Study Of The Relationship Between Protein Sequence And Structure

It is the main topic to study the relationship between protein sequence and structure in molecular biology. Although researchers have been made huge success in the study of protein in the half of a century,there still have some basic problems unrevealed, such as whether protein sequences are random chains or not, and how to predict the secondary and tertiary structures of proteins directly from their amino acid sequences, and so on. So, it is a meaningful task to study the relationship between protein sequence and structure. Firstly, we analyze the distribution of the components of amino acids of protein sequence and find that the probabilities of the 20 kinds of amino acids are not equal. We also find there is large difference between the number of known protein sequence and that of random sequences which made up randomly by 20 kinds of amino acids. All these revealed that the protein sequence is not uniform randomness in some extent. In order to demonstrate our conclusion further, we bring forward nonlinear prediction method to study protein sequence. We apply nonlinear prediction method to all alpha proteins, all beta proteins and Tim-barrel proteins and find that there exists nonlinear correlation between protein sequences and their structures. It is showed that most proteins contain certain length correlation segments only if its three-dimension structure is symmetric, and the length of these correlation segments is also corresponding to the length of these symmetrical regions. For finding out the essence of the correlation between protein sequences and the symmetry of their structures, we use the modified nonlinear prediction method and recurrent plot method to the beta-trefoil family proteins whose structures contain perfect three-fold symmetry. The results show that these protein sequences also contain hidden three-fold symmetry and the hidden symmetry is corresponding to their structural three-fold symmetry. So we believe that their structural symmetry is possibly coded by their sequence hidden symmetry. In summary, we studied the relationship of protein sequence and structure. The results not only reveal part of the relations between symmetries of sequences and symmetries of structures of proteins, but also show that the nonlinear correlation method may provide a new approach to detect the essential characters and inherent rules of protein sequences.