Skip to main content
Fig. 1 | BMC Evolutionary Biology

Fig. 1

From: A coarse-graining, ultrametric approach to resolve the phylogeny of prokaryotic strains with frequent homologous recombination

Fig. 1

Illustration of the procedure of the proposed CGP algorithm. a The algorithm takes n aligned sequences as input, which can be nucleotide sequences or amino acid sequences; substitutions on the sequences are represented by coloured markers. b Each of the n(n-1)/2 genome pairs is divided into equal-sized segments, and the pairwise substitutions on each segment is enumerated to obtain the distribution of local SSPs density (denoted as g(x)). c The algorithm aims to infer the distance matrix of the genome sequence pairs from the n(n-1)/2 SSP distributions. d In particular, the algorithm fits the empirical SSP distributions with a model; the input of this model involves a matrix of n(n-1)/2 coalescent times and other model parameters (mutation rate μ, recombination rate ρ, average population divergence θ and transfer efficiency δTE). e In the fitting process, the n(n-1)/2 coalescent times are constrained (matrix cells with the same colour have the same value), such that the matrix can be bijectively mapped to a UPGMA tree. f the algorithm explores the model parameter space and tree space to obtain the best fit ultrametric tree

Back to article page