Whilst one may have believed that associative learning requires a nervous system, this paper shows that chemical networks can be evolved in silico to undertake a range of associative learning tasks with only a small number of reactions. The mechanisms are surprisingly simple. The networks can be analysed using Bayesian methods to identify the components of the network responsible for learning. The networks evolved were simpler in some ways to hand-designed synthetic biology networks for associative learning. The motifs may be looked for in biochemical networks and the hypothesis that they undertake associative learning, e.g. in single cells or during development may be legitimately entertained.
McGregor S, Vasas V, Husbands P, Fernando C (2012) Evolution of Associative Learning in Chemical Networks. PLoS Comput Biol 8(11): e1002739. http://dx.doi.org/10.1371/journal.pcbi.1002739