Globular proteins themselves form compact molecular structures, the packing density of the atoms in which is close to the density of packing of atoms in crystals. We suppose that for the existence of such properties in natural polypeptide chains, there must exist both the areas in which local interactions mainly take place, and the areas in which the interaction between remote chain residues mainly takes place. To test this prediction, we developed a program for analyzing contacts between residues in the polypeptide chain, depending on the local density of structural information. To estimate the density of the structural information, the earlier proposed in [1] method of analysis of the information structure of protein sequences was used. The developed program uses a set of files in PDB for input, which initially pass through a correction stage (insertions and deletions are corrected, the non-canonical residues names are restored, etc.) and then information structure of the regarded protein is calculated, and a matrix of contacts for residues belonging to the sections of the polypeptide with a different density of structural information is built. The program is able to build a matrix of contacts for residues of various information type, i.e. located at different distances from the local extremes in the information structure of the protein under study. This paper discusses in detail the contacts between residues located in the vicinity of the extremes of the information structure, i.e. residues of different information types. The developed program allows exploring the contacts between amino acid residues depending on their information types. It is shown that the information type of residue is an important factor in determining the amino acid residue ability to interact effectively with other amino acid residues of the polypeptide chain. The matrices of contacts were shown for residues of various information types, what confirms the hypothesis of the existence of the regions of the polypeptide chain implementing mainly short- or long-distance interactions within the protein molecule. The results obtained can be widely used in protein computer modeling works and in the design of natural protein molecules. 1. A. N. Nekrasov et al. (2004) Analysis of the information structure of protein sequences: a new method for analyzing the domain organization of proteins, J. Biomol. Struct. Dyn. 2004, 21(5), pp. 615-624. A. N. Nekrasov, A. A. Zinchenko (2010) Structural features of the interfaces in enzyme-inhibitor complexes. J Biomol Struct Dyn. 2010, 28(1), pp. 85-96.
Moscow Conference on Computational Molecular Biology (MCCMB'13) Moscow, Russia July 25–28, 2013

Specific protein-protein interaction is essential for the function of life systems. A variety of computational methods are being extensively used now-a-days to investigate this interaction and to identify structural features of binding sites. In this paper, the informational structure analysis method was applied to the study of protein-protein interaction interfaces in enzyme-inhibitor complexes. The analysis of amino acid sequence by informational structure analysis method reveals three types of sites (ADD+, NORMAL and ADD-) which differ in the density of first rank elements in the informational structure. ADD+, NORMAL and ADD- sites also differ in their ability towards adaptive conformational reorganization which contributes to the formation of protein-protein interaction interfaces in enzyme-inhibitor complexes. The study of hydrolytic enzymes in complex with their protein inhibitors shows that at least one of the interaction interface sites is of ADD- type. ADD- sites possess an increased ability towards adaptive conformational changes thus enabling effective protein interaction.
J Biomol Struct Dyn. 2010 Aug;28(1):85-96.