Controls
Presets
Parameters
Quality
Camera
Model checks
The science
The niche. Haematopoietic stem cells (HSCs) live in bone-marrow niches where stromal cells present SCF, CXCL12 and thrombopoietin (TPO) signals. A strong signal keeps an HSC dormant and anchored (quiescent, Q); a weak one (for example after G-CSF mobilisation) lets it cycle (active, A). The model uses Q → A at a = a0/(1+s)2 and A → Q at r = r0(1+s), so the quiescent fraction is r/(a+r).
Division outcomes. An active cell divides at rate λ. With probability p(1 - N/K) both daughters stay stem cells (symmetric renewal; it needs a free niche site), with probability q both differentiate and leave (shed progeny; q is capped at 1 - p), otherwise the division is asymmetric and N is unchanged. With the Q/A split near its fast equilibrium the mean-field equation is dN/dt = g N - b N2 with g = λφ(p - q), b = λφp/K, φ = a/(a+r); for g > 0 it is the logistic curve with set point N* = K(p - q)/p, and the time to regain 90 % of N* after injury is t90 = ln(9(N* - N0)/N0)/g. A strong niche signal lowers φ and therefore slows regeneration: dormancy protects the stem cells from cell-cycle poisons (5-FU-like) at the price of slow recovery.
Simplifications. No spatial competition between sites, no ageing clock, no separate long- and short-term stem-cell classes, no niche remodelling after injury; rates are illustrative (not fitted mouse data) and the time unit is days. Newborn cells start active, so the exact stochastic mean follows the coupled Q/A equations, which differ slightly from the quasi-steady logistic curve. Checks: stationary quiescent fraction from single-cell Gillespie; RK4 versus the exact solution; cell-level ensemble versus the coupled Q/A mean-field ODE; recovery time closed form versus bisection; long-run time average versus N*.