Likelihood-based evaluation of character recoding schemes for phylogenetic analysis

A likelihood-based framework statistically evaluates amino-acid character recoding schemes, quantifying how reducing 61-state codon models into 20-state (or 6-state) models affects available evolutionary information.

• Why it matters: Recoding is widely used, but collapsing state spaces can discard evolutionary signal, especially when amino-acid frequencies vary across sites or lineages.
• What they did: The method builds a likelihood framework to assess recoding schemes, focusing on mapping 61-state codon substitution models into 20-state amino-acid replacement models and extending to 6-state groupings.
• The result: It enables formal, statistical comparison of recoding options, helping determine when reduced state-space models may be too information-losing for phylogenetic inference.

AI summary of the abstract · may contain errors · not medical advice · read the paper.