Amino acid correlation functions in protein structures

Abstract

Understanding the spatial folding of proteins from their amino acid sequences has an enormous potential in contemporary life sciences. The ability to predict secondary and tertiary structures from the primary through the use of computers will enable a much faster and more efficient discovery of organic substances with therapeutic or otherwise bioactive potential, largely eliminating the need for synthesis and testing of large numbers of organic substances for physiological effects. Our manuscript presents an application of correlation function analysis, usually used to describe properties of liquids, to protein structures in order to elucidate statistically favored distances among amino acids. Pairwise distribution functions were calculated between C-alpha atoms of 20 amino acids in a large ensemble of Protein Data Bank structures. The correlation functions show characteristic distances in amino acid interactions.