[R-sig-ME] Maternal Genetic Effects in MCMCglmm

[Eryn McFarlane]

Hi Jarrod,

Thanks so much for your help so far.

I tried your advice but I get an error message that I don't understand:

Error in MCMCglmm(LRS ~ 1, random = ~dam + animal, ginverse = list(animal = invA,  : 
  animal ginverse appears but pedigree has been passed

I think that this means that I have replaced my pedigree with the ginverse? Is there something that I am missing to make sure that both are included in the model?

Below is some of my code. I hope that I am making a blatantly obvious mistake.

data$dam<-factor(data$dam, levels=levels(data$animal))

invA<-inverseA(ped)

model.LRS.a<-MCMCglmm(LRS~1, random=~dam+animal, ginverse=list(animal=invA, dam=invA),  ped = ped, data=data, prior=prior.LRS.a, nitt=nitt, thin=thin, burnin=burnin, verbose=FALSE, family="poisson", pr=FALSE) 

Thank you for your help,
Eryn



On 2011-10-11, at 4:27 PM, Jarrod Hadfield wrote:

> Hi Eryn,
> 
> You can now fit  maternal genetic effects by using the ginverse argument:
> 
> ginverse=list(animal=invA, dam=invA)
> 
> invA is the inverse relationship matrix which you can get using inverseA.
> 
> you need to make sure that the dam factor has the same levels as animal even if they are not represented:
> 
> your_data$dam<-factor(your_data$dam, levels=levels(your_data$animal))
> 
> Two problems you may have are:
> 
> i) slow speed and mixing
> ii) it is not yet possible to fit the covariance between maternal genetic and direct genetic effects
> 
> Cheers,
> 
> Jarrod
> 
> 
> 
> 
> On 11 Oct 2011, at 20:55, Eryn McFarlane wrote:
> 
>> 
>> Dear List,
>> 
>> I hope that you can help me. I would like to use MCMCglmm to estimate Maternal Genetic Effects.
>> I think that what I want to do is use an A matrix for the maternal effects, and I think that I want to do this by linking my "dam" random effect to the pedigree somehow.
>> Does anyone know if there is a way to do this that is already written into MCMCglmm? I have not seen anything previously in the list about this, so if it exists else where, then that would be excellent as well.
>> 
>> Thank you for your help,
>> 
>> Eryn
>> 
>> Eryn McFarlane
>> MSc Candidate
>> Department of Integrative Biology
>> University of Guelph
>> Guelph, ON, Canada N1G 2W1
>> 
>> _______________________________________________
>> R-sig-mixed-models at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
>> 
> 
> 
> -- 
> The University of Edinburgh is a charitable body, registered in
> Scotland, with registration number SC005336.
>