[R-sig-ME] observation level random effects/kinship model

Ryan King c.ryan.king at gmail.com
Wed Mar 14 13:28:00 CET 2012


When the observation-level random effects are independent then they
are the same as the noise. i.e.
y = xb + u +e can just be rewritten y= xb + e', with e' = u+e. Since
the sum of two normals is normal, the model is unchanged from usual
OLS.

With kinship that symmetry breaks, and observation-level random
effects are identifiable.

.I am currently using R to do such genetics models to do association
.mapping. I ask to other people that have done that before me and if I
.understand well, no packages allows to specify such a variance/covariance
.matrix for a random effect except ASREML.

You can also use MCMCglmm and R-INLA.

Ryan King




More information about the R-sig-mixed-models mailing list