Study On The Relationship Between Rna Structure And Function By Gaussian Network Model

Ribonucleic acid(RNA)is an important biological macromolecules in living organisms.A large number of studies have shown that only less than 2% of RNA molecules encode proteins,the rest RNA molecules acting as non-coding RNA play a variety of biological functions in many biological processes,such as catalytic and regulation functions.The most important application fields of the non-coding RNA researches are the prevention and treatment of the related human diseases,which include understanding the occurrence and progress for the disease,as well as looking for the new biomarkers and drug targets for the new diagnostic and treatment methods of the disease.With the increase of the number of available RNA tertiary structures in the database,how to predict its biological function from the three-dimensional structures of RNA is becoming more and more important.In order to understand the mechanism of the biological functions of RNA,it is necessary to obtain the intrinsic dynamical properties of the RNA,besides its static tertiary structures.Therefore,it is urgent to develop effective theoretical method and computation technologies to investigate the dynamic properties encoded in the RNA structure.In this thesis,we learn the successful experience for the application of the Gaussian network model(GNM)in proteins,and then develop the effective method based on the GNM to investigate the structure-function relationship for RNA.Firstly,the different types,structures and functions of non-coding RNAs are introduced.Based on the structural properties of RNA,the study subject of this thesis is focused on the rRNA,tRNA and riboswitch RNA with single chain.Then,in our studies,the native structure of RNA is simplified as an elastic network model based on the GNM,in which the P atom in each nucleotide acid is represented by a node.If the distance between two nodes is less than the cutoff value,they are connected by a spring.In the GNM,the spring constant is a variable parameter.In this thesis,the interactions between different nodes are classified into two types: the covalent and non-covalent interactions,where the spring constants for the covalent and non-covalent interactions are c? and ?,respectively.In this study,the value of c? is changed,and then the correlation coefficient between the theoretical and experimental temperature factors is calculated.The optimized value of c? is obtained through maximizing the correlation coefficient of the temperature factors.After that,the distribution of the values of c? is analyzed,and the distribution characteristics for rRNA,tRNA and single-chain riboswitch RNA are obtained.Finally,based on the different distribution characteristics of spring constants for these different structure type RNAs,the GNMs are constructed for these different RNA structures.In this thesis,the GNMs for the rRNA and riboswitch RNA structures are developed,which is helpful for the further studies of the intrinsic dynamics encoded in RNA structures and the exploration of the biological functions of these RNAs.