MEMM

MIXED EFFECTS MATING MODEL (MEMM)

 

Supporting scientific articles

MEMM implements a Bayesian statistic method first presented in Klein et al. (2008). It estimates the pollen dispersal function and the variance in male fecundity on the basis of spatial information (positions of sampled plants, positions of all putative fathers in the study plot) and genetic information (genotypes of the sampled plants, putative fathers and sampled seeds).
Klein et al. (2011) investigates the performance of the method on simulated datasets and illustrate how neglecting the random component in male fecundity can lead to erroneous conclusions.

Klein, E.K., Desassis, N., Oddou-Muratorio, S. (2008) Pollen flow in the wildservice tree, Sorbus torminalis (L.) Crantz. Whole inter-individual variance of male fecundity estimated jointly with dispersal kernel. Molecular Ecology. 17: 3323-3336.
Klein, E.K., Carpentier, F.H., Oddou-Muratorio, S. (2011) Estimation of the whole variance of male fecundity from genotypes of progeny arrays: accuracy and limits of the Bayesian forward approach. Methods in Ecology and Evolution. 2, 349–361

Version v2.0

Version 2.0 changes :
- Parameter file change to xml format.
- A GUI is added to fill parameter file.
- Code factoring.
- Packaging tools.

Documentations, sources and binaires for Linux, MacOS and Windows are available here.

Beta version v2.0 (for Rocio...)

Download the zipped file

 

Version v1.1

Download version v1.1 for MacOSX, for Windows ofr for Linux

Download the pdf documentation below

What's new in version v1.1

·      A bug in version V1.0 was fixed. This bug was leading to excessively low values of the shape parameter b and excessively high values of the dispersal distance d and of the variance of fecundity dobs/de.
·      The gamma distribution is now implemented
·      In is now possibility to define weighting variables (Be careful, Parameters.txt needs one more information even when not using weighting variables)
·      It is no more necessary to define a useless file name when allelic frequencies are computed from adult individuals.
·      The CI for the parameter 1 was corrected.
 

Help needed ?

Do not hesitate to contact us for any question or comment about MEMM. You can mail us at Etienne Klein and/or Sylvie Oddou-Muratorio and Jean-François Rey (development question only). You can also leave a comment below.