Skip to main content
. 2010 Mar 2;5(3):e9501. doi: 10.1371/journal.pone.0009501

Table 2. Identification of positively selected sites in piscidins by maximum likelihood analysis using various models of evolution.

Method Model Parameter estimates Ln Likelihood Model comparison Positively selected sites a
CODEML M0: neutral ω = 0.28 −809.60 None
M1: nearly neutral ω0 = 0.14, ω1 = 1.00p0 = 0.67, p1 = 0.33 −794.42 Not allowed
M2: positive selection ω0 = 0.15, ω1 = 1.00, ω2 = 13.76p0 = 0.66, p1 = 0.31, p2 = 0.03 −793.20 M2 vs M12ΔLnL = 2.44, df = 2, p = 0.29 29
M3: discrete ω0 = 0.05, ω1 = 0.29, ω2 = 2.25p0 = 0.22, p1 = 0.59, p2 = 0.19 −792.42 M3 vs M02ΔLnL = 34.36, df = 4, p = 0.00 25, 29, 43
M7: β p = 0.73, q = 1.27 −795.43 Not allowed
M8: β + ωS>1 p = 0.73, q = 1.27ω = 2.31p0 = 0.83, p1 = 0.17 −792.29 M8 vs M72ΔLnL = 5.68, df = 2, p = 0.04 29
HYPHY REL ω = 1.14 25, 29
a

Only positively selected sites with Bayesian posterior probabilities equal or greater than 95% are indicated. Sites with a posterior probability greater than 99% are highlighted in bold.