Table 1.

Likelihood Log ratios and PAML site models for positive selection.

Gene	N	Lc	S	Model	Categories	2Δl M3 vs. M0	2Δl M2 vs. M1	2Δl M8 vs. M7	Parameter estimates	Positively selected sites
MAVS	45	560	2.27	M1	Nearly Neutral (2 categories)	114.827712	31.398468	32.374384	P1 = 0.34017; P2 = 0.65983	14, 45*, 60*, 74, 78, 82*, 90*, 116*, 126, 149**, 163*, 312**, 438**, 465*, 513*
									ω1 = 0.06983; ω2 = 1.00000
				M2	Positive Selection (3 categories)				P1 = 0.33817; P2 = 0.57824; P3 = 0.08360
									ω1 = 0.09769; ω2 = 1.00000; ω3 = 3.81373
				M3	Discrete (3 categories)				P1 = 0.23466; P2 = 0.65644; P3 = 0.10890
									ω1 = 0.00868; ω2 = 0.82550; ω3 = 3.45775
				M7	Beta (10 categories)				P = 0.04076; q = 0.01519
				M8	Betaandw>1 (11 categories)				P0 = 0.90653; p = 0.23986; q = 0.13299
									p1 = 0.09347; ω1 = 3.64677

The beta distribution parameters p and q, and the fraction of sites under positive selection (p1). Positive selection parameters are bolded. p: significant at 5%; p: significant at 1%. Positive selection sites found by model M8 are listed by human sequence numbering. Sites with high posterior probability. 0.8–0.9 in bold, and 0.5–0.7 in normal print. In this case, the critical values are 5.99, 9.21, and 13.82, respectively, for a l distribution with 2 degrees of freedom.

^**significant at 1%;

^*significant at 5%.