View article

Biological organisms can adapt to unstable conditions. This adaptability results from a combination of heterogeneity and plasticity within a population, which means perturbations are unlikely to have the same effect on all population members, allowing time for adaptations to evolve. In this work we show that this behaviour can be duplicated by introducing stochasticity to generate functional heterogeneity in artificial gene regulatory networks, a biologically inspired tool used to solve computational problems.

We show that by introducing this stochasticity, artificial gene regulatory networks are more robust to perturbations when applied to complex dynamical tasks meaning that they are less affected than their deterministic counterparts. Moreover, we analyse the structure of the artificial gene regulatory networks to understand what effect stochasticity has on the structure of the networks and whether we …