I have the data for both morphological and reproductive traits for 22 species and the phylogenetic tree built by maximum likelihood with branch length. I need to estimate the evolution rate between these two trait categories to show whether reproductive traits evolve faster than morphological traits. Which parameter I should calculate it for comparison? dose any body have R codes for doing this kind of comparison?
Any hints or recommendations?
Thanks everyone!
What kind of characters do you have? Are they all continuous? If so you could estimate the rate parameter (sigma) of the Brownian Motion model using AUTEUR, for example. It depends on your data, but your phylogeny is pretty small... see this publication attached
http://sysbio.oxfordjournals.org/content/62/2/181
Thank you Dr. Provete for your suggestion.
All my data are continuous trait like wing length, wing width for morphological & testis size for reproductive traits. As you mentioned parameter (sigma) you mean by Variance? Please can I have your email address, I prefer to be in contact with you through email, and this is my email:
It's variance, but I think it's not what you're imagining. Take a look at Brian O'Meara's papers about Brownie to understand the Brownian Motion model
You could do an ancestral state reconstruction (ASR) of the two traits on the phylogeny and compare the outcomes. Programs like BayesTraits and BEAST will do ASRs for you. If your traits are continuous, you may use a Brownian motion prior (or Gamma or Cauchy or other if you need an asymmetrica distribution), or if your traits are discrete you may use a symmetrical or asymmetrical transition rate model prior.
I think the best way to obtain evolutionary rates between traits is to compute the phylogenetic independent contrasts for all traits and computing the covariance matrix of those contrasts. The resulting covariance matrix will contain the maximum likelihood estimate of the brownian motion rate parameter on the diagonals and covariances among characteres on the off-diagonals.
Is not as elaborate as more model-based approaches, but it gives the same result in the case of browniam motion and, because you compute covariances, you can say something also about trait association.
On R this can be done as follows
#####################
library(ape) #loaging ape package
data # data matrix with species on lines and traits on collums
phylo #your phylogeny
pics
- Badges
- Science topic
- Similar topics
- Biological Science
- Evolutionary Biology