DEX: a consensus-based amino acid exchangeability measure for improved codon substitution modelling

Physicochemically similar amino acids undergo more frequent substitutions compared to dissimilar amino acid pairs. Despite their clear potential, amino acid similarity matrices remain underused for certain molecular evolution applications. One key potential application that is understudied is in quantifying the strength of natural selection based on amino acid substitution patterns. This is partially due to the high number of proposed amino acid distance measures and the lack of agreement on which are most accurate. In this study, we assessed the performance of 30 amino acid distance measures, including a new amino acid distance measure we developed based on recent deep mutational scanning data. We compared these measures across codon substitution models fit to alignments spanning Streptococcus , Drosophila , and mammalian lineages, as well as segregating variants across Escherichia coli strains and human genotypes. We further constructed consensus matrices from combinations of top-performing measures in this analysis using the DISTATIS approach and retested these matrices. Our results show that experimentally-derived measures, particularly our new measure, DMS-EX and the existing experimental exchangeability measure, best fit codon substitution patterns across diverse lineages. We found that a consensus measure based on these two approaches, which we named DEX, performed best overall. We also explored the value of asymmetric exchangeabilities in DMS-EX for predicting allele frequencies of replacement polymorphisms across diverse lineages, including when conditioned on buried vs. exposed sites. Overall, we provide a systematic comparison of the performance of existing measures. The amino acid distance measures we introduce constitute a substantial improvement for exploring novel methods for quantifying the strength of natural selection and for providing improved baselines for future benchmarking approaches. Significance Protein-coding genes have long been a focus for researchers studying the strength and direction of selection. By studying non-synonymous substitutions, those that change amino acids, it is possible to estimate the relative strength of selection. Despite widespread interest in such approaches, information on which amino acids are exchanged is underused in most molecular evolution applications. This is partly because many different measures exist for quantifying amino acid distances, particularly those based on physicochemical properties. A newer class of amino acid distance measures is derived from deep mutational scanning datasets, where virtually every possible substitution is tested for its impact on protein function. We characterised and compared 30 amino acid distance measures, including a novel measure based on deep mutational scanning data. We highlight differences in how well these measures fit real substitution data. Overall, we find that DEX, which is a consensus of our new measure and an existing experimental exchangeability measure, performed best in these models. This work will serve as the basis for future improved methods for inferring selection efficacy from protein-coding alignments.