Table 2. The overall rate of nonsynonymous (dN) to synonymous (dS) substitutions and sites under positive selection in the respective protein-coding genes.

			PAML					FUBAR
			M7 vs M8		M8 vs M8a
Gene	n	dN/dS (95% CI)	LRT (2Δl)	p-value	LRT (2Δl)	p-value	Positively selected sites	Positively selected sites
N	15	0.19 (0.13–0.28)	0.053	0.97	0.037	0.85	None	None
P	6	0.49 (0.34–0.68)	1.99	0.37	1.88	0.17	None	None
M	13	0.27 (0.18–0.39)	10.9	0.004	10.9	0.001	331, 335	335
F	19	0.17 (0.12–0.24)	0	1	0	1	None	None
H	16	0.20 (0.15–0.26)	0	1	0	1	None	None
L	8	0.16 (0.13–0.20)	0	1	0	1	None	None

n:Number of sequences; LRT: likelihood ratio test statistic for null model versus alternative model; 2Δl: Differences in the likelihood scores of null and alternative models; Null models (Neutral model): M7 and M8a; Alternative model (Selection model): M8; Degrees of freedom for M7-M8 and M8-M8a are 2 and 1, respectively. Sites with p < 0.05 and posterior probability > 0.95 for PAML and FUBAR respectively are considered to be under positive selection. Sites that are detected to be under positive selection by both methods are in bold.