Re: specifying linear contrasts in ASReml-R

From: Arthur <asremlforum_at_VSNI.CO.UK>
Date: Wed, 18 Feb 2009 02:10:29 +0000

Dear Andre,

My response relates to how you do this in Standalone ASReml but it should point the way.

To simply test a contrast, ASReml has the qualifier

!CONTRAST gdom genotype 0.5 0.5 -1

assuming genotype codes the levels A B AB

Then you can include gdom in the model.
Maybe also define
!CONTRAST gadd genotype -1 1 0
and include both terms in the model.

Now your query discussed the predict statement.

Given that your model is as you have (based on 'genotype',
You can get ASReml to calculate the contrast (and its SE hence a t test in the fixed context) using the weights

PREDICT DPP !weight genotype 0.5 0.5 -1

This should point in the right direction.


Arthur Gilmour

Principal Research Scientist (Biometrics)
NSW Dept Primary Industries

-------------------- m2f --------------------

Sent using Mail2Forum (

Read this topic online here:

-------------------- m2f --------------------
Received on Thu Feb 18 2009 - 02:10:29 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.)