[R-sig-ME] indirect maternal genetic effects

[Jarrod Hadfield]

Hi,

You are correct. You are need to associate the inverse with both terms:

invA<-inverseA(ped)$Ainv


m1<-MCMCglmm(..., random=~dam+animal, ginverse=list(animal=invA, dam=invA),  ...)

if you want to estimate the direct-maternal genetic covariance you need:


m2<-MCMCglmm(..., random=~str(dam+animal), ginverse=list(animal=invA, dam=invA),  ...)

Cheers,

Jarrod
  


On 19/12/2017 16:26, Walid Mawass wrote:
> Hello everyone,
>
> I am trying to model direct genetic effects and indirect maternal 
> genetic effects on infant survival, using MCMCglmm. From what I know 
> and understood from the ASReml code on how to model this, I should 
> include the identity of my focal individual ("animal") to estimate the 
> direct genetic effect, include the identity of his mother ("dam") to 
> estimate the maternal environment effect and then add another variable 
> which should represent the genetic effect of the mother on the focal 
> individual. I assumed that I should add the same "dam" variable but 
> give it its own incidence matrix through ginverse(), but I am not sure 
> about this. I would appreciate some help on how to code this 
> relationship.
>
> Thank you
>

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <https://stat.ethz.ch/pipermail/r-sig-mixed-models/attachments/20171221/8d35b1bc/attachment.ksh>