Table 3.

Likelihood ratio test (LRT) statistics for F3x4 and F61 models in Plasmodium phylogenies.

Gene	N° strain/isolates	Model	M1a vs M2aa		M7 vs M8a		Positively selected sitesc
			−2ΔlnLb	p value	−2ΔlnLb	p value
KAHRP	16
		F3x4	34.81	2.78 × 10−08	35.78	1.70 × 10−08	P123, K443, S467, V492, G516, S603
		F61	36.05	1.92 × 10−9	37	1.18 × 10−9
MESA	11
		F3x4	13.54	1.15 × 10−03	13.71	1.05 × 10−03	N315
		F61	6.61	0.010	7.24	0.0071
EMP3	15
		F3x4	1.70	0.43	1.76	0.41	—
		F61	1.50	0.22	6.39	0.011

Notes: ^aM1a is a nearly neutral model that assumes one ω class between 0 and 1 and one class with  ω = 1; M2a (positive selection model) is the same as M1a plus an extra class of ω > 1; M7 is a null model that assumes that 0 < ω < 1 is beta distributed among sites; M8 (positive selection model) is the same as M7 but also includes an extra category of sites with ω > 1; ^b2ΔlnL: twice the difference of the natural logs of the maximum likelihood of the models being compared; c. Positions refer to the Pf_3D7 strain sequence.