Microbiome Diversity Metrics (Alpha and Beta Diversity)

Alpha Diversity—Within-Sample#

Alpha diversity measures the diversity within a single sample. Three dimensions. Richness (how many species): Chao1, ACE, observed OTUs/ASVs.

Evenness (how evenly distributed): Pielou's J. Combined (richness + evenness): Shannon index (most commonly reported in this wiki), Simpson index (probability two random sequences are different species).

Clinical pattern: Reduced alpha diversity is the most consistent microbiome finding across disease states—IBD, CRC, depression, schizophrenia, ASD, obesity, CVD, and endometriosis all show lower Shannon indices compared to healthy controls.

Oral alpha diversity is inversely associated with breast cancer risk (OR 0.86 per SD).^{[1]Wu et al. 2022 — The Oral Microbiome and Breast Cancer in the Ghana Breast Health StudyZeni Wu, Doratha A. Byrd, Yunhu Wan et al. · 2022Open reference 1 ↓} However, reduced diversity is not always pathological—the healthy vaginal microbiome is naturally low-diversity (Lactobacillus-dominant), and high vaginal diversity indicates Dysbiosis.

Beta Diversity—Between-Sample#

Beta diversity measures how different two microbial communities are from each other. Bray-Curtis dissimilarity: Based on abundance differences (ignores phylogeny). UniFrac (weighted/unweighted): Incorporates phylogenetic distances between taxa.

Jaccard index: Presence/absence only (ignores abundance).

Visualized via PCoA, NMDS, or PERMANOVA. Disease vs. control groups typically show significant beta-diversity separation (PERMANOVA p < 0.05), indicating distinct community structures.

Limitations#

Alpha/beta diversity are summary statistics—they tell you the community is different but not how or why. Two communities with identical Shannon indices can have completely different species compositions.

16S rRNA Sequencing primer choice affects diversity estimates. Relative abundance data (standard in 16S) can inflate apparent diversity changes when a single dominant taxon shifts.