Environmental heterogeneity may contribute to the high levels of genetic variation in glucosinolate genes found in Arabidopsis thaliana.

Minimal modifications of common regulatory circuits can allow bacteria to learn from the past to better predict the future, on a physiological, rather than evolutionary, timescale.

Experimentally evolved budding yeast populations reveal the role of contingency and chance in shaping the emergence and dynamics of pleiotropic effects of adaptation.

Environmental conditions strongly impact the fitness effects of Hsp90, resulting in the selection of Hsp90 sequences in nature that are robust to a variety of stressful conditions.

The impact of changing gene expression noise on fitness reveals beneficial or deleterious effects in a stable environment depending how close the average expression level is to the fitness optimum.

An interdisciplinary approach combining high throughput genotype-to-phenotype mapping, population genetic simulations, and experimental evolution provides an answer to the question of how populations evolve new functions by providing tests of the role antagonistic coevolution plays in pressuring populations to innovate.

Single-cell experiments and mathematical modelling show that cellular signalling networks are vulnerable to trade-offs in speed versus accuracy, but that these vulnerabilities can be overcome by distributing the two tasks to different, although interacting, subnetworks.

In bacteria, frequent adaptive copy-number mutations can hinder the fixation of beneficial point mutations and hence the divergence of duplicated DNA sequences.

Quorum sensing enables heterogeneous production of autoinducers in microbial populations, suggesting an alternative mechanism to stochastic gene expression in bistable gene-regulatory circuits to control phenotypic heterogeneity.

The timescale dependence of d_N/d_S in bacteria is better explained by adaptive than purifying dynamics, suggesting comparative genomics can underestimate past adaptation.