Back
to index
Constraints
Site constraints:
Site constraints can either penalize or reward a residue for being
at the interface, and decoys that fail to meet the constraint are
rejected. A common use is when you are using a fragment of
a multidomain protein for one docking partner. In this case, you
might block the N and C termini of the fragment to reflect the fact
that in reality, there is another part of the protein there.
A common case to use a rewarding site constraint is if experimental
evidence has identified a given residue(s) as having a high probability
of being at the interface. You can use a rewarding constraint to
force that residue to be at the interface. My experience with
rewarding constraints,
however, is that they lead to a narrow diversity of predicted
structures. A better strategy is to make a loose distance
constraint between that residue and some other residue on the other
docking partner.