Table 1.

Likelihood analysis of the site models in the PAML program for the TET gene sequence data.

Model	Parameter Estimates a	LnL	2ΔlnL (p-Value)	Positively Selected Sites b,c
TET1
M0	ω = 0.385	−87,143.158462		None
M1a	ω0 = 0.205, p0 = 0.623, ω1 = 1.000, p1 = 0.377	−86,007.766228		Not allowed
M2a	ω0 = 0.205, p0 = 0.623, ω1 = 1.000, p1 = 0.298, ω2 = 1.000, p2 = 0.078	−86,007.766228	M2a vs. M1a	None
			0 (p = 1)
M7	p = 0.679, q = 0.983	−85,718.609725		Not allowed
M8	p0 = 0.962, p = 0.744, q = 1.197, p1 = 0.038, ω1 = 1.294	−85,707.99818	M8 vs. M7	524F, 811M
			21.22309 (p < 0.001)
TET2
M0	ω = 0.346	−75,677.716183		None
M1a	ω0 = 0.184, p0 = 0.664, ω1 = 1.000, p1 = 0.336	−74,593.700860		Not allowed
M2a	ω0 = 0.184, p0 = 0.664, ω1 = 1.000, p1 = 0.292, ω2 = 1.000, p2 = 0.044	−74,593.700860	M2a vs. M1a	None
			0 (p = 1)
M7	p = 0.617, q = 1.064	−74,286.497185		Not allowed
M8	p0 = 0.978, p = 0.671, q = 1.250, p1 = 0.022, ω1 = 1.558	−74,265.863135	M8 vs. M7	123R*, 469V, 673V, 968P, 1002T, 1205S, 1767S
			41.2681 (p < 0.001)
TET3
M0	ω = 0.115	−48,719.653509		None
M1a	ω0 = 0.056, p0 = 0.860, ω1 = 1.000, p1 = 0.140	−47,597.069946		Not allowed
M2a	ω0 = 0.056, p0 = 0.859, ω1 = 1.000, p1 = 0.139, ω2 = 999.000, p2 = 0.002	−47,599.818203	M2a vs. M1a	None
			−5.496514 (p = 1)
M7	p = 0.237, q = 1.380	−47,268.934744		Not allowed
M8	p0 = 0.999, p = 0.237, q = 1.380, p1 = 0.001, ω1 = 999.000	−47,270.552239	M8 vs. M7	None
			−3.23499 (p = 1)

^a p and q denote the parameters of the β distribution, β (p, q), which determine that of ω values; ^b Amino acid sites inferred to be under positive selection are shown with a posterior probability PP > 0.95; * Posterior probability (PP) > 0.99; ^c The referenced sequences for the TET1, TET2 and TET3 genes were human TET1, human TET2, and human TET3, respectively.