Accuracy and Power of Bayes Prediction of Amino Acid Sites Under Positive Selection

Abstract

Bayes prediction quantifies uncertainty by assigning posterior probabilities. It was used to identify amino acids in a protein under recurrent diversifying selection indicated by higher nonsynonymous (d(N)) than synonymous (d(S)) substitution rates or by omega = d(N)/d(S) > 1. Parameters were estimated by maximum likelihood under a codon substitution model that assumed several classes of sites with different omega ratios. The Bayes theorem was used to calculate the posterior probabilities of each site falling into these site classes. Here, we evaluate the performance of Bayes prediction of amino acids under positive selection by computer simulation. We measured the accuracy by the proportion of predicted sites that were truly under selection and the power by the proportion of true positively selected sites that were predicted by the method. The accuracy was slightly better for longer sequences, whereas the power was largely unaffected by the increase in sequence length. Both accuracy and power were higher for medium or highly diverged sequences than for similar sequences. We found that accuracy and power were unacceptably low when data contained only a few highly similar sequences. However, sampling a large number of lineages improved the performance substantially. Even for very similar sequences, accuracy and power can be high if over 100 taxa are used in the analysis. We make the following recommendations: (1) prediction of positive selection sites is not feasible for a few closely related sequences; (2) using a large number of lineages is the best way to improve the accuracy and power of the prediction; and (3) multiple models of heterogeneous selective pressures among sites should be applied in real data analysis.

Large-Scale Analyses of Positive Selection Using Codon Models

“…This procedure guaranties that the simulated dataset has the same distribution of parameters as the real data set, including potential confounding factors such as codon usage bias or long branches. These results and those of Anisimova et al (Anisimova, Bielawski et al 2002;Anisimova and Yang 2007) showed that the maximum likelihood estimate is robust to dS saturation, even for large divergence as shown by our simulations with doubled branch lengths. This is probably due to the use of more sequences, which "break" the long branches of the gene tree.…”

“…Simulations suggest a minimum number of six sequences in the alignment to have enough power and accuracy (Anisimova, Bielawski et al 2001;Anisimova, Bielawski et al 2002;Anisimova and Yang 2007). This is in itself a major issue for genomic studies involving too few species (e.g.…”

Population genetics of microbial pathogens estimated from multilocus sequence typing (MLST) data

“…Empirical results reported by Yang et al (2000a) and simulations by Anisimova et al (2001Anisimova et al ( ,2002 indicated that the LRTs and the inference of sites under positive selection do not seem to be sensitive to the assumed tree topology (a neighbor-joining tree in our analyses), even if a star tree is used. Hence, presumably, our results are not biased by whichever phylogenetic process (clonal, epidemic, or panmictic) drives the population structure of the studied pathogens.…”

A Maximum Likelihood Method for Detecting Functional Divergence at Individual Codon Sites, with Application to Gene Family Evolution

“…This impacts Bayesian site identification, as ML parameter estimates are used to compute the posterior probabilities. Simulation studies showed that low accuracy in Bayesian site identification occurs when sequence divergence is very low or too few sequences are sampled because under such conditions the sampling errors in ML parameter estimates are too high (Anisimova et al 2002). Similarly, suboptimal parameter estimates, based on local optimum, also could lead to low accuracy in Bayesian site identification.…”