary Table 6). Function (opportunistic pathogen, SCFA, probiotic) was taken strictly from PubMed and is likely oversimplified. Microbiota have been studied under a narrow lens of what is already known about them. Opportunistic pathogens are often looked for in clinical specimen with infection, SCFA bacteria are studied intensively for their anti-inflammatory and other protective effects, and probiotics are understudied but highly advertised. The full function of the microbiota are not yet fully understood. In comparing results across published studies, note that a “genus” classified by one study may not be the same as the genus by the same name in another study. Taxonomic classifications and nomenclature are not standardized across reference databases, e.g., “Prevotella”, as annotated in some databases including NCBI, is further divided by SILVA (used here) into several non-monophyletic groups that SILVA calls, Prevotella_2, Prevotella_6, Prevotella_7, Prevotella_9, and Prevotella (see Discussion).
ANCOM analysis of composition of microbiomes. FC fold change in patients (MRA in patients/MRA in controls). FDR Benjamini–Hochberg false discovery rate (multiple testing corrected P-value). KW Kruskal–Wallis. MWAS microbiome-wide association study. MRA mean relative abundance in controls. NC not uncultured (uncharacterized). Opp path opportunistic pathogen (often commensal microorganism that can become pathogenic in immune-compromised individuals). Probiotic carbohydrate-metabolizing bacteria commonly known as probiotics. SCFA short-chain fatty acid-producing bacteria. W ANCOM score indicating the number of times a genus achieved FDR 0.05 as compared with other genera (maximum W possible: 444 in dataset 1, 560 in dataset 2, threshold 0.8 was used for significance, all shown genera were above significance threshold).