I’ll be posting R code that has been used in my papers on this page. click on the image to download the code.

MECCA is an R package that uses Approximate Bayesian methods to estimate rates of trait evolution from incompletely sampled comparative data. A paper describing the approach is in press at Evolution at the moment. The package comes with an example dataset for carnivore body masses. Note that to do the posterior regression adjustment, you will need to download the package ABCtoolbox. I’ll be posting a more detailed tutorial on how to use MECCA in the near future.

This download is a folder containing a modified version of geiger’s disparity through time code (see Harmon et al. 2003, Science) that produces a DTT plot with pretty shaded 95% credible intervals. The code will also return a p-value for the morphological disparity index (MDI). This approach was described in our paper on Cetacean body size evolution (Slater et al. 2010 Proc Roy Soc). The data from that paper and an example script are also provided.

Significance Testing with Disparity Through Time

MECCA - modeling rates of trait evolution with incompletely sampled trees and data using Approximate Bayesian Computation

fitContinuousMCMC is a Bayesian version of the geiger fitContinuous function that used Maximum Likelihood to fit models to comparative trait data. In this MCMC version, ancestral state values can be estimated under the available models (BM, SSP, Trend, ACDC exponential and linear) and informative node priors can be integrated, for example from the fossil record. Our paper at Evolution shows that these in particular can have a huge impact on model selection performance.

fitContinuousMCMC - fit models of continuous trait evolution to comparative data while integrating fossil information as informative node priors