Hi all,
I am interested in comparing estimation of genetic parameters for binary
traits under various models (animal, sire, dam, sire-dam). However, I am
under some confusion as to the appropriate pedigree file to use for each
model, given that the pedigree file for a sire model is described in the
manual in terms of a sire-MGS structure.
It seems to me that providing the full pedigree (as in an animal model)
also provides the necessary pedigree information for sire, dam and sire-dam
models. Consequently, I assume that the same pedigree file may be used for
all models, with the only change required being in the model statement for
the trait.
For example:
read in:
animal !P
sire !P
dam !P
Animal model defined by: trait ~ fixed effects !r animal
Sire model: trait ~ fixed effects !r sire
Dam model: trait ~ fixed effects !r dam
trait ~ fixed effects !r sire dam (similar outcome to animal model given
full pedigree)
You could also consider dropping relationships between sires or dams by
using ide(sire) etc. rather than sire, but this is probably not appropriate
under selection.
My question is, will the pedigree provided for an animal model be correctly
used under the alternative pedigree models - the additional information in
the pedigree file for animal models simply resulting in redundant elements
in ainverse for sire or dam models?
Does anyone have a comment?
Cheers
Kim
Kim Bunter
PhD Student
Animal Genetics and Breeding Unit
University of New England
Armidale, NSW, 2351
AUSTRALIA
Ph: (02) 6773 3788
Fax: (02) 6773 3266
email: kbunter@metz.une.edu.au
--
Asreml mailinglist archive: http://www.chiswick.anprod.csiro.au/lists/asreml