The genetic code can cause systematic bias in simple phylogenetic models

摘要: Phylogenetic analysis depends on inferential methodology estimating accurately the degree of divergence between sequences. Inaccurate estimates can lead to misleading evolutionary inferences, including incorrect tree topology and poor dating historical species divergence. Protein coding sequences are ubiquitous in phylogenetic inference, but many standard methods commonly used describe their evolution do not explicitly account for dependencies sites a codon induced by genetic code. This study evaluates performance several datasets simulated under simple substitution model, describing range different types selective pressures. approach also offers insights into relative when there acting data. Methods based statistical models performed well was no or limited purifying selection (low dependency codon), although more biologically realistic tended outperform simpler models. exhibited greater variability strong (high codon). Simple substantially underestimate sequences, underestimation pronounced internal branches tree. resulted some performing poorly exhibiting evidence systematic bias inference. Amino acid-based nucleotide that contained generic descriptions spatial temporal heterogeneity, such as mixture hidden Markov models, coped notably better, producing accurate topology.