Chemistry Reference
In-Depth Information
van der Waals, electrostatics and desolvation correlation functions as well as
spherical harmonics-based docking. Other programs use coarse-grained
potentials/decoys.
Effectively the scoring functions use chemical properties, solvation terms,
evolutionary conservation and shape complementarity. They can be divided into
empirical, forcefield-based and knowledge-based. Compared to receptor-drug
complexes, with concave/convex interfaces, the flat interfaces in some protein-
protein interactions makes shape complementarity harder to detect.
Recent work reports a protocol that the authors suggest improves the state of the
art in initial stage rigid body exhaustive docking search, scoring and ranking.
Improvements are introduced in electrostatic affinity functions; shape-
complementarity, novel knowledge-based filters, GBSA reranking procedure and
a new knowledge-based interface propensity term with fast Fourier transform
formulation. Algorithms including dynamic packing grids can be used to speed up
the calculations with error bounds [788].
Targeting protein-protein interactions is an attractive new approach to drug design
due to increasing available information regarding structures of protein complexes
and signaling pathways. In this approach drugs must be designed that can
specifically interfere with, for example, dimer and oligomer formations (multimer
protein forms), cause disruption of antibody-antigen interactions (fast
pharmaceutical segment) and alter a signaling pathway by targeting protein
interactions [787].
Different approaches should be introduced since the binding sites differ from
traditional small ligand-protein interactions. It is noteworthy, there are already
various databases of protein-protein inhibitors. Nonetheless, for designing drugs
to interfere with protein-protein interaction there are many challenges. Extensive
databases of starting structures are required. The type of small ligands that
comprise the traditional protein-ligand databases are inappropriate since typical
protein-protein interfaces are approximately 750-1500 A
2
. In addition, more
shallow surface areas define protein-protein interfaces.
