Hybrid approach to analysis of beta-sheet structures based on signal processing and statistical consideration

Abstract

A number of biotechnology applications are based on protein design. For this design, the relationship between a proteins primary structure and its conformation is of vital importance. A beta-sheet is a common feature of a proteins two-dimensional structure; therefore, elucidating the principles governing beta-sheet structure and its stability is critical for understanding the protein-folding process. In the three-dimensional representation of protein molecules, C(alpha) carbon coordinates (carbon atom immediately adjacent to the carboxylate group) have often been employed instead of the complete set of coordinates for the corresponding residues. Using the C(alpha) carbon coordinates, we showed that particular amino acids are not randomly distributed within a beta-sheet structure. On the basis of a new statistical approach for the analysis of a spatial distribution of amino acids in a protein, presented by their physico-chemical parameters, the electron-ion interaction potential (EIIP) and hydrophobicity, are described here. The relationship between amino acid positions inside the beta-sheet and the EIIP and hydrophobicity parameters was established. The correlation between amino acid propensities related to the beta-sheet was examined using multiple cross-spectra analysis. We also applied the continuous wavelet transform for the analysis of selected beta-sheet structures using the EIIP and hydrophobicity parameters. The findings provide new insight into conformational propensities of amino acids for the adaption of beta-sheet structures.