Evolutionary Rescue Race: Mutant Establishment in a Declining Population

A flask on a bench built in Blender with thousands of simulated cells: watch the wild type fade and resistant mutants rescue the population, or fail, trial after trial. Drag to orbit.

Switch to 2D
Controls

Presets

Parameters

Quality

Camera

Model checks

The science

The race. After an environmental change (a drug, a heat wave, a new pollutant) a population of wild-type cells shrinks at net rate −d per generation. Evolutionary rescue happens when resistant mutants, growing at net rate +s, appear and establish before the wild type has gone extinct. The population therefore dips, reaches a minimum and either recovers (U-shaped curve) or disappears. Each wild-type birth produces a mutant with probability μ, and some resistant cells may already exist at the start (standing variation p0).

Equations. With birth rate 1: Nw(t) = N0 e−(d+μ)t and Nm(t) = n0 est + μ N0(est − e−(d+μ)t)/(s + d + μ) are the mean trajectories, but a single new mutant is lost by chance most of the time: starting from one cell the lineage reaches n cells with probability s / (1 − (1−s)n) (gambler’s ruin of a birth-death process), about s for large n. Mutants arise as a Poisson stream, so the expected number of lineages that succeed is Λ = π(n0 + μ N0(1 − e−DT)/D) and the rescue probability is 1 − e−Λ. The animated run is a tau-leap simulation with Poisson births and deaths and a soft ceiling at 20 000 cells; the page repeats trials automatically and counts how many were rescued.

Numerical checks. (1) The closed-form means are compared with an RK4 integration; (2) an exact birth-death chain of one mutant gives the lineage probability; (3) exact Gillespie runs of the whole population (population size limited to 300 for speed, plus a fixed reference case) are compared with 1 − e−Λ with a 4-sigma binomial tolerance; (4) the minimum of the mean population is found by root-finding and by a grid scan.

Simplifications. Mutants are never lost back to the wild type, there is one mutation step, no cost of resistance, no sexual recombination, no spatial structure, and a single constant environment; a real population also changes its birth rate with density. Rescue is declared at 25 mutants.

Loading model...
Loading Blender model...
-
-
Made with 🟠 Blender · Three.js