Use of residue pair potentials to rerank docked complexes
After the global search for shape complementarity and favourable electrostatic interaction, a list of several putative complexes is generated. A model close to native (see below for quantitative measures) is rarely at the top ofthe list of complexes ranked by shape complementarity. It is necessary, therefore, to develop further scoring functions to identify good solutions from the list of alternatives. The strategy we have developed is to use a residue-residue scoring scheme to evaluate the docked complexes. Such an approach has an energy landscape that is smoother than an approach based on all (heavy) atoms but incorporates information that cannot be included in the grid-based first step. The terms to quantify the residue-residue interactions are derived empirically from known protein structures and these terms are often known as empirical pair potentials.
The theory of empirical pair potentials is that since they are derived from observations they will incorporate the dominant thermodynamic effects without explicitly having to partition and quantify the effects. Empirical pair potentials have been widely used in protein structure prediction, particularly in fold recognition following their introduction by Sippl  and their use
Continue reading here: Info
Was this article helpful?