We also simulated the evolution of a protein coding sequence consisting of L independent codon sites, in order to produce data that could be analyzed by the the PAML package [7] (link). We simulated populations for each site independently. In the final generation of each simulation, two individuals were sampled (either from a single population or from two divergent populations), and the corresponding codons were concatenated. A set of such simulations produces a pair of nucleotide sequences of length 3L.
In each simulation at a site, an individual could carry one of the 64 codons. The mutation probability was μ per nucleotide per generation. The fitness of an individual was determined by the encoded amino acid: we assumed that only two amino acids, alanine and valine, were allowed at the site; one of them was the resident amino acid and conferred fitness 1, the other was the mutant amino acid and conferred fitness 1+s; codons encoding other amino acids or stop codons were assumed lethal (non-reproductive). In all other respects the codon-based simulation was identical to the two-allele simulation. We initiated all simulations with a population monomorphic for codon GTT, which determined the initial resident allele (valine). The following parameter values were used:
s ∈ {−0.005,−0.002,0,0.002,0.005}, μ ∈ {10−7,10−6}. We ran the single population simulations for L = 104 sites for T = 5×105 generations. We ran the two population simulations for L = 103 sites for T = 0.25 μ−1 generations.
We used the CODEML program from the PAML package to infer the most likely dN/dS ratio for each pair of sequences. We used the likelihood ratio test, based on the χ2 distribution, to determine the 95% confidence interval on the estimated dN/dS ratio.
In each simulation at a site, an individual could carry one of the 64 codons. The mutation probability was μ per nucleotide per generation. The fitness of an individual was determined by the encoded amino acid: we assumed that only two amino acids, alanine and valine, were allowed at the site; one of them was the resident amino acid and conferred fitness 1, the other was the mutant amino acid and conferred fitness 1+s; codons encoding other amino acids or stop codons were assumed lethal (non-reproductive). In all other respects the codon-based simulation was identical to the two-allele simulation. We initiated all simulations with a population monomorphic for codon GTT, which determined the initial resident allele (valine). The following parameter values were used:
s ∈ {−0.005,−0.002,0,0.002,0.005}, μ ∈ {10−7,10−6}. We ran the single population simulations for L = 104 sites for T = 5×105 generations. We ran the two population simulations for L = 103 sites for T = 0.25 μ−1 generations.
We used the CODEML program from the PAML package to infer the most likely dN/dS ratio for each pair of sequences. We used the likelihood ratio test, based on the χ2 distribution, to determine the 95% confidence interval on the estimated dN/dS ratio.
Free full text: Click here
