Inverse Ising techniques to infer underlying mechanisms from data
Zeng Hong-Li1, 2, †, Aurell Erik3, 4, ‡
       

(a) Temporal behavior of all allele frequencies defined as fi[1]. Data recorded every 5 generations. (b) An example of pairwise correlation changing with time. With finite population size, there exists strong fluctuations in the system. (c) Scatter plot for the reconstructed against the tested fitness with DCA-nMF (red dots) and DCA-PLM (blue dots) algorithm for Jij’s. Parameters: the number of loci L = 25, the number of individuals N = 200, mutation rate μ = 0.01, recombination rate r = 0.1, crossover rate ρ = 0.5, standard deviation of epistatic fitness σ = 0.002.