multivariate genetic analysis using SPLUS asreml

From: Derek Roff <derekr_at_UCR.EDU>
Date: Wed, 17 Oct 2007 10:14:45 -0700

I am having a lot of trouble fitting the multivariate model
and calculating the appropriate values for the standard
errors of the correlations.

Does anyone have SPLUS or R coding to do this?

I am using the command

HS.asr <- asreml(cbind(TRAIT1,TRAIT2)~1,
                  random=~us(trait):ped(Animal)+ Dam + Cage,
                  ginverse=list(Animal=AINV), data=Data)

I am disturbed because the answer I get is very sensitive to
scaling (e.g.. I get different answers if I multiply TRAIT2
by 10 vs 100).

A sample of the HS.asr$varcomp output is
component std.error
trait:ped(Animal)!trait.TRAIT1:TRAIT1 5.460194e-001
5.460194e-001 0.09656404
trait:ped(Animal)!trait.TRAIT2:TRAIT1 9.132571e-002
9.132571e-002 0.04712336
trait:ped(Animal)!trait.TRAIT2:TRAIT2 1.164721e-001
1.164721e-001 0.03610538
                                  Dam 7.626141e-008
7.626141e-008 NA
                                 Cage 1.452081e-001
1.452081e-001 0.02356835
                           R!variance 1.000000e+000
1.000000e+000 NA
                R!trait.TRAIT1:TRAIT1 1.149725e+000
1.149725e+000 0.07331402
                R!trait.TRAIT2:TRAIT1 3.243728e-001
3.243728e-001 0.03528615
                R!trait.TRAIT2:TRAIT2 5.165126e-001
5.165126e-001 0.02799466

trait:ped(Animal)!trait.TRAIT1:TRAIT1 5.654480 Unconstrained
trait:ped(Animal)!trait.TRAIT2:TRAIT1 1.938013 Unconstrained
trait:ped(Animal)!trait.TRAIT2:TRAIT2 3.225892 Unconstrained
                                  Dam NA Boundary
                                 Cage 6.161147 Positive
                           R!variance NA Fixed
                R!trait.TRAIT1:TRAIT1 15.682198 Unconstrained
                R!trait.TRAIT2:TRAIT1 9.192636 Unconstrained
                R!trait.TRAIT2:TRAIT2 18.450394 Unconstrained

It seems strange that gamma and component have the same
value. Am I missing something?

Thanks for any help you can give me.
Derek Roff
Derek Roff,
Department of Biology,
University of California,
CA 92521
Received on Sun Oct 17 2007 - 10:14:45 EST

This webpage is part of the ASReml-l discussion list archives 2004-2010. More information on ASReml can be found at the VSN website. This discussion list is now deprecated - please use the VSN forum for discussion on ASReml. (These online archives were generated using the hypermail package.)