Simple methods for testing the molecular evolutionary clock hypothesis.

TLDR

The authors develop simple statistical tests to evaluate the molecular evolutionary clock hypothesis for both nucleotide and amino acid sequences. These tests employ chi‑square statistics that remain valid when substitution rates vary among sites or when the outgroup is unknown, and were benchmarked by simulation against the likelihood ratio and relative‑rate tests. Simulation results show that the new tests achieve power comparable to the likelihood ratio and relative‑rate tests, demonstrating their practical usefulness.

Abstract

Abstract Simple statistical methods for testing the molecular evolutionary clock hypothesis are developed which can be applied to both nucleotide and amino acid sequences. These methods are based on the chi-square test and are applicable even when the pattern of substitution rates is unknown and/or the substitution rate varies among different sites. Furthermore, some of the methods can be applied even when the outgroup is unknown. Using computer simulations, these methods were compared with the likelihood ratio test and the relative rate test. The results indicate that the powers of the present methods are similar to those of the likelihood ratio test and the relative rate test, in spite of the fact that the latter two tests assume that the pattern of substitution rates follows a certain model and that the substitution rate is the same among different sites, while such assumptions are not necessary to apply the present methods. Therefore, the present methods might be useful.