For paired data analysis using limma with multiple factor, if I was interested in the interaction effect between Condition and Time would the structure look like this?

mm <- model.matrix(~Patient + Condition*Time + Batch)

This design is not full rank:

a) condition and time is the same, drop one of it

b) if you're doing a paired analysis then you're anyway comparing within patient so batch is not needed as the two samples per patient are always in the same batch.

Do ~Patient+Condition to test condition, accounting for patient differences.