Directed Evolution Lab | mysimulator.uk

Directed Evolution

Library design

Library size 24 variants

Mutation rate (error-prone PCR) 0.60

Selection stringency 20%

Round history

Round

Survivors

Best fitness

–

Mean fitness

–

Pool diversity

–

Rounds to plateau

–

How it works

Each glowing dot is a protein variant placed at coordinates that stand in for its position in sequence space (a 2D projection, the way real variant libraries get visualised after clustering). Height and colour show fitness — activity, stability, whatever the screen selects for — mapped out as a rugged landscape with several local peaks and one global optimum.

Every round: the current survivors are each mutated library size times using Gaussian steps scaled by the mutation rate (this stands in for error-prone PCR, recombination or saturation mutagenesis — bigger steps explore further but risk destructive mutations). The library is screened, and only the top slice — set by selection stringency — survives to seed the next round's library, exactly like a real select-and-amplify cycle.

child = parent + N(0, mutationRate²)
keep  = top stringency% of library by fitness(child)
next round's parents = keep

Mutation rate too low → the population creeps and gets stuck on the nearest small peak (premature convergence).
Mutation rate too high → most mutants land in dead sequence space; fitness gain stalls even after many rounds.
Stringency too high → the pool collapses to one clone fast, losing the diversity needed to find a better peak later.
Stringency too low → weak variants keep diluting the next library, slowing the climb.

This is the core trade-off behind real directed-evolution campaigns: library breadth versus screening throughput, and exploration versus exploitation of the fitness landscape.

Round 0

Best-ever variant Surviving parents Screened-out variants

⚠ Couldn't load the 3D engineThree.js failed to load from the CDN. Check your connection and reload.