Overview

Sunder is a computer program that estimates parameters in a statistical model that quantifies the relative effect of geographic versus ecological isolation on genetic differentiation. It is a modelbased alternative to the Mantel tests. The package consists mostly of five functions:
SimSunderData: data simulation under the geostatistical models with binomial/multinomial marginal distribution
MCMCCV: Bayesian inference with MCMC and model selection by crossvalidation under the model above
MLCV: maximum likelihood estimation and model selection by crossvalidation under the assumption of joint normality of the data
Reformat23: reformat a two dimensional matrix into a three dimensional array
Reformat32: reformat a three dimensional array into a two dimensional matrix
Installation

Open R and in the shell type:
install.packages('Sunder')
References

Botta, F., Eriksen, C., Fontaine, M. C., & Guillot, G. (2015). Enhanced computational methods for quantifying the effect of geographic and environmental isolation on genetic differentiation. Methods in Ecology and Evolution, 6(11), 12701277
Guillot G., R. Schilling, E. Porcu, M. Bevilacqua Validity of covariance models for the analysis of geographical variation. Methods in Ecology and Evolution Volume 5, Issue 4, pages 329335, April 2014.
Guillot G. and F. Rousset. Dismantling the Mantel tests. Methods in Ecology and Evolution. 4(4), 336344, 2013
Input and output

To use Sunder, you need
An array containing allele counts data. This should be an R array with dimensions (n,l,a) n: number of geographical locations, l: number of loci, a: max number of alleles
A matrix of geograpical distances
A matrix of environmental distances
Contact:

b i o s t a t i s t i c s @ ipri.org