Enhanced Sampling Methods Integrate Small-angle X-ray Scattering Data To Simulate The Dynamic Structures Of Highly Flexible Proteins

Biological macromolecules(proteins,nucleic acids,etc.)usually participate in many life activities,and the three-dimensional(3D)structure and function of these macromolecules are closely related.Structural biologists continue to develop new technical methods to analyze the 3D structures of biological macromolecules,trying to better explain the mechanisms through the analysis of their 3D structures.These methods for analyzing macromolecular structures mainly include two categories:Firstly,methods to obtain high-resolution structural information including X-ray crystallography and cryo-electron microscopy(cryo-EM).Secondly,small-angle X-ray scattering(SAXS),chemical cross-linking coupled mass spectrometry(CX-MS),fluorescence resonance energy transfer(FRET)and other methods that only obtain low-resolution structural information are included.With the development of computer technology,structural biologists put forward the concept of "integrative structural modeling"(ISM).Its main purpose is to analyze the precise macromolecular structures by combining computer technology with one or more kinds of different resolution experimental structure information.ISM mainly includes four steps:building a model,determining a scoring function,conformational space sampling,and verifying the model.Among them,the process of conformational space sampling is the "speed-limiting step" of ISM,and the enhanced sampling methods are used to seek short-term and efficient sampling.How to improve the sampling efficiency of conformational space is still a hot and difficult point today.In the second chapter of this article,we first integrated an enhanced sampling method and small-angle X-ray scattering to study the unique long region 11(UL11)protein,which is an intrinsically disordered proteins(IDPs).We constructed an all-atom model of the system,using the combination ofA99SB force field and OPC water model,and using accelerated molecular dynamics(aMD)simulation.Then,integrative modeling is conducted based on iterative multiple independent simulations.SAXS are used to design a scoring function for screening conformations and thus guide the simulations toward an ensemble that fits the experimental data well.It was found that the resuts fit well with the SAXS data,and can obtain various states of UL11.Our simulations were carried out on the basis of the all-atom models,which are great significance for understanding the structural dynamic changes.Although the object of our current research is the monomer UL11,we only need to modify the algorithm to simulate the phase separation process of multiple proteins.Subsequently,we extended an enhanced conformational sampling of macromolecules by cluster-guided iterative multiple independent molecular dynamics(clust-MIMD)simulations in chapter 3.We have determined three clustering strategies:intermediate value(center,CEN),Close to target(C2T)and farthest to initinal(F2I),and selected the representative conformation of each cluster to start the next cycle.The clust-MIMD simulations have more sufficient sampling than conventional molecular dynamics(cMD)simulations in the AdK protein system,especially in F2I strategy.You can also see obvious intermediate states and transition pathways in the map of conformational distribution.At the same time,we also applied this sampling methods to BS69 protein,which is more flexible system.The results show that the data of our enhanced sampling method fits better with SAXS data than cMD simulations.