General scoring functions for PP, PL, and LL

[Guosheng Wu <wu_guosheng2002=at=yahoo.com>]

Hi there,
 In the computation of protein(P)-ligand(L) binding
 free energy, one would like to have a quick and
 accurate scoring function to evaluate all of those
 non-bond interactions between atomic pairs of PP, PL,
 and LL, for the strain energy of the target, binding
 energy, and strain energy for ligand, respectively. So
 that a more reasonable comparison with the
 experimental activities can be justified.
 However, things are often very complicated. It seems
 most of current scoring functions are focused on the
 PL part. For the LL part, simple energy difference
 (between the solution state & binding state) from MM
 is often claimed to be good enough. For the PP part,
 it may be more difficult to quantify and validate, or
 it simply adds more noise rather than useful
 information. One may argue that the PL part is usually
 the dominant component. Still it's not that simple in
 many cases, otherwise, no one would really care about
 the business of flexibility.
 Does anyone notice any kind of "general" approach for
 this task? Physically, should there be much difference
 for these three kinds of potentials?
 I would like to get your comments or reminding of
 relevant literature.
 Thanks a lot.
 Guosheng
 __________________________________________________
 Do You Yahoo!?
 Tired of spam?  Yahoo! Mail has the best spam protection around
 http://mail.yahoo.com