Table 2.

Summary of the ω estimation and likelihood ratio test (2ΔL) between two-ratio (ω ₀ ≠ ω ₁ = ω ₂) and three-ratio (ω ₀ ≠ ω ₁ ≠ ω ₂) models.

Hypothesis	np	lnL	2ΔL	p	ω	Supporting hypothesis in Table 1
TFL1 vs. magnoliids						1. Functional constraint hypothesis
ω 1 = ω 2	135	−12594.3259			ω 0 = 0.1044, ω 1 = ω 2 = 0.13221
ω 1 ≠ ω 2	136	−12593.9631	0.7256	0.3258	ω 0 = 0.1021, ω 1 = 0.1361, ω 2 = 0.1678
CEN vs. magnoliids						1. Functional constraint hypothesis
ω 1 = ω 2	135	−12597.2165			ω 0 = 0.1173, ω 1 = ω 2 = 0.1072
ω 1 ≠ ω 2	136	−12597.2164	0.0002	0.9887	ω 0 = 0.1054, ω 1 = 0.1193, ω 2 = 0.1674
RCN1 vs. magnoliids						1. Functional constraint hypothesis
ω 1 = ω 2	135	−12596.0957			ω 0 = 0.1170, ω 1 = ω 2 = 0.0891
ω 1 ≠ ω 2	136	−12595.6616	0.8682	0.2774	ω 0 = 0.1140, ω 1 = 0.0841, ω 2 = 0.1665
RCN2 vs. magnoliids						1. Functional constraint hypothesis
ω 1 = ω 2	135	−12597.6561			ω 0 = 0.1137, ω 1 = ω 2 = 0.1164
ω 1 ≠ ω 2	136	−12597.5681	0.176	0.8708	ω 0 = 0.1137, ω 1 = 0.1078, ω 2 = 0.1240
RCN3 vs. magnoliids						1. Functional constraint hypothesis
ω 1 = ω 2	135	−12597.2484			ω 0 = 0.1126, ω 1 = ω 2 = 0.1314
ω 1 ≠ ω 2	136	−12596.6628	1.1712	0.2052	ω 0 = 0.1126, ω 1 = 0.1079, ω 2 = 0.1535
Eudicots vs. magnoliids						1. Functional constraint hypothesis
ω 1 = ω 2	135	−12592.2355			ω 0 = 0.0915, ω 1 = ω 2 = 0.1255
ω 1 ≠ ω 2	136	−12592.0526	0.3658	0.5494	ω 0 = 0.0916, ω 1 = 0.1085, ω 2 = 0.1264
Monocots vs. magnoliids						1. Functional constraint hypothesis
ω 1 = ω 2	65	−12592.9536			ω 0 = 0.1249, ω 1 = ω 2 = 0.0931
ω 1 ≠ ω 2	66	−12592.7766	0.354	0.5617	ω 0 = 0.1249, ω 1 = 0.1073, ω 2 = 0.0917

ω ₁, ω ₂, and ω ₀ are the Ka/Ks ratio of the branches of the eudicot TFL1 (or eudicot CEN, monocot RCNs), magnoliid TFL1-like, and background lineages, respectively.

np: number of parameters

p: p-value obtained from fitted model using χ ² test.