Id say that trying to 'explain the irreducible complexity' should be your opponent job in as much as he needs to show that some exists at all. The other major requirement for that would be to have a rigorous definition of IC, I know that Behe himself has had a couple of different ones.
One issue which will clearly heavily impact any discussion of signalling is crosstalk. Very many signalling pathways use the same downstream effectors and consequently there can be considerable interplay between different pathways and the loss of one element of a particular signalling pathway need not lead to a total loss of function for the upstream receptor.
I don't think you should be required to try and explain away the supposed irreducible complexity of a system until it has actually been shown to be irreducibly complex.
TTFN,
WK
™ Version 4.2
