Measuring environmental complexity by using multivariate mutual information. (A) Multivariate mutual information between environmental signals and nutrient presence was found to correlate with the evolutionary rate, in experiments with 1024 cells, 5 environments, 107 time units, and 64 experiments per environment (replicates). Insert shows the lack of correlation of the evolutionary rate with other information measures (as reported in Table 1). (B) Fitness trajectories for AND, OR, A, and XOR environments averaged over 64 experiments; evolutionary rate is defined as the time constant τ of the exponential fit (w=A0-α*exp(-t/τ)) on the averaged fitness trajectory of the evolved populations (dashed curves and inset table).
Mozhayskiy and Tagkopoulos BMC Bioinformatics 2012 13(Suppl 10):S10 doi:10.1186/1471-2105-13-S10-S10