Bayesian and maximum likelihood phylogenetic analyses of protein sequence data under relative branch-length differences and model violation

Table 1 Linear fit equations for data in Figure 1. Slope, y-intercept, significance, and R² values for linear equations relating bootstrap proportion and Bayesian posterior values shown in panels A, D and E of Figure 1, i.e. for all nodes subtending identical subtrees among the 21 empirical protein-sequence datasets, regardless of whether the corresponding full ML and Bayesian trees are topologically identical or not.

Data ¹	Panel	Slope	SE ²	Signif	y-Intcpt	SE	Signif	Mult R ²	Adj R ²
JTT model, all data	A	0.4993	0.0536	0.001	51.823	4.568	0.001	0.6207	0.6136
	D	0.5150	0.0398	0.001	50.125	3.441	0.001	0.5279	0.5247
	E	0.5101	0.0322	0.001	50.625	2.773	0.001	0.5508	0.5486
EQ model, all data	A	0.4557	0.0572	0.001	55.661	4.871	0.001	0.5452	0.5367
	D	0.4517	0.0352	0.001	56.269	3.050	0.001	0.5227	0.5195
	E	0.4531	0.0298	0.001	56.077	2.569	0.001	0.5297	0.5274
JTT model, BP <85%	A	0.7536	0.1771	0.001	38.180	10.819	0.01	0.4880	0.4610
	D	0.7816	0.1258	0.001	34.931	8.115	0.001	0.4081	0.3976
	E	0.7694	0.1020	0.001	36.112	6.487	0.001	0.4251	0.4176
EQ model, BP <85%	A	0.6909	0.1809	0.01	43.124	11.054	0.001	0.4342	0.4044
	D	0.6837	0.1099	0.001	43.086	7.089	0.001	0.4088	0.3982
	E	0.6843	0.0922	0.001	43.174	5.866	0.001	0.4170	0.4095

¹All data: values based on all data shown in the respective panel in Figure 1; BP <85%: values based on only those data for which the value of the PROML bootstrap proportion is less than 85%.
²SE, standard error; Signif, significance (probability level that estimate >|t|); y-Intcpt, y-intercept; Mult R², multiple R²; Adj R², adjusted R².

ISSN: 2730-7182