In recent years, Microsoft Research Asia has conducted a series of studies on the new crown virus.
In 2021, Microsoft Research Asia and Tsinghua University jointly published a cover article in Advanced Theory and Simulations, "Exploring the Regulatory Function of the N-terminal Domain of SARS-CoV-2 Spike." Protein Through Molecular Dynamics Simulation）[1]。
This is a computational biology study jointly carried out by the team of Professor Gong Haipeng of Microsoft Research Asia asia and Tsinghua University School of Life Sciences, which mainly studies the role of N-terminal domain NTD (N-Terminal Domain) in the human mechanism of new coronavirus infection through molecular dynamics simulation.
NTDs that regulate infection
It is understood that there is a layer of S protein on the surface of the new crown virus, and the virus can only enter the human body after it is recognized by the body's receptor protein ACE 2. The stronger the combination of the two, the easier it is to cause infection.
The form of the S protein is somewhat like the letter "Y", and the downward vertical area, the S2 region, is embedded in the surface of the virus and plays a fixed role. The two "branches" that protrude are called NTD and RBD (Receptor Binding Domain).
Figure | S protein structure of the new coronavirus (left is the S protein trimer structure, composed of three chains; the right picture is the S protein monomer structure, composed of NTD, RBD and S2) (Source: Microsoft Research Asia)
Prior to this study, it had been known experimentally that RBD could directly play a role in recognizing human receptor proteins, and the specific role of NTD was not very clear.
Microsoft Research Asia and Professor Gong Haipeng's laboratory collaborated to build a stable simulation system with millions of atoms, and after molecular dynamics simulations of billions of steps (1 step is 1 femtosecond, that is, one hundred trillionths of a second), a "wedge" model hypothesis was proposed, which can be simply regarded as a "wedge" inserted at the bottom of RBD and "Y".
Studies have found that RBD has two states, down and up, and only the up state is likely to be recognized with human receptor proteins. The study found that RBD tends to move from an up state to a down state, and when NTD is crammed into the lower end of RBD like a wedge, it blocks this tendency of RBD, that is, NTD keeps RBD up, further giving it the ability to infect.
It can be said that NTD plays a regulatory role in the process of virus infection, regulating the morphology of RBD, thereby changing the tendency of the virus to infect the human body.
Figure | Schematic of NTD's regulatory role in changes in S protein conformational changes in SARS-CoV-2 (Source: Microsoft Research Asia)
It is foreseeable that the interface of NTD regulation of RBD may become a potential drug target. If a drug molecule can be designed to pull the NTD outward, that is, to pull out the "wedge", then the RBD will tend to be down and become a conformation without the ability to infect, and the ability of the virus to infect people will be much reduced.
"Our work is the first in the world to propose a regulatory model of NTD in the process of viral infestation. In addition, according to this potential drug target, we designed a virtual screening algorithm for drugs, and screened some potential drugs of the new crown virus, which provided a certain reference value for the development of new coronavirus drugs. Wang Tong, a research fellow at Microsoft Research Asia, said.
The new model reveals the reason for the strong infectivity of Ami kerong
Recently, Microsoft Research Asia and Professor Wang Xinquan of tsinghua University School of Life Sciences and Professor Zhang Linqi of the School of Medicine of the university have collaborated to make new breakthroughs in the interpretation of the mechanism of strong contagion of the Aomi Kerong variant, and their papers have been accepted by Cell Research, a top journal in the field of biology.
It is understood that in this cooperation, Professor Wang Xinquan's research group took the lead in analyzing the crystal structure of Themi Kerong high resolution, which laid the foundation for the study of the infection mechanism of Theomi Kerong.
Compared with the static crystal structure, Wang Tong's team used molecular dynamics simulation to simulate and analyze the structural variation of Omikeron and its infection mechanism from a "dynamic perspective".
In the study, researchers at Microsoft Research Asia first constructed two simulation systems, using the original structure of the new coronavirus and the structure of Omikeron as the starting structure, and conducted long-term, hundreds of millions of steps of molecular dynamics simulations in parallel with these two structures, simulating the movement process at the atomic level, so as to observe the real changes of the virus in the human body.
(Source: Pixabay)
"The method of simulating molecular motion can be divided into two types, classical molecular dynamics simulation and quantum simulation based on primordiality," Wang Tong said, "for the study of protein macromolecules in the new crown virus, large-scale conformational change movement, the classical dynamic simulation method is a more suitable means." ”
According to Wang Tong, classical simulations are suitable for large systems, like thousands of atoms in proteins, such as this simulation system contains millions of atoms. In terms of time, classical simulations do simulations with hundreds of millions to billions of steps. In contrast, quantum simulations are relatively accurate, but the time-consuming process of computation makes it applicable only to very small systems, such as calculating properties of systems with only more than ten atoms or performing short simulations.
After building a simulation system, Microsoft Research Asia used self-developed algorithms to analyze the ability of the original coronavirus and Omilon to infect humans, as well as their very slight differences in structure.
It is worth noting that the amount of data generated during the research process is extremely large, and the entire simulation has at least hundreds of millions of structures, resulting in hundreds of millions to billions of frames of structural changes.
To this end, the researchers creatively proposed a new Markov model algorithm to analyze the data. It has a stronger characterization ability for the data analysis of molecular dynamics simulations, and can more realistically simulate the dynamic change process of proteins.
After screening and analyzing the massive amount of data, the researchers aggregated billions of structures into several representative types of structures, and then analyzed the differences between them and how they were converted to each other, as well as the ability to bind to human receptor proteins, so as to better understand the molecular mechanism of Omiljung infection.
These aggregated representative conformations are called "sub-states", and the researchers have made qualitative binding free energy calculations to reflect the infective capacity of the virus to some extent.
Specifically, in the study, both the Olmikeron and the original virus after clustering had three substates. Since Omilon itself is only a variant of the original virus, they each have the highest proportion of a substate, both in terms of structure and the ability to bind to human receptor proteins.
Thus, the key differences in the Structure of Theomexjong are in two other sub-states. Compared with the two substates of the original virus, the two substates of Omikeron interacted more with the human receptor protein ACE2 and were more able to bind (less free binding energy), thus leading to a stronger ability to infect.
At the same time, the study also found that the transition between the three sub-states of Omikeron is very fast, and it is easy to switch from the main sub-state similar to the original virus to the other two sub-states with stronger binding ability. This also explains from a dynamic perspective why Aumechon is so infectious.
Figure | Molecular dynamics simulation reveals the state changes and infection mechanism of Omikejong (the left picture is the new coronavirus S protein, the right picture is the S protein of the Omilon variant, the State3 of the two systems is basically the same and the main conformation, and the other two substates of the Omilon have significantly higher binding free energy than the two substates of the new crown virus) (Source: Microsoft Research Asia)
In the latest research work on Omikejong, Wang Tong said, "This study is the first time in the world that the method of 'combination of dry and wet' and dynamic perspectives is used, starting from the perspectives of structural biology and computational biology at the same time, proposing the molecular mechanism of Omilon with strong infectivity." ”
"Wet and dry combination" promotes scientific research
The "wet and dry combination" mentioned above is also a combination of "dry experiments" (computational experiments) and "wet experiments" (traditional scientific experiments).
It is understood that the "dry experiment" can verify and supplement some of the existing conclusions of the "wet experiment", and can also do some research that the "wet experiment" cannot do or needs to spend a lot of time and manpower to do.
For example, in the study of Omi kerong, if you analyze the changes in the structure of mutation sites one by one in the "wet experiment" laboratory, it may take several months, and in the case of such a severe epidemic situation, it is obviously not conducive to the prevention and control of the epidemic. Using computational means to simulate the effects of this mutation may only take a week or two.
In addition, "wet experiments" are often exploratory in the early stage, and even in the direction of research, they are very confused. At this time, the "dry experiment" can guide the experimental design of the "wet experiment" to a certain extent, helping it to find a more successful method.
It is worth noting that the two do not replace each other, but are equally important, guiding and complementary. "We feel this way in the process of working with research partners, and we think that the 'combination of wet and dry' will play a spiraling, wave-like impetus for the whole scientific research." Wang Tong said.
Figure | Wang Tong, Head Of Microsoft Research Asia (Source: Wang Tong)
He also added that in the research of the new crown virus, the cooperation with many teachers at Tsinghua University is a very good starting point. Microsoft Research Asia is very much hoping to have more in-depth cooperation and cross-field exchanges with experts and scholars from universities and research institutes at home and abroad.
It is understood that Microsoft Research Asia will also carry out relevant research on why the new coronavirus is so infectious throughout the evolution of the coronavirus.
Interdisciplinary development places higher demands on talents
Finally, wang tong also expressed his views on the requirements and development trends of talents in interdisciplinary fields such as computational biology.
He believes that interdisciplinary development is mainly around a key word, that is, "intersection". Crossover is not simply "addition", but "multiplication". There must be a certain chemical reaction and deep integration between the two disciplines.
In the case of computational biology, it is essentially the intersection of computer science and biology, and it is necessary to combine the knowledge of the two fields and truly digest and understand in order to have a real academic impact in the field of computational biology and solve real scientific problems.
This also puts forward a higher requirement for the cultivation of talents or the demand for talents. If researchers only grasp one aspect of knowledge, the understanding of another discipline is not deep enough, and only stay at the level of simple tool use, then the research work can only float on the surface, and it is easy to cause problems such as incorrect setting of calculation parameters, resulting in the final calculation results and experimental results are completely different.
In the future, the cross-integration between various disciplines will be more and more, in the specific scientific research work, talents from different backgrounds need to listen more to each other's knowledge and demands in the process of cooperation and exchange, only based on mutual trust and open mind, can we promote the sustainable development of interdisciplinary research.
-End-
reference:
1、Y. Li，T. Wang，J. Zhang，B. Shao，H. Gong，Y. Wang，X. He，S. Liu，T. Liu.Exploring the Regulatory Function of the N-terminal Domain of SARS-CoV-2 Spike Protein through Molecular Dynamics Simulation.Advanced Theory and Simulations4（2021）.
https://doi.org/10.1002/adts.202100152