A likelihood method for detecting trait-dependent shifts in the rate of molecular evolution

Itay Mayrose, Sarah P. Otto

Research output: Contribution to journalArticlepeer-review


Rate heterogeneity within groups of organisms is known to exist even when closely related taxa are examined. A wide variety of phylogenetic and dating methods have been developed that aim either to test for the existence of rate variation or to correct for its bias. However, none of the existing methods track the evolution of features that account for observed rate heterogeneity. Here, we present a likelihood model that assumes that rate variation is caused, in part, by species' intrinsic characteristics, such as a particular life-history trait, morphological feature, or habitat association. The model combines models of sequence and character state evolution such that rates of sequence change depend on the character state of a lineage at each point in time. We test, using simulations, the power and accuracy of the model to determine whether rates of molecular evolution depend on a particular character state and demonstrate its utility using an empirical example with halophilic and freshwater daphniids.

Original languageEnglish
Pages (from-to)759-770
Number of pages12
JournalMolecular Biology and Evolution
Issue number1
StatePublished - Jan 2011
Externally publishedYes


  • character mapping
  • dating
  • evolutionary models
  • molecular clock
  • rate shift, substitution rates

All Science Journal Classification (ASJC) codes

  • Ecology, Evolution, Behavior and Systematics
  • Molecular Biology
  • Genetics


Dive into the research topics of 'A likelihood method for detecting trait-dependent shifts in the rate of molecular evolution'. Together they form a unique fingerprint.

Cite this