trouble matching results with those from simple mixed model

[rkb965]

Hi, thank you so much for this valuable package.

(1) How do I specify covariates as adjustment variables without rMVP doing any processing (eg calculating PCs) on them?

When I use a simplified kinship matrix that is valid for lme4, I get nearly identical results between rMVP and lme4 when I do not include additional covariates. However, when I include additional covariates, my results differ substantially. I suspect that I am mis-specifying the rMVP model somehow.

My model is specified like this:

Covariates <- model.matrix.lm(~as.factor(breed)+as.factor(sex)+as.numeric(weight), data=yourdata, na.action = "na.pass")

MVP(
    phe=my_phe,
    geno=my_geno, 
    map=my_map, 
    K = my_kin, 
    CV.MLM = Covariates,
    maxLoop=3, 
    method=c("MLM"), 
    file.output=FALSE, 
    ncpus=1
  )

(2) My kinship matrix was calculated on methylation data and is centered around 0 (IQR -0.2, 0.2) with a few values that are large and positive (up to 15). Do you think this is problematic for model fit? Do you have any suggestions for either my kinship matrix or evaluating model fit?

Thank you so much for your time!