Extra base hits: widespread empirical support for instantaneous multiple-nucleotide changes

Abstract

Despite many attempts to introduce evolutionary models that permit substitutions that instantly alter more than one nucleotide in a codon, the prevailing wisdom remains that such changes are rare and generally negligible (or are reflective of non-biological artifacts, such as alignment errors), and codon models continue to posit that only single nucleotide change have non-zero rates. We develop and test a simple hierarchy of codon-substitution models with non-zero evolutionary rates for only one-nucleotide (1H), one-and two-nucleotide (2H), or any (3H) codon substitutions. Using 35,000 empirical alignments, we find widespread statistical support for multiple hits: 58% of alignments prefer models with 2H allowed, and 22% -with 3H allowed. Analyses of simulated data suggest that these results are not likely to be due to simple artifacts such as model misclassification or alignment errors. Further modeling revealed that synonymous codon island jumping among codons encoding serine, especially along short branches, contributes significantly to this 3H signal. While serine codons were prominently involved in multiplehit substitutions, there were other common exchanges contributing to better model fit. It appears that a small subset of sites in most alignments have unusual evolutionary dynamics not well explained by existing model formalisms, and that commonly estimated quantities, such as dN/dS ratios may be biased by model misspecification. Our findings highlight the need for continued evaluation of assumptions underlying workhorse evolutionary models and subsequent evolutionary inference techniques. We provide a software implementation for evolutionary biologists to assess the potential impact of extra base hits in their data in the HyPhy package.

Natural selection in the evolution of SARS-CoV-2 in bats created a generalist virus and highly capable human pathogen

“…This finding is consistent with evidence of positive selection operating on Orf1ab in Middle East Respiratory Syndrome (MERS) [ 44 ], and Spike and N proteins being essential for immune recognition. All segments show a nontrivial signal of synonymous rate site-to-site variation, highlighting the importance of accounting for synonymous rate variation in this study, as recent literature [ 45 ] indicates that when the coefficient of variation (CV) for the distribution of site-to-site synonymous rates exceeds 0.5. Importantly, not accounting for synonymous rate variation risks mistaking it for evidence of selection.…”

“…Some synonymous rate variation may also be attributed to the 5′ and 3′ context-specific mutation rate variation observed in SARS-CoV-2 [ 47 ]. Because multinucleotide mutations could be taking place in SARS-CoV-2 genomes, e.g., trinucleotide substitutions in N at positions 28881–28883 [ 48 ]), and not considering them in models can lead to false positives in certain cases [ 49 ], we modified BUSTED[S] to include them and to estimate the extent of multinucleotide mutations in the SARS-CoV-2 data [ 45 ]. For 11 out of 20 segments ( S3 Table ), this model provides a better information theoretical fit to the data, including 2 of 5 segments where BUSTED[S] found evidence of positive selection (N and ORF1ab/regE).…”

“… For each segment, 2 rows of numbers are presented: top—for the BUSTED[S]-HMM model and bottom—for the BUSTED[S] model, allowing for multinucleotide substitutions (see Lucaci and colleagues [ 45 ] for details). The model with the better AIC-c and significant test results ( p -value) are shown in bold.…”

Selection, drift, and constraint in cypridinid luciferases and the diversification of bioluminescent signals in sea fireflies

“…At the branch and site levels, all three aBSREL, FEL and MEME tests incorporate site‐level heterogeneity into their models. Finally, at single sites, we used multihit (Lucaci et al., 2020) to assess the evidence supporting a double or triple substitution over a single substitution at every site (i.e., incidences of MNS).…”

Elevated rates of positive selection drive the evolution of pestiferousness in the Colorado potato beetle (Leptinotarsa decemlineata, Say)

“…As designed, the aBSREL test does not test for changes in specific codons, but instead the proportion of sites that have evolved under positive selection for each branch, and thus, it should be less sensitive to multinucleotide mutations. However, we still screened our data for such a bias by running multinucleotide mutation-aware models on the 673 putative insecticide-resistant genes using an updated aBSREL program in Hyphy v2.5.9(Lucaci et al, 2020). The likelihood values of the multinucleotide mutation model were compared with single nucleotide mutation model using the Akaike information criteria (AIC).Putatively selected genes were subsequently grouped by GO terms using the Leptinotarsa decemlineata official gene set and the Gene Ontology project v1.2 database(Ashburner et al, 2011;Schoville et al, 2018; The Gene Ontology Consortium, 2019;Yang, 2007).…”