Study Of Relationship Between Sequence Co-evolution And Unbound-state And Bound-state Structures Of Proteins

Protein sequence determines its structure,but proteins usually participate in life activities in the form of complex with biological macromolecules such as proteins and RNA.Therefore,each protein molecule has two states: unbound state and bound state.Some proteins have similar structures of unbound state and bound state,while others have different or even very different structures of unbound state and bound state.So,does the protein sequence determine the unbound state or the bound state?Studies have shown that the interactions between amino acid residues in protein molecules can be stored in homologous sequences in the form of co-evolution.The co-evolutionary residue pairs in protein sequences contain important biological functions and structural information.In recent years,many algorithms have been proposed to use co-evolution information in protein sequences to predict their residue contacts.Among them,direct coupling analysis(DCA)is the most widely used algorithm.The successful application of DCA not only significantly improved the accuracy of predicting the structure of protein monomers and protein complexes,but also revealed some protein biological functions and mechanisms,which greatly promoted the development of drug design.So do these co-evolved residues correspond to the residues in the unbound state structure or the contacting residues in the bound state structure? In this article,the DCA under Pseudo likelihood approximation(Plm DCA)method with better accuracy was used to analyze the co-evolutionary residue pairs in protein sequences,and then compared with the contact residue pairs formed in the unbound state and the bound state structures.The contact residue pairs are compared.Through the statistical analysis of the prediction accuracy(accuracy,ACC))of the contact residues in the protein structure in the data set,we found that for proteins with similar structures in the unbound state and the bound state,the ACC values of most proteins in different states Very high,which means that DCA can be used to analyze the relationship between co-evolution of protein sequence and the structure of free and bound states.On this basis,we studied proteins with different structures in the unbound state and the bound state in the protein-protein complex,and found that the co-evolution information in the protein sequence has a strong correlation with the residues in the bound state structure.We have reached the same conclusion for proteins with different structures in the unbound state and the bound state in the protein-RNA complex.However,due to the small number of proteins in the data set,the conclusions drawn here need to be further verified.