Tests for positive selection using ω-ratio tests: candidate amino acid sites identified by all the ω-ratio tests are underlined.

aAll the PAML and HyPhy analyses were run on four non-recombinant PgiC1 segments that are delimited on the basis of the three recombinant breakpoints (nucleotide positions 195, 561 and 855) identified by GARD.bWithin the PAML package, two nested tests, each including one neutral (M1a/M7) and one selection model (M2a/M8), were used to test for selection. Within each of the two nested tests, 2dL is the likelihood-ratio statistic used to test whether the selection model fits the data significantly better than the neutral model (df = 2).Candidate targets of positive selection identified in M2a, M8 are those with posterior probabilities >50% in Bayes Empirical Bayes analyses. Sites with posterior probabilities >99% are indicated in bold text.*: 0.01 p cThe model REL within the Hyphy package was used to test for selection. Candidates targets of positive selection are those with Bayes factors > 50, candidates with Bayes factors larger than 500 are shown in bold text.dCandidate targets of positive selection identified in omegaMap are those with posterior probabilities >50%. Sites with posterior probabilities >99% are shown in bold text.Tests for positive selection using ω-ratio tests: candidate amino acid sites identified by all the ω-ratio tests are underlined.