[R-sig-ME] glmer vs. MCMCglmm

[David Duffy]

On Wed, 1 Apr 2009, Hervé CHAPUIS wrote:

> I intend to estimate the genetic variance for a binary trait (disease 
> resistance) in a simulated population.
> First, I have used lmer and glmer.  However, sometimes I have false 
> convergences, and  I can't specify "nAGQ=5" when fitting two random effects 
> (sire AND dam).

I presume you have data for just two generations, then?

> This is the reason why I have decided to give a glimpse at MCMCglmm.
>
> I am still trying to implement an animal model as specified in the MCMCglmm 
> manual, but I can't figure out how an heritability can be estimated so high 
> without a big mistake. But I can't see it.
>

Heritability for binary traits is a pretty unsatisfactory concept, but I 
don't see how to get it over 1 ;)  I presume you're following the vignette 
for PlodiaRB, where priors are discussed, but fitting the animal model. 
Does your job work for that dataset?

Cheers, David Duffy.

PS You may be interested in comparing results from the MCMC GLMM for 
pedigree data I have included in my Sib-pair program 
(http://www.qimr.edu.au/davidD).

-- 
| David Duffy (MBBS PhD)                                         ,-_|\
| email: davidD at qimr.edu.au  ph: INT+61+7+3362-0217 fax: -0101  /     *
| Epidemiology Unit, Queensland Institute of Medical Research   \_,-._/
| 300 Herston Rd, Brisbane, Queensland 4029, Australia  GPG 4D0B994A v