Although the coarse-grained method has been successful for many systems, some important systems, especially biological macromolecular systems, have yet to be developed. Biological macromolecular systems, such as DNA, protein, etc., have the characteristics of spatial inhomogeneity, long time characteristic scale, and multiple scales. It makes it difficult for basic all-atom molecular dynamics to simulate these systems. These characteristics are also the difficulty in establishing an accurate and effective coarse-grained biomolecular system model. In addition, how to effectively deal with the specific structures and specific interactions of these biological macromolecules is also a challenge for coarse-grained methods. It is still a difficult long-term task to establish an effective coarse-grained force field for biological macromolecular systems like the current all-atom molecular force field, so as to achieve semi-quantitative prediction.
In some cases, it is necessary to consider the multi-scale simulation problem of a mixture of the all-atom model and the 5-coarse-grained model. Similar to the QM/MM method, under the overall framework of coarse-grained simulation, high-level simulation that can reflect more molecular details can be realized in some areas. At present, there have been beneficial attempts in related aspects, but they have not yet been able to establish a universal theoretical framework and general algorithms. In addition, it may be possible to use the advantage of the coarse-grained model to efficiently sample the free energy surface. After the coarse-grained simulation is completed, the all-atom model can be reconstructed according to the coarse-grained configuration to accelerate the all-atom simulation. How to effectively transform the coordinates of all-atoms and coarse-grained particles at the junction and how to effectively connect the full-atom and coarse-grained force fields is the difficulty and key to the realization of multi-scale simulation of all-atoms and coarse-grained particles.
The current research on coarse-grained algorithms is mostly focused on reconstructing the structure of the system as accurately as possible. How to accurately describe the kinetic properties at the same time at the level of coarse-grained and how to coarse-grain the kinetic properties has not yet been fully studied. For this problem, our understanding is that it is necessary to combine the advantages of coarse-grained methods for sufficient potential energy surface sampling and the advantages of all-atom methods to handle rare events to effectively study the dynamics of large time scales that are difficult to directly deal with in all-atom simulations problem.
Considering the continuous and rapid growth of computing power, the goal of the coarse-grained method should not only be the application of computational techniques to accelerate molecular simulations. Establishing the theory of the computational complexity of the coarse-grained method will help identify very important systems and problems from the perspective of the development of the coarse-grained method.
The ideal coarse-grained force field should be like a general all-atom force field, with good portability, that is, a set of coarse-grained force field can be used for different thermodynamics without modification or after very simple modification. The combination of environment (such as various temperatures, pressures, concentrations, etc.) and different molecular structures (such as various amino acid residue sequences). At present, most of the coarse-grained methods are dedicated to accurately reproducing the structural information of a given system at the all-atom simulation level, and there is no guarantee of portability in the theory. How to ensure the portability of the established coarse-grained force field is another important issue that needs to be studied in the coarse-grained methodology.