We developed a new method named Smoothed Atomic Volume Percentage (SAVP) to describe geometric properties of protein surface. SAVP was an improved algorithm based on the idea of protrusion index, and this new method was proved to be effectively on the representation of the properties in convex or concave region on protein surface. Another work in this paper was the fine structure description of protein interface, which described protein interface into two different remarkable regions, marginal interface and inner interface. Then we analyzed several geometric properties, such as protrusion index, roughness and ASA, among these different interface regions using the known data set, including homodimers, convex interface heterodimers and concave interface heterodimers. The statistical results demonstrated that significant distribution difference of those geometric properties mentioned above was observed among marginal interface, inner interface as well as non-interface surface separately. Dramatically, we found that difference of SAVP was more significant than the other, therefore SAVP might be used as a new tool for prediction of potential protein interface.
|