[R] How to run mixed model with related independent variables

[Sariya, Sanjeev]

I've data that look like:

Outcome             V1_AA  V1_EU  V1_NA  V2_AA  V2_EU  V2_NA
0              0.046     1.001     0.954     0.045     1.001     0.954
0              0.007     1              0.993     0.007     1              0.993
1              1.774     0.217     0.009     1.774     0.217     0.009
1              0.004     1.996     0              0.004     1.996     0
1              1.001     0.997     0.002     1.001     0.997     0.002
0              0.94        0.998     0.061     0.94        0.998     0.061
0              0.587     1.407     0.006     0.587     1.408     0.006
1              0.019     1.978     0.003     0.018     1.979     0.003

Column Outcome is dependent variable. It's dichotomous either 0 or 1.
Column 1st, 2nd, 3rd are three components of V1, similarly next three for V2. Sum of V1_AA, V1_EU and V1_NA will be two at each row. Similarly for V2 variable's components.
Here the three predictors (I'm referring as component) are related. If AA is higher then, NA and EU will be lesser.

I cannot simply use multivariate regression as:

Y ~ V1_AA + V1_EU + V1_NA

How do I proceed with these data?

Thanks!


	[[alternative HTML version deleted]]