[BioC] LIMMA MZ Twin design

[Mario Falchi]

Hi,

I'm trying to use Limma to analyse some nimblegen methylation data on MZ 
twins (imported through Ringo).
The design is:
Two colours microarray. One channel with a twin, the other channel with 
the other twin
After 22 arrays we will observe the same pair (a twin for channel) but 
10 years later:

1a 1b
2a 2b
....
1aAfter10 1bAfter10
2aAfter10 2bAfter10
...

I'd like to see:
How similar are the MZ twins within each array 1a Vs 1b, etc
How they change after 10 years (diverge? converge?)
How each individual changes after 10 years (1a against 1aAfter10)

The data are organized in regions of interest, therefore a variable 
number of probes exist for each region.
It would be interesting comparing the methylations a whole, but also 
comparing the different region of interest.
I find extremely difficult generating the design for this experiment and 
I'd be extremely grateful if you could help me.

Best regards,
Mario