R help - Nov 2011 - Help to inputting a pre-defined correlation structure in a Mixed Model

I'm working in a Gen/Marker-Phenotype association study in wheat and I'm
using a Mixed Model Approach to estimate the effect of the markers. My model
has the atribute measured as y (response variable), the markers and the
blocks (of a complete random block design) as fixed and the genotypes and
the residuals as random. In one hand I'm assuming that there is no
correlation between residuals and that they have a normal distribution with
mean equals to zero and a common variance equals to (sigma e)^2. On the
other hand I'm also asumming that the genotypes are normally distributed
with mean equals to zero but with variance equals to a G matrix that
considers the relatdness between individuals. This G matrix is expressed as
2 * (sigma g)^2 * K, where K is a matrix with the reladness coefficients
(kinship) between all individuals in the study. The K matrix is calculated
outside of R environment, and I'm using it as an input. I'm using the
lme
function of the nlme library in R, but I want to know how can I gap the
structures that R offers me (as corrAR1, corrSumm, etc.) as default
structures, because I already have this structure and I only have to
estimate the (sigma g)^2 parameter, possibly using REML. I'm new in the use
of mixed models and I will be very pleased if someone could help me with
this particular question. Thanks a lot in advance. 

--
View this message in context:
http://r.789695.n4.nabble.com/Help-to-inputting-a-pre-defined-correlation-structure-in-a-Mixed-Model-tp4097147p4097147.html
Sent from the R help mailing list archive at Nabble.com.