# Repeated measures interaction betwen SNP and treatment with R

This is my first question is this forum. I have the a problem that I want to solve with the lme function of the nlme package of R.

The goal of my study is analyse if the cholesterol level (quantitative) change in a series of patients depending of 3 different treatments and the genetic profile (SNPs, a factor variable with three categories). We have two repeated measures of the cholesterol (before the intervention and 1 year after).

The general question is to know if some of the treatment "works" (change in the cholesterol) for some of the 3 categories of the SNP.

My analysis in R is the next, but I don't know if it is correct:

w.lme.1 <- lme( colesterol ~  tiempo * treatment + SNP ,
random = ~1 | ID , data=xx, na.action=na.omit, method="ML")

w.lme.2 <- lme( colesterol ~  tiempo * treatment * SNP ,
random = ~1 | ID , data=xx, na.action=na.omit, method="ML")

anova(w.lme.1, w.lme.2)


where:

• ID , identification of the individuals
• time, 2 factors : baseline time and 1-year after
• treatment, 3 levels
• SNP, 3 levels

R output:

> head(xx)
ID colesterol tiempo treatment SNP
1   1     312.47      0         1   1
2   1     221.61     12         1   1
3  10     221.33      0         1   1
4  10     227.67     12         1   1
5 100     229.07      0         2   1
6 100         NA     12         2   1
> str(xx)
'data.frame':   594 obs. of  5 variables:
$ID : int 1 1 10 10 100 100 101 101 102 102 ...$ colesterol: num  312 222 221 228 229 ...
$tiempo : Factor w/ 2 levels "0","12": 1 2 1 2 1 2 1 2 1 2 ...$ treatment : Factor w/ 3 levels "1","2","3": 1 1 1 1 2 2 1 1 3 3 ...
\$ SNP       : Factor w/ 3 levels "0","1","2": 2 2 2 2 2 2 1 1 1 1 ...

-
Welcome to the site! I removed your signature from your post as the site adds one automatically. – Peter Flom Oct 16 '12 at 10:03
Please could you post the outputs of str(xx) and head(xx) so that we see the essential of your data ? – Stéphane Laurent Oct 16 '12 at 11:30
Thanks, Stephane. I have created a dataframe with the variables I use in this question, and I put the info – Jesus Herranz Valera Oct 17 '12 at 9:05