t 0 hH′ 0pH 0SCFA 0 mMMissions 0/3FPS 0

Drag to orbit, wheel or pinch to zoom. Scale is NOT true: a real bacterium is ~1–5 µm, a villus ~0.5–1 mm. Colours: blue fibre fermenter, green fast sugar user, orange acid-tolerant, teal opportunist, purple lactate converter, violet cells = immune cells, amber = metabolites.

Gut Neighbourhood: A Microbial Street Beside the Villi

Five model taxa share a gut segment. Change what arrives (fibre, sugar, protein), how fast it flows and see who wins, who feeds whom, how pH moves and how the immune cells respond.

Educational model, not medical. Not a microbiome test, diagnosis, diet, probiotic or antibiotic advice. Taxa are generic, numbers are illustrative model values.

What arrives each hour (inflow, mmol/L)

Gut conditions

Community

Composition (share of biomass)

pH

SCFA and lactate (mmol/L)

Missions

Independent numerical checks

Controls

Drag / one finger: orbit; wheel / pinch: zoom. Hero / Street / Top move the camera.
Sliders change the diet or gut transit; the community responds over model hours. "Model hours per second" speeds time up. Antibiotic course kills sensitive taxa for 24 h.
Layers tint the villi by pH, SCFA or immune signal.

The science

Chemostat. Each taxon i grows on one resource S with Monod kinetics μ = μ_max S / (K_s + S) · f_pH; dN_i/dt = (μ_i − D − d_i) N_i + m (D = transit dilution, d = loss, m = tiny immigration from a reservoir). Resources: dS/dt = D(S_in − S) − Σ N_i μ_i / Y_i.

Competition (R* rule). On a single limiting resource the taxon with the lowest S* = K_s(D + d)/(μ_max − D − d) drives the resource below everyone else's break-even level and excludes them. Coexistence needs different resources, cross-feeding, or pH niches.

Cross-feeding. The fibre fermenter releases sugar and lactate; sugar users turn sugar into lactate and SCFA; the lactate converter lives on lactate and makes SCFA. A product of one is the food of the next.

pH. pH = 7.4 − 0.045 (SCFA + lactate) (illustrative), SCFA are absorbed by the host at 0.3 per hour; the cardinal pH model f = (pH−pH_min)(pH−pH_max) / [(pH−pH_min)(pH−pH_max) − (pH−pH_opt)²] gives each taxon a pH niche.

Immune loop. Signal σ = N_D/(N_D + K); activation obeys dI/dt = (σ − I)/τ, τ = 8 h, and adds a loss rate to the opportunist only (tolerance to the others).

Diversity. Shannon H′ = −Σ p ln p; more even communities have higher H′ (max ln 5 = 1.61). Diversity alone does not say "good" or "bad".

Simplifications. Well-mixed compartment, five generic taxa, one limiting resource each, no spatial gradients in the equations (the 3D neighbourhood is a drawing of a sample of the cells), no host genetics, no real species. Scale in the picture is not true to life. Models built with Blender, scene by Three.js.