Abstract
Based on the results of several groups using different docking methods, the key properties that determine the expected success rate in protein–protein docking calculations are measures of conformational change, interface area, and hydrophobicity. A classification of protein complexes in terms of these measures provides a prediction of docking difficulty. This classification is used to study the targets of the CAPRI docking experiment. Results show that targets with a moderate expected difficulty were indeed predicted well by a number of groups, whereas the use of additional a priori information was necessary to obtain good results for some very difficult targets. The analysis indicates that CAPRI and other relatively large‐scale docking studies represent very important steps toward understanding the capabilities and limitations of current protein–protein docking methods. Proteins 2005;60:176–180.

This publication has 19 references indexed in Scilit: