Are Third Positions Really That Bad? A Test Using Vertebrate Cytochrome b

Area of Science:

• Molecular Evolution
• Bioinformatics
• Phylogenetics

Background:

• Third codon positions in DNA sequences are frequently downweighted or excluded in phylogenetic analyses.
• This practice is based on the assumption that higher substitution rates at third positions introduce more noise than valuable phylogenetic signal.

Purpose of the Study:

• To empirically test the common assumption that third codon positions are less reliable for phylogenetic inference compared to first and second positions.
• To evaluate the phylogenetic content of individual codon positions (1st, 2nd, and 3rd) in vertebrate DNA sequence data.

Main Methods:

• Ten vertebrate datasets were partitioned into three subsets, each containing only first, second, or third codon positions.
• Phylogenetic performance of these subsets was assessed using established methods for evaluating phylogenetic data content.

Main Results:

• Contrary to expectations, third codon positions demonstrated superior phylogenetic performance (3rd > 1st > 2nd).
• The results indicate that downweighting third positions in favor of second positions is not supported by the data.

Conclusions:

• The study challenges the default downweighting of third codon positions in phylogenetic analyses.
• Unless specific evidence of misleading information exists, third codon positions should be retained and analyzed for their phylogenetic value.