Question about YANK restrains

[mmagithub]

Hi, I am calculating the binding affinity of a small molecule sandwiched between two proteins. In the absence of the small molecule, these two proteins do not bind under physiological conditions. I am observing that YANK somehow overestimates the binding affinities of such ligands. I am thinking this could be due to an extra stabilization effect (in addition to the usual binding affinities) and wondering if there could be a smarter workaround (if it is a real problem). I am using a standard Imatinib like input parameters with harmonic restrains?

I am also asking if we can directly compare experimental values (deltaG values from Kd or IC50 data) to YANK outputs ?

Any suggestions, Thanks