Skip to main content


Figure 3 | BMC Evolutionary Biology

Figure 3

From: Detecting coevolution without phylogenetic trees? Tree-ignorant metrics of coevolution perform as well as tree-aware metrics

Figure 3

Performance of method, alphabet combinations for detecting signal at ( i , i + 3) and ( i , i +4). Each combination of method and alphabet is represented as a single point, with coordinates defined by the -log(p-value) at (i, i + 3) (x-axis) and (i, i + 4) (y-axis) for the (A) tetrapod myoglobin alignment, (B) randomized tetrapod myoglobin alignment, (C) chordate myosin rod alignment, and (D) randomized chordate myosin rod alignment. Blue lines indicate a significance threshold of α = 0.01. SCA cutoffs included for (A-B) are 0.9 and 0.8, and for (C-D) are 0.4 and 0.6. These represent the best cutoff values at (i, i +3) and (i, i +4), respectively, for each positive control data set. Label key: method, alphabet, so S8, CHARGE_HIS_2 refers to SCA with cutoff = 0.80, and the CHARGE_HIS_2 alphabet. Method abbreviations: AS: Ancestral States; L07: LnLCorr07; L99: LnLCorr99; MI: Mutual Information; NMI: Normalized Mutual Information; RMI: Resampled Mutual Information; Sn: Statistical Coupling Analysis, cutoff = n/10; MIP, Corrected Mutual Information; Gn: Generalized Continuous-Time Markov Process Coevolutionary Algorithm, ε = n/10; CM: CoMap. Alphabet definitions in Table 5.

Back to article page