Choosing appropriate substitution models for the phylogenetic analysis of protein-coding sequences

Area of Science:

• Computational Biology
• Evolutionary Biology
• Genomics

Background:

• Phylogenetic inference of protein-coding sequences often neglects the genetic code's complexity.
• Standard nucleotide substitution models are frequently used, despite their limitations for coding sequences.
• Codon-based models, while more accurate, are computationally intensive and often excluded from model selection.

Purpose of the Study:

• To investigate codon position (CP) models as an efficient alternative to standard nucleotide substitution models for phylogenetic inference.
• To evaluate the statistical performance of CP models compared to nucleotide and codon models.
• To determine the most appropriate evolutionary models for diverse gene alignments.

Main Methods:

• Evaluated 11 substitution models, including standard nucleotide, codon, and four codon position (CP) models.
• Applied models to alignments of 177 RNA virus genes and 106 yeast genes.
• Assessed model appropriateness using established statistical criteria.

Main Results:

• The majority of analyzed gene alignments were best described by CP substitution models.
• CP models provided a statistically superior fit compared to standard nucleotide models.
• CP models offer a computationally realistic alternative to full codon models.

Conclusions:

• Codon position (CP) substitution models are a computationally feasible and statistically robust alternative for phylogenetic analysis of coding sequences.
• Incorporating CP into evolutionary models improves the accuracy of phylogenetic inference for genes.
• These findings have significant implications for understanding molecular evolution and phylogenetic reconstruction.