Statistical methods

• QGglmm

This R package to infer quantitative genetic parameters from Generalised Linear Mixed Models (GLMM), e.g. for non Gaussian phenotypic traits. It allows for the computation of quantitative genetic parameters such as additive genetic variance, heritabilities, intra-class correlation coefficients, G-matrices (for multivariate analyses), but also for evolutionary predictions. The package is available on CRAN. The source code is available on GitHub. Please report issues here.

• BayeScEnv
The new environmental version of BayeScEnv can be found on GitHub. This genome scan method allows to detect high population differentiation (i.e. high FST) due to a high environmental differentiation (local adaptation scenario). It is also less error-prone than BayeScan.


• Estimating heritability using MCMCglmm

This tutorial is intended for students or researchers in the domain of evolutionary ecology, interested in using the animal model to estimate the eritability of biological traits in a wild population. It aims at bringing theoretical and practical help on three main issues: (i) understanding what heritability is, what it quantifies and how the animal model works; (ii) learning by practice how to implement animal models using the MCMCglmm R package; and (iii) introducing Bayesian statistics (priors, Markov Chain Monte Carlo, etc.).