Installing ASReml

ASReml 3.0 and ASReml-R are powerful software packages for mixed model analyses. Installing them is not always so straightforward, though. Therefore I have put together a brief manual, in which I describe in some detail what to download where, how to install both the standalone ASReml and ASReml-R and what to do with the license file, both for Windows and Mac OS. This manual was written for students participating in ECO334: Mixed effects modelling with R and ASReml, but because I figured it might be of interest to others as well, it is also available here (pdf).


Test for isolation by distance with MantelPiece.R

Mantel tests are often used to test whether there is a significant relationship between geographic and genetic distance. However, sometimes we want to compare the strength of isolation by distance (i.e. the slope) for two different types of markers. So far people have been using IBD (Bohonak, 2002) or more recently IBDWS (Jensen et al., 2005) to obtain 95% confidence intervals using bootstrapping or jackknifing, and to subsequently test whether the confidence intervals for the two slopes are overlapping. MantelPiece does something similar, but additionally provides 95% confidence intervals for the difference in slopes. This pairwise comparison should provide additional statistical power. Furthermore, MantelPiece performs a randomisation test (sensu Mantel) to test whether the slopes are significantly different. The latter seems to provide most statistical power.



Postma et al. 2011 Sex ratio R scripts

These are the two supplementary R scripts, accompanying Postma et al. 2011. Disentangling the effect of genes, the environment and chance on sex ratio variation in a wild bird population. Proc. Roy. Soc. B. For more details, read the paper and have a look at the scripts.

Postma_et_al_Supp_R_code_1.R contains the code to run the MCMCglmm analysis.

Postma_et_al_Supp_R_code_2.R contains the code to quantify and visualise any overdispersion in sex ratio.

Please e-mail me with any comments, questions or critisisms you may have.