Table 2.

Tests for positive selection among TaDofs gene family using site-specific models

Models	np	Estimates of parametersa	lnL	2△lnL	Positively selected sitesb
M0 (one-ratio)	191	ω = 0.05384	− 4539.929	883.03 (M3 vs M0)	Not allowed
M3 (discrete)	195	p0 = 0.43174, p1 = 0.26150, p2 = 0.30676 ω1 = 0.00039, ω2 = 0.03024, ω3 = 0.23081	− 4098.415		None
M7 (beta)	192	p = 0.19976, q = 2.34994	− 4083.224	2036.983 (M8 va M7)	Not allowed
M8 (beta & ω)	194	p0 = 0.99999, p = 0.82856, q = 1.22843(p1 = 0.00001), ω = 2.55223	− 5101.716		30A**, 31 T**, 33A**, 45 K, 67S, 82H, 85S, 102G*, 103 T, 104S**, 105D

Note: *, p < 0.05 and **, p < 0.01 (χ2 test). lnL, log likelihood. 2△lnL, twice the log-likelihood difference of the models. a, ω was estimated under model M0, M3, M7, M8. b, the number of amino acid sites estimated to have undergone positive selection. np, number of free parameters