Serine codon-usage bias in deep phylogenomics: pancrustacean relationships as a case study.

Area of Science:

• Evolutionary Biology
• Genomics
• Bioinformatics

Background:

• Phylogenomic analyses commonly use amino acid data, but the optimal data type (nucleotides vs. amino acids) for ancient relationship inference is debated.
• Clarifying pancrustacean (crustaceans and insects) evolutionary history is a key challenge in arthropod phylogeny.

Purpose of the Study:

• To investigate the discrepancy between nucleotide- and amino acid-based phylogenomic trees for pancrustaceans.
• To determine the optimal data type for resolving ancient evolutionary relationships within Pancrustacea.

Main Methods:

• Utilized a dataset of 62 protein-coding genes from Regier et al. (2010) for pancrustacean phylogenomic analysis.
• Compared phylogenetic trees inferred from both nucleotide and amino acid data.
• Analyzed synonymous codon usage bias, particularly for serine, arginine, and leucine, across different arthropod lineages.

Main Results:

• The nucleotide data inferred a highly supported pancrustacean tree distinct from the poorly supported amino acid tree.
• Discrepancies were attributed to substitutions within synonymous codon families, especially serine (TCN and AGY).
• Serine codon usage bias correlated with the nucleotide tree topology but not the amino acid tree, suggesting a compositional bias affecting nucleotide data.

Conclusions:

• Synonymous codon usage bias, particularly in serine, appears to influence nucleotide-based phylogenomic analyses, potentially affecting tree topology.
• Amino acid data may also be impacted by this bias due to intermediate substitutions.
• Neither nucleotide nor amino acid data from this dataset robustly resolves pancrustacean relationships, which remain largely unresolved.