Skip to main content

Table 2 Positive site-specific selection identified with CodeML models, BUSTED and MEME, and MrBayes

From: Divergent and non-parallel evolution of MHC IIB in the Neotropical Midas cichlid species complex

Model logL Parameter estimates PSS
M1aa − 3068.35 p0 = 0.606, p1 = 0.394
ω0 = 0.128, ω1 = 1
NA
M2aa − 2985.41 p0 = 0.379, p1 = 0.366, p2 = 0.255
ω0 = 0.073, ω1 = 1, ω2 = 4.138
1, 3, 10, 12, 13, 22, 28, 35, 40, 42, 44, 45, 46
M7a − 3052.97 p = 0.171, q = 0.184 NA
M8a − 2984.37 p0 = 0.749, p1 = 0.251, p = 0.194
q = 0.219, ω = 3.767
1, 3, 10, 12, 13, 22, 28, 35, 40, 42, 44, 45, 46
Constrainedb − 2885.7 p1 = 0.369, p2 = 0, p3 = 0.631
ω1 = 0, ω2 = 0.54, ω3 = 1
NA
Unconstrainedb/MEME − 2805.1 p1 = 0.351, p2 = 0.543, p3 = 0.106
ω1 = 0.37, ω2 = 0.37, ω3 = 43.67
1, 3, 5, 12, 13, 16, 21, 22, 35, 36, 39, 42, 44, 45, 46, 47
MrBayes   p = 0.356, pN = 0.359, p+  = 0.284
ω = 0.082, ωN = 1, ω+ = 4.577
1, 3, 13, 22, 28, 35, 36, 42, 45, 46
  1. logL, log-likelihood value; PSS, positively selected sites; NA, not allowed; PSS indicated for the unconstrained model are identified with MEME. Model parameters are: M1a and M2a: p0 = proportion of sites with 0 < ω0 < 1, p1 = proportion of sites with ω1 = 1, p2 = proportion of sites with ω2 > 1; M7 and M8: p, q = β distribution parameters, p0 = proportion of sites with ω within the β distribution, p1 = proportion of sites with ω > 1; constrained and unconstrained: p1, p2 = proportion of sites with 0 ≤ ω1 ≤ ω2 ≤ 1, p3 = proportion of sites with ω3 = 1 (constrained) or ω3 > 1 (unconstrained); MrBayes: p = proportion of sites with 0 ≤ ω < 1, pN = proportion of sites with ωN = 1, p+ = proportion of sites with ω+  > 1
  2. aCodeML models
  3. bBUSTED models