Molecular anti-aggregation simulation method, three-dimensional grid structure hardware and medium

Abstract

The invention discloses a molecule anti-aggregation simulation method, three-dimensional grid structure hardware and a medium, and the method comprises the steps: recognizing a plurality of limited atoms, and determining a corresponding target calculation unit according to the position description information of each limited atom; adding additional information into the data parameter information, and storing the data parameter information into an associated storage unit; calculating Van der Waals force and electrostatic force in parallel according to the data parameter information in the associated storage unit through each target calculation unit, and determining the stress among the limited atoms according to the result; according to the stress and the data parameter information, parallel computing new position information of each limited atom to serve as an anti-aggregation simulation result under the current time step, and additionally storing the anti-aggregation simulation result into the data parameter information of the associated storage unit; and after the new target calculation unit and the association storage unit are updated according to the new position information, the operation of parallel calculation of the Van der Waals force and the electrostatic force is returned until simulation of the preset time step number is completed, and the simulation efficiency, accuracy and stability are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing and molecular dynamics simulation optimization, and particularly relates to a molecular anti-aggregation simulation method, a three-dimensional grid structure hardware and a medium. BACKGROUND

[0002] Swimming system usually includes a protein and a plurality of small molecule ligands. Molecular dynamics simulation based on the swimming system reveals the binding mechanism and dynamic behavior of small molecule ligands and proteins by simulation, and plays an important role in drug design and protein ligand interaction research.

[0003] However, the system often faces the problem of aggregation of small molecule ligands in long-time simulation, especially in high-concentration solution or hundreds of nanosecond simulation. For example, four imatinib molecules aggregate near the protein, which interferes with the correct sampling of the drug binding process and affects the accuracy of the simulation.

[0004] The related art mainly processes the swimming system through the traditional molecular dynamics simulation method. Due to the dependence on general computing architecture and conventional parallel algorithm, it is difficult to efficiently carry the surge of computing load when facing large-scale swimming system such as high-concentration small molecule solution, resulting in low simulation efficiency. The conventional parallel algorithm cannot fully adapt to the characteristics of modern hardware and cannot maximize the parallel computing power, further limiting the improvement of computing efficiency. When dealing with the aggregation problem, global intervention or empirical parameter adjustment is often used, which is difficult to adapt to the dynamic behavior of small molecule ligands, resulting in insufficient simulation accuracy and difficulty in accurately reproducing the dynamic behavior of the actual system, causing non-physical aggregation phenomenon to exist or excessive inhibition of normal interaction, reducing the accuracy of long-time simulation results. SUMMARY

[0005] The present application provides a molecular anti-aggregation simulation method, a three-dimensional grid structure hardware and a medium to solve the problems of low simulation accuracy, low efficiency and low accuracy of molecular anti-aggregation simulation.

[0006] According to an aspect of an embodiment of the present application, a molecular anti-aggregation simulation method is provided, comprising:

[0007] Analyzing the simulation results of the protein system molecular dynamics, identifying a plurality of restricted atoms that need to be treated by anti-aggregation in the small molecule ligands that meet the aggregation conditions;

[0008] According to the position description information in the data parameter information of each restricted atom extracted in the simulation results, determining the target computing unit corresponding to each restricted atom in the three-dimensional grid structure hardware;

[0009] The data parameter information of each restricted atom is stored in the associated storage unit of the target calculation unit after additional information for calculating the additional repulsive force between the restricted atoms is added in the data parameter information of each restricted atom;

[0010] The van der Waals force and the electrostatic force between each restricted atom are calculated in parallel by each target calculation unit according to the data parameter information in the associated storage unit, and the force between each restricted atom is determined according to the calculation result;

[0011] The new position description information of each restricted atom is calculated in parallel by each target calculation unit as the anti-aggregation simulation result at the current time step according to the force between each restricted atom and the data parameter information in the associated storage unit, and is additionally stored in the data parameter information of the associated storage unit;

[0012] After the new position description information is updated, the new target calculation unit and the associated storage unit corresponding to each restricted atom are updated, and the operation of calculating the van der Waals force and the electrostatic force between each restricted atom in parallel according to the data parameter information in the associated storage unit is performed until the simulation of the preset number of time steps is completed.

[0013] According to another aspect of the embodiment of the present application, a molecular anti-aggregation simulation device is provided, comprising:

[0014] The restricted identification module is configured to analyze the simulation result of the protein system molecular dynamics, and identify a plurality of restricted atoms that need to be subjected to anti-aggregation treatment in the small molecule ligand satisfying the aggregation condition;

[0015] The unit determination module is configured to determine the target calculation unit corresponding to each restricted atom in the three-dimensional grid structure hardware according to the position description information in the data parameter information of each restricted atom extracted in the simulation result;

[0016] The additional information module is configured to store the data parameter information of each restricted atom in the associated storage unit of the target calculation unit after additional information for calculating the additional repulsive force between the restricted atoms is added in the data parameter information of each restricted atom;

[0017] The force calculation module is configured to calculate the van der Waals force and the electrostatic force between each restricted atom in parallel by each target calculation unit according to the data parameter information in the associated storage unit, and determine the force between each restricted atom according to the calculation result;

[0018] The additional storage module is configured to calculate the new position description information of each restricted atom as the anti-aggregation simulation result at the current time step in parallel by each target calculation unit according to the force between each restricted atom and the data parameter information in the associated storage unit, and additionally store the new position description information in the data parameter information of the associated storage unit;

[0019] The execution module is returned to update the new target calculation unit and the associated storage unit corresponding to each limited atom according to the new position description information, and then return to perform the operation of calculating the van der Waals force and the electrostatic force between each limited atom according to the data parameter information in the associated storage unit in parallel until the simulation of the preset number of time steps is completed.

[0020] According to another aspect of the embodiments of the present application, a three-dimensional grid structure hardware is provided, which comprises:

[0021] The first dimension x the second dimension x the third dimension grid points; each grid point comprises a calculation unit and an associated storage unit; each calculation unit or associated storage unit is communicatively connected with an adjacent calculation unit or associated storage unit; the calculation unit or associated storage unit located at the edge of the three-dimensional grid structure hardware is communicatively connected with the calculation unit or associated storage unit located at the opposite edge of the three-dimensional grid structure hardware; and a global control unit communicatively connected with each calculation unit; the global control unit is used to issue parameters to the calculation unit and coordinate the calculation task;

[0022] Each associated storage unit is communicatively connected with the corresponding calculation unit; wherein the associated storage unit stores the data parameter information according to any embodiment of the present application, and a computer program executable by the at least one calculation unit, the computer program is executed by the at least one calculation unit to make the at least one calculation unit execute the molecular anti-aggregation simulation method according to any embodiment of the present application.

[0023] According to another aspect of the embodiments of the present application, a computer readable storage medium is provided, which stores computer instructions for making a processor execute the molecular anti-aggregation simulation method according to any embodiment of the present application.

[0024] According to another aspect of the embodiments of the present application, a computer program product is also provided, which comprises a computer program, the computer program is executed by a processor to implement the steps of the method according to any embodiment of the present application.

[0025] The technical scheme of the embodiment of the present application is characterized in that, by analyzing the simulation result of the protein system molecular dynamics, a plurality of restricted atoms needing to be prevented from aggregation are identified in the small molecule ligand meeting the aggregation condition, the corresponding target computing unit is allocated in the three-dimensional grid structure hardware based on the position information, and the additional information for calculating the additional repulsive force is added to the data parameters and stored in the associated storage unit; each target computing unit calculates the van der Waals force, electrostatic force and stress between the restricted atoms in parallel, and then obtains new position information as the current time step anti-aggregation simulation result and stores it additionally, while the target computing unit and the associated storage unit are dynamically updated according to the new position, and the calculation is repeated until the preset time step is completed. By positioning the restricted atoms needing to be prevented from aggregation and allocating the target computing unit for them, the hardware parallel computing capability can be fully utilized, the calculation and storage are bound nearby to reduce the delay, the dynamic adaptation to the change of atomic position ensures the continuity of the simulation, and the efficiency and accuracy of the anti-aggregation simulation are improved, so that the stable simulation of long time scale can be ensured, the non-physical aggregation of the small molecule ligand in the simulation process can be effectively prevented, and the accuracy of the small molecule ligand aggregation behavior simulation is improved.

[0026] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present application, nor is it used to limit the scope of the present application. Other features of the present application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0027] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creating laborious work.

[0028] Figure 1 is a flow chart of a molecular anti-aggregation simulation method provided by the first embodiment of the present application;

[0029] Figure 2 is a flow chart of another molecular anti-aggregation simulation method provided by the second embodiment of the present application;

[0030] Figure 3 is a flow chart of still another molecular anti-aggregation simulation method provided by the third embodiment of the present application;

[0031] Figure 4 is a structural schematic diagram of a molecular anti-aggregation simulation device provided by the fourth embodiment of the present application;

[0032] Figure 5 is a two-dimensional structural schematic diagram of a three-dimensional grid structure hardware for implementing the molecular anti-aggregation simulation method of the present application. Detailed Implementation

[0033] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0034] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0035] Example 1

[0036] Figure 1 This is a flowchart of a molecular anti-aggregation simulation method provided in Embodiment 1 of the present invention. This embodiment is applicable to simulating the dynamic behavior of molecules and the aggregation behavior of small molecule ligands. The method can be executed by a molecular anti-aggregation simulation device, which can be implemented in hardware and / or software, and is generally configured in a three-dimensional mesh structure hardware. Specifically, the three-dimensional mesh structure hardware can be understood as: an architecture that organizes hardware computing resources into a three-dimensional mesh form to improve the efficiency and performance of simulation calculations.

[0037] Correspondingly, such as Figure 1 As shown, the method includes:

[0038] S110. Analyze the simulation results of molecular dynamics of the protein system and identify multiple restricted atoms that need to be treated to prevent aggregation in small molecule ligands that meet the aggregation conditions.

[0039] In the embodiments of the present application, the simulation result of the protein system molecular dynamics can be understood as follows: basic data about the protein and the small molecule ligand are obtained through experimental or theoretical modeling molecular dynamics simulation, or directly from a file storing molecular structure information, and can specifically include: atom type (such as carbon, oxygen, nitrogen, and other atom types), atom quantity, charge amount carried by the atom, and the coordinates of the atom in space, and the like.

[0040] The small molecule ligand satisfying the aggregation condition can be understood as follows: in the process of molecular dynamics simulation, or from the initial structure, the small molecule ligand that has a mutual approach and aggregation trend. For example, the atomic distance of multiple small molecule ligands is continuously reduced during simulation, and there is a condition of being less than a preset aggregation condition threshold, which has the possibility of forming non-physiological aggregation, and is considered to satisfy the aggregation condition. Or, according to the structure of the small molecule ligand itself (for example, containing many hydrophobic groups), it is inferred that the small molecule ligand is easy to aggregate, and is screened out as a small molecule ligand satisfying the aggregation condition. Or, according to the actual experimental requirements (for example, the physiological environment requires the ligand to be dispersed), the small molecule ligand that needs to be prevented from aggregation is reversely screened out.

[0041] The restricted atom can be understood as follows: from the small molecule ligand satisfying the aggregation condition, the atom that plays an important role in judging and processing aggregation is selected. The screening method of the restricted atom can be as follows: selecting a functional group center atom, such as the oxygen atom in the hydroxyl group and the nitrogen atom in the amino group; selecting an atom near the center of gravity, that is, an atom that can represent the overall position of the small molecule ligand, by monitoring these atoms, whether the ligand is aggregated can be judged; selecting an atom with the highest contact probability with other ligands, for example, due to the hydrophobic effect of the hydrophobic end, the ligands are easy to approach and aggregate with each other, so the carbon atom of the hydrophobic end can be selected.

[0042] Specifically, the simulation result of the protein system molecular dynamics is analyzed to obtain the basic structure data of the protein and the small molecule ligand. Based on these data, small molecule ligands with aggregation tendency (satisfying the aggregation condition) are screened out, and multiple restricted atoms that need to be prevented from aggregation are identified from these small molecule ligands.

[0043] In S120, the position description information in the data parameter information of each restricted atom extracted in the simulation result is used to determine a target computing unit corresponding to each restricted atom in the three-dimensional grid structure hardware.

[0044] In the embodiment of the present application, the data parameter information can be specifically understood as a set of various data and parameters related to the restricted atoms for calculation and analysis in the process of molecular dynamics simulation, and can specifically include: basic properties of atoms, such as the type, charge (static parameter for calculating electrostatic force), and mass of the atom; position description information, such as the coordinates and velocity of the atom in three-dimensional space; anti-aggregation related parameters, such as the pre-set cutoff radius and information for calculating repulsive force; and attribution and association information, such as the molecule to which the atom belongs, and the interaction relationship with other atoms (such as whether there is a near-neighbor interaction relationship between the restricted atoms). The computational unit (CU) can be specifically understood as a unit in a three-dimensional grid structure hardware for processing the calculation of atoms in a specific spatial region.

[0045] Specifically, based on CU(x,y,z)=floor(position / box_size×[X,Y,Z]), the grid coordinates CU(x,y,z) of the restricted atom are determined according to the position description information in the data parameter information of each restricted atom extracted in the simulation result, that is, the restricted atom is assigned to the target CU corresponding to the grid coordinates.

[0046] Wherein, position represents the coordinates of the atom in the simulation box, box_size represents the side length of the simulation box, [X,Y,Z] represents the node number of the three-dimensional grid structure hardware in three dimensions, and floor(·) represents the floor function, that is, the atomic coordinates are converted into integer grid coordinates (for example, (1,2,1) is obtained after calculation, which corresponds to the calculation unit of the node CU(1,2,1) in the three-dimensional grid structure hardware).

[0047] Wherein, the simulation box can be specifically understood as a virtual three-dimensional space region for defining the substances such as proteins, small molecule ligands, and solvent molecules in the three-dimensional grid structure hardware.

[0048] In the data parameter information of each restricted atom, after adding the additional information for calculating the additional repulsive force between the restricted atoms, the data parameter information is stored in the associated storage unit of the target calculation unit.

[0049] In the embodiment of the present application, the additional information can be specifically understood as new parameter information added to prevent the restricted atoms (key atoms of small molecule ligands prone to aggregation) from gathering together, and the core of the additional information is the adjustment of the van der Waals force parameter. Specifically, it can include the newly added van der Waals parameters σ' (that is, the equilibrium distance parameter for adjusting the range of repulsive force) and ε' (that is, the energy parameter for adjusting the strength of the repulsive force). The van der Waals force calculated by the two parameters will be stronger than the original parameters, which can play a role in applying stronger force between the aggregated atoms to disperse the aggregated atoms.

[0050] The associated storage unit can be understood as a storage module bound to the target computing unit in the three-dimensional grid structure hardware, used to store atomic data processed by the CU nearby for quick calling by the CU.

[0051] Specifically, after adding the new parameter for calculating stronger repulsive force as additional information in the data parameter information of each restricted atom, the complete data is stored in the associated storage unit of the corresponding grid point. Optionally, the associated storage unit can maintain a parameter table and an atomic list respectively to record the corresponding data parameter information.

[0052] S140, by each target computing unit, the van der Waals force and the electrostatic force between each restricted atom are calculated in parallel according to the data parameter information in the associated storage unit, and the force between each restricted atom is determined according to the calculation result.

[0053] Specifically, since different restricted atoms are distributed in different regions of the three-dimensional grid, each target computing unit can independently calculate the interaction between the restricted atoms in the region.

[0054] Each target computing unit directly reads data parameter information such as coordinates, charges and van der Waals parameters of restricted atoms from the bound associated storage unit. For each pair of restricted atoms, the target computing unit calculates the corresponding van der Waals force and electrostatic force with additional repulsive force to prevent aggregation, superimposes the two vectors, and obtains the total force of each atom.

[0055] Since the new van der Waals parameter is used for the restricted atom, the calculated van der Waals force will be stronger than that between ordinary restricted atoms, especially when the atomic spacing is small (i.e. close to aggregation), the repulsive force will increase significantly, thereby pushing away the atoms that may aggregate, achieving the core goal of preventing aggregation, while the electrostatic force calculation is still based on the original charge parameter, ensuring that the normal electrostatic interaction is not disturbed, and the physical rationality of the simulation is considered.

[0056] S150, by each target computing unit, the new position description information of each restricted atom is calculated in parallel as the anti-aggregation simulation result at the current time step according to the force between each restricted atom and the data parameter information in the associated storage unit, and is appended to the data parameter information in the associated storage unit.

[0057] Specifically, each target computing unit reads the parameters such as atomic position position(t), velocity velocity(t) and atomic mass mass of the current time step from the associated storage unit, combines the calculated force force(t), and based on the Velocity Verlet algorithm, parallelly calculates the new position and velocity of each confined atom in the region. That is, based on position(t+Δt) = position(t) + velocity(t) x Δt + 0.5 x force(t) / mass x Δt^2 to calculate the new time step position, based on velocity(t+Δt / 2) = velocity(t) + 0.5 x force(t) / mass x Δt to calculate the new half-step time step velocity, based on the new time step position position(t+Δt) to calculate the new time step force force(t+Δt), and based on velocity(t+Δt) = velocity(t+Δt / 2) + 0.5 x force(t+Δt) / mass x Δt to calculate the new time step velocity. Wherein, Δt represents the time step length, that is, the time interval between two adjacent calculation times.

[0058] It can be understood that, due to the addition of an extra repulsive force in the anti-aggregation strategy, the atomic motion may be intensified, thereby causing the temperature to rise. Based on the above calculation, each target computing unit can adjust the updated velocity velocity(t+Δt) through a temperature coupling algorithm (such as the Berendsen temperature coupling algorithm), that is, if the three-dimensional grid structure hardware temperature is higher than the target value, the velocity is reduced; if it is lower than the target value, the velocity is increased, so as to ensure that the simulation temperature is stable at the preset value (such as the physiological environment temperature), and to avoid the influence of non-physical temperature changes on molecular behavior, such as excessive dispersion of ligands due to too high temperature, deviating from the real state.

[0059] After completing the position and velocity update, each target computing unit appends the new position and velocity as new position description information to the data parameter information in the associated storage unit, so as to retain the complete atomic motion trajectory, facilitate subsequent analysis of whether the ligand still has an aggregation trend, and provide the latest position and velocity data for the calculation of the next time step.

[0060] S160, after updating the new target computing unit and the associated storage unit corresponding to each confined atom according to the new position description information, returning to perform the operation of parallelly calculating the van der Waals force and the electrostatic force between each confined atom according to the data parameter information in the associated storage unit, until the simulation of the preset number of time steps is completed.

[0061] Specifically, after the position of the restricted atom changes, the restricted atom may exceed the jurisdiction of the original target computing unit and needs to be re-assigned to a new target computing unit and a corresponding associated storage unit, so as to ensure that the restricted atom is always processed by the target computing unit to which the current position of the restricted atom belongs, and to ensure the spatial correspondence of the calculation.

[0062] After updating the new target computing unit and the associated storage unit corresponding to each restricted atom, through global synchronization, all target computing units are waited to complete the operation of the current time step, and then through consistency checking, it is verified that the total number of restricted atoms in the three-dimensional grid structure hardware is unchanged and there is no repetition and loss, and the molecular structure is complete (such as the atom belonging of each small molecule is not disordered), so as to ensure the reliability of the data after the current time step processing.

[0063] Optionally, on the basis of each of the above embodiments, the associated storage unit can maintain a near-neighbor atom index for recording other restricted atoms within a certain range from the restricted atom, that is, the restricted atoms having interaction relationship.

[0064] Correspondingly, on the basis of each of the above embodiments, after global synchronization and consistency checking, the near-neighbor atom index in each associated storage unit is updated based on the new position description information, and the related counter and flag are reset, and the operation of calculating the van der Waals force and the electrostatic force between each restricted atom in parallel according to the data parameter information in the associated storage unit is returned to be executed until the simulation of the preset number of time steps is completed.

[0065] It can be understood that the total simulation time = preset time step number x Δt, and by adjusting the time step, the simulation accuracy can reach the microsecond or millisecond level, which is sufficient to capture slow processes such as small molecule ligand and protein binding.

[0066] Optionally, on the basis of each of the above embodiments, the data parameter information can include: atom information, atom group information, near-neighbor atom index, van der Waals force parameter, to-be-sent information of target atoms in each storage unit, and electrostatic parameter.

[0067] In the embodiment of the application, the atom information can be specifically understood as a list recording the basic attributes of a single atom, such as atom_list, which can specifically include parameters such as atom serial number id, atom type type, atom coordinates position[3], atom velocity velocity[3], atom force force[3], atom mass mass, and atom charge charge. Wherein, [3] represents three dimensions.

[0068] The atom group information can be specifically understood as a list recording the atom belonging, such as molecule_list, which can specifically include parameters such as atom group serial number id, atom group order type, and atom serial number atom_ids (that is, the atom serial number id in atom_list).

[0069] The near-neighbor atom index can be specifically understood as recording a list of other atoms around each atom that need to be calculated for interaction, such as neighbor_list, which can include parameters such as limited atom serial number id and corresponding near-neighbor atom serial number id.

[0070] The van der Waals force parameter can be specifically understood as recording a parameter table for calculating van der Waals force, such as vdw_param_table, which can specifically include original parameters σ (interatomic equilibrium distance) and ε (potential well depth, representing the strength of attraction), which are suitable for normal interatomic interaction; and new parameters σ' and ε', which are anti-aggregation parameters for limited atoms (aggregating atoms), and the calculated repulsive force is stronger, which is specifically used to prevent aggregation.

[0071] The to-be-sent information of the target atom in each storage unit can be specifically understood as a parameter table for recording the information sent by the atom participating in the specified position type calculation task in the storage unit to the specified calculation unit, such as bond_information. The specified position type calculation task can be specifically understood as a calculation task that needs to be specifically corrected for atomic position or interaction. Correspondingly, the information that the target atom needs to send to the specified calculation unit is the to-be-sent information.

[0072] The electrostatic parameter can be specifically understood as a parameter table recording the atomic charge q (corresponding to charge in atom_list), such as elec_param_table.

[0073] The atomic information provides a basis for calculating the force and updating the position of the atom, ensuring the feasibility of the simulation, the atomic group information clearly defines the molecular attribution, avoiding the misapplication of additional forces to atoms in the same molecule, ensuring the accuracy of molecular interaction calculation, the near-neighbor atom index records the atoms that need to be calculated for interaction in advance, greatly reducing invalid operations, reducing the cost of calculation, and improving the efficiency and stability of long-time scale simulation, the van der Waals force parameter contains both original parameters and new parameters, which can ensure the physical authenticity of normal molecular interaction and specifically inhibit aggregation, the to-be-sent information of the target atom in each storage unit ensures that the calculation units participating in cooperation share consistent data, avoiding deviation and reducing resource waste, and the electrostatic parameter ensures the completeness of the atomic force calculation. Through the synergistic effect of these information, the accuracy and reliability of long-term simulation are improved, providing a calculation basis for protein-ligand interaction analysis.

[0074] The technical scheme of the embodiment of the present application is that, by analyzing the simulation result of the protein system molecular dynamics, a plurality of restricted atoms needing anti-aggregation treatment are identified in the small molecule ligand meeting the aggregation condition, the corresponding target computing unit is allocated in the three-dimensional grid structure hardware based on the position information, and the additional information for calculating the additional repulsive force is added to the data parameters and stored in the associated storage unit; each target computing unit calculates the van der Waals force, electrostatic force and stress between the restricted atoms in parallel, and then obtains new position information as the anti-aggregation simulation result at the current time step and additionally stores it, while dynamically updating the target computing unit and the associated storage unit according to the new position, and repeating the calculation until the preset time step is completed. By positioning the restricted atoms needing anti-aggregation treatment and allocating target computing units for them, the hardware parallel computing capability can be fully utilized, the calculation and storage are bound nearby to reduce the delay, the dynamic adaptation to the change of atomic position ensures the continuity of the simulation, and the efficiency and accuracy of the anti-aggregation simulation are improved, which ensures the stable simulation of long time scale and effectively prevents the non-physical aggregation of small molecule ligands during the simulation process, thereby improving the accuracy of the small molecule ligand aggregation behavior simulation.

[0075] Embodiment two

[0076] Figure 2 The flowchart of another molecular anti-aggregation simulation method provided by the second embodiment of the present application is a refinement of the above-mentioned "updating the new target computing unit and the associated storage unit corresponding to each restricted atom according to the new position description information" in the above-mentioned embodiment, which can specifically include: selecting the out-of-bound atoms exceeding the coverage boundary of the target computing unit from the restricted atoms by each target computing unit according to the new position description information, and determining the new target computing unit based on the new position description information of the out-of-bound atoms; constructing each target migration list based on the data parameter information of each out-of-bound atom and the corresponding new target computing unit by the target computing unit of each out-of-bound atom, and sending each target migration list to the corresponding new target computing unit; and appending the data parameter information of the out-of-bound atoms in the received target migration list to the data parameter information in the associated storage unit by each new target computing unit.

[0077] Correspondingly, as shown in Figure 2 , the method comprises:

[0078] S210, analyzing the simulation result of the protein system molecular dynamics, and identifying a plurality of restricted atoms needing anti-aggregation treatment in the small molecule ligand meeting the aggregation condition.

[0079] S220, determining the target computing unit corresponding to each restricted atom in the three-dimensional grid structure hardware according to the position description information in the data parameter information of each restricted atom extracted in the simulation result.

[0080] S230, after adding the additional information for calculating the additional repulsive force between the confined atoms in the data parameter information of each confined atom, storing the data parameter information into the associated storage unit of the target calculation unit.

[0081] S240, through each target calculation unit, calculating the van der Waals force and the electrostatic force between each confined atom according to the data parameter information in the associated storage unit in parallel, and determining the force between each confined atom according to the calculation result.

[0082] S250, through each target calculation unit, calculating the new position description information of each confined atom as the anti-aggregation simulation result at the current time step according to the force between each confined atom and the data parameter information in the associated storage unit in parallel, and appending the new position description information to the data parameter information in the associated storage unit.

[0083] S260, through each target calculation unit, selecting the out-of-boundary atom from the confined atoms according to the new position description information, and determining the new target calculation unit based on the new position description information of the out-of-boundary atom.

[0084] It can be understood that in the three-dimensional grid structure hardware, each target calculation unit is responsible for processing the atoms in a specific region in the three-dimensional space, and the boundary of the region is determined by the grid division rule. For example, if the lengths of the simulation box in the x, y and z directions are Lx, Ly and Lz respectively, and the node numbers of the three-dimensional grid in the three dimensions are X, Y and Z, then the coverage boundary of CU(x, y, z) is: x direction: [x×(Lx / X), (x+1)×(Lx / X)), y direction: [y×(Ly / Y), (y+1)×(Ly / Y)), and z direction: [z×(Lz / Z), (z+1)×(Lz / Z)]. If the position of an atom falls within the range, it belongs to the target CU; if it exceeds the range, it is an out-of-boundary atom.

[0085] Specifically, each target calculation unit iterates the data parameter information (such as atom_list) of its associated storage unit to check whether the new position description information of each confined atom exceeds its coverage boundary, and if so, it is determined as an out-of-boundary atom that exceeds the coverage boundary of the target calculation unit.

[0086] When the out-of-boundary atom does not exceed the overall boundary of the three-dimensional grid hardware, i.e., only exceeds the boundary of the current target calculation unit, the new grid coordinates corresponding to the new target calculation unit can be calculated directly from the new position coordinates in the new position description information of the out-of-boundary atom based on CU(x, y, z) = floor(position / box_size×[X, Y, Z]).

[0087] When the out-of-bound atoms exceed the overall boundary of the three-dimensional grid hardware, i.e., exceed the boundary of the simulation box, the periodic boundary condition can be used to correct the atom positions back into the simulation box by using a periodic mapping formula:

[0088] position_new = position_old - box_size * floor(position_old / box_size) to correct the atom positions back into the simulation box, and then determine the corresponding new target computing unit. Here, box_size is the length of the simulation box in the corresponding direction, position_old is the position coordinate of the out-of-bound atom in the corresponding direction before correction, and position_new is the position coordinate of the out-of-bound atom in the corresponding direction after correction. Based on CU(x, y, z) = floor(position / box_size * [X, Y, Z]), the new grid coordinates are calculated according to the corrected position coordinates in the corresponding direction, corresponding to the new target computing unit.

[0089] It can be understood that the boundary of the three-dimensional grid hardware corresponds to the boundary of the simulation box, and the molecular dynamics simulation uses the periodic boundary condition, i.e., the outside of the box is a mirror image of the box. When the confined atoms exceed the boundary of the box, they are mapped into the opposite boundary and assigned to the corresponding CU, which is equivalent to the atoms going out of one edge of the box and coming in from the opposite edge, maintaining the continuity of the system and avoiding simulation distortion caused by the disappearance of the boundary.

[0090] S270, through the target computing unit of each out-of-bound atom, based on the data parameter information of each out-of-bound atom and the corresponding new target computing unit, a target migration list is constructed for each out-of-bound atom, and the target migration list is sent to the corresponding new target computing unit.

[0091] Specifically, after the new target computing unit of the out-of-bound atom is determined, the original target computing unit creates a target migration list for each out-of-bound atom based on the data parameter information of each out-of-bound atom and the corresponding new target computing unit. The target migration list can specifically include: the serial number id of the out-of-bound atom, the coordinates of the new target computing unit, the latest position description information, the electrostatic parameter and the van der Waals force parameter (such as additional information including new parameters), etc. The original target computing unit sends the constructed target migration list to the corresponding new target computing unit through near-neighbor communication.

[0092] Correspondingly, after the target migration list is sent, the original target computing unit can delete the information related to the out-of-bound atom from its associated storage unit, such as the atom_list and neighbor_list data structures, to avoid calculation conflicts caused by the same atom being processed by two computing units at the same time.

[0093] S280, by each new target computing unit, after appending and storing the data parameter information of the out-of-bound atoms in the received target migration list to the data parameter information of the associated storage unit, return to execute the operation of calculating the van der Waals force and the electrostatic force between each confined atom in parallel according to the data parameter information in the associated storage unit until the simulation of the preset number of time steps is completed.

[0094] Specifically, when the new target computing unit receives the target migration list sent by the original target computing unit, the complete data parameter information of the out-of-bound atoms is extracted therefrom and appended to the data structure of the data parameter information of the associated storage unit of the new target computing unit, for example, appended to atom_list, while updating data tables such as neighbor_list and molecule_list, to ensure that the out-of-bound atoms are included in the processing range of the new target computing unit.

[0095] Return to execute the operation of calculating the van der Waals force and the electrostatic force between each confined atom in parallel according to the data parameter information in the associated storage unit until the simulation of the preset number of time steps is completed.

[0096] Further, on the basis of each of the above embodiments, the molecular anti-aggregation simulation method can further include:

[0097] By each target computing unit, the data parameter information of the confined atoms of the associated storage unit is sent to the global control unit in the three-dimensional grid structure hardware;

[0098] By the global control unit, the received data parameter information is integrated as a file output;

[0099] By the global control unit, the system performance index is calculated based on the received data parameter information.

[0100] In the embodiments of the application, the global control unit can be specifically understood as: a computing unit in the three-dimensional grid structure hardware for overall planning of global parameters, centralized management of data and execution of system-level analysis, which can be undertaken by a specific computing unit (such as a computing unit with CU(0,0,0) coordinates), is a core computing unit related to simulation result output and performance evaluation, and executes a main process or a root process corresponding to the parallel program executed by each target computing unit.

[0101] Specifically, each target computing unit sends the data parameter information of the confined atoms of the associated storage unit to the global control unit in the three-dimensional grid structure hardware at a preset time period (such as every preset number of time steps or a preset simulation time).

[0102] After receiving the data parameter information of all target computing units, the global control unit integrates the local level data of each target computing unit into complete system level data, which can specifically include: atomic information of the whole system (such as the position and velocity information of all restricted atoms), total energy of the system, kinetic energy, potential energy and thermodynamic parameters (such as temperature and pressure), etc. The integrated data is written into an output file, such as a trajectory file and an energy file, and is stored in a standardized file for subsequent offline visualization or in-depth analysis.

[0103] The global control unit calculates system performance indicators based on the received data parameter information, performs system level performance evaluation, which can specifically include: long period evaluation and short period monitoring.

[0104] The long period evaluation can be specifically understood as: performing evaluation operation according to a longer preset period, which can specifically include: calculating ligand-protein binding free energy, evaluating the binding strength of ligand and protein, detecting whether the ligand can be stably combined; tracking the complete path of the ligand from free diffusion to the binding site, identifying the key intermediate state; analyzing whether the ligand remains dispersed in a long time scale (such as whether the aggregation event frequency and aggregation scale are lower than the threshold value), and evaluating the long-term effect of the anti-aggregation strategy.

[0105] The short period monitoring can be specifically understood as: performing evaluation operation according to a shorter preset period, which can specifically include: calculating the aggregation degree of small molecule ligand by calculating the average distance between ligands and the number of aggregation clusters (such as a cluster formed by more than 3 ligands is considered as aggregation), and evaluating whether the anti-aggregation strategy is effective in real time; counting the contact frequency (such as the frequency of hydrogen bond and hydrophobic interaction) between ligand and protein active site, judging whether the ligand is correctly combined or randomly aggregated, and evaluating the contact situation between protein and ligand; monitoring whether the temperature and energy fluctuation are within a reasonable range, evaluating the stability of the three-dimensional grid structure hardware, and ensuring the physical reasonableness of the simulation.

[0106] The local level restricted atom data of each target computing unit is sent to the global control unit to provide information basis for global analysis, avoiding the evaluation of the overall state of the system due to data fragmentation; the global control unit integrates the data and outputs it in the form of a file to form a complete system level record; on this basis, the global control unit calculates the system performance indicators, which can quantitatively evaluate the physical effectiveness of the simulation and the actual effect of the anti-aggregation strategy, and can provide decision basis for specific application scenarios such as drug design (such as ligand binding strength and identification of key intermediate states), improving the reliability and application value of molecular anti-aggregation simulation.

[0107] The technical scheme of the embodiment of the present application is that, by analyzing the simulation result of the molecular dynamics of the protein system, a plurality of restricted atoms needing anti-aggregation treatment are identified in the small molecule ligand meeting the aggregation condition, the corresponding target calculation unit is allocated in the three-dimensional grid structure hardware based on the position information, and the additional information for calculating the additional repulsive force is added to the data parameters and stored in the associated storage unit; each target calculation unit calculates the van der Waals force, electrostatic force and stress between the restricted atoms in parallel, and then obtains new position information as the anti-aggregation simulation result at the current time step and appends the storage, through each target calculation unit, the out-of-bound atoms are selected from the restricted atoms according to the new position description information, and the new target calculation unit is determined; through the target calculation unit of each out-of-bound atom, each target migration list is constructed and sent to the new target calculation unit; through each new target calculation unit, the data parameter information in the received target migration list is appended to the associated storage unit; the calculation is repeated until the preset time step is completed. By screening the out-of-bound atoms and reallocating the target calculation unit, the calculation discontinuity after the atoms cross the boundary can be avoided, the continuity and accuracy of the simulation are ensured, the target migration list is constructed based on the out-of-bound atom information and is sent directionally, the unordered communication across units can be reduced, the data migration efficiency is optimized, the new target calculation unit appends the migration data to the associated storage unit, the data integrity is ensured, the near storage is realized, the data access delay is reduced, and the adaptive ability of the hardware architecture to the dynamic motion of the atoms is enhanced, especially the stability of the simulation of the large-scale and long-time complex Swimming system is improved.

[0108] Embodiment three

[0109] Figure 3 The flowchart of another anti-aggregation simulation method of molecules provided by the third embodiment of the present application, this embodiment is a refinement of the above-mentioned embodiment "parallel calculation of van der Waals force and electrostatic force between each restricted atom according to the data parameter information in the associated storage unit", which can specifically include: when the distance between the restricted atoms calculated according to the data parameter information is less than the preset cutoff radius, the van der Waals force between the restricted atoms is calculated according to the additional information for calculating the additional repulsive force between the restricted atoms in the data parameter information and the distance between the restricted atoms, and the van der Waals force is accumulated in the short-range force between the restricted atoms; the electrostatic force between the restricted atoms is calculated according to the electrostatic parameters of the restricted atoms in the data parameter information and the distance between the restricted atoms, and the electrostatic force is accumulated in the short-range force between the restricted atoms; return to perform the operation of calculating the van der Waals force between the restricted atoms according to the additional information for calculating the additional repulsive force between the restricted atoms in the data parameter information and the distance between the restricted atoms, until each van der Waals force and electrostatic force between the restricted atoms in the data parameter information is calculated.

[0110] Correspondingly, as shown in Figure 3 , the method comprises:

[0111] S310, analyze the simulation result of the protein system molecular dynamics, and identify a plurality of restricted atoms needing anti-aggregation treatment in the small molecule ligand meeting the aggregation condition.

[0112] S320, determine a target computing unit corresponding to each restricted atom in the three-dimensional grid structure hardware according to the position description information in the data parameter information of each restricted atom extracted in the simulation result.

[0113] S330, after adding additional information for calculating the additional repulsive force between the restricted atoms in the data parameter information of each restricted atom, store each data parameter information to the associated storage unit of the target computing unit.

[0114] S340, through each target computing unit, when the distance between the restricted atoms calculated according to the data parameter information is less than the preset cutoff radius, calculate the van der Waals force between the restricted atoms according to the additional information for calculating the additional repulsive force between the restricted atoms in the data parameter information and the distance between the restricted atoms, and accumulate the van der Waals force to the short-range force between the restricted atoms.

[0115] It can be understood that in molecular dynamics, the interaction force (such as van der Waals force) between atoms quickly decays with the increase of the distance between atoms. Therefore, the preset cutoff radius can be used as a calculation threshold in the simulation: when the distance between two restricted atoms is less than the radius, the van der Waals force (belonging to short-range force) between the restricted atoms is calculated; if the distance is greater than the cutoff radius, the interaction can be ignored to avoid meaningless calculation and improve efficiency.

[0116] Specifically, each target computing unit obtains the positions of each restricted atom and calculates the distance between the restricted atoms (r ij ) according to the data parameter information, and when the distance between the restricted atoms is less than the preset cutoff radius, queries the new parameters σ' and ε' in the additional information from the data parameter information (such as vdw_param_table), σ' is usually smaller than the original σ (so that the repulsive force acts earlier), and ε' is greater than the original ε (to enhance the strength of the repulsive force), so that a stronger repulsive force is generated when the atoms are close (close to aggregation), which mechanically prevents aggregation. Based on the van der Waals force F vdw between the restricted atoms is calculated, where, represents the attractive force component, which quickly decays with the increase of the distance, represents the repulsive force component, which plays a leading role at close distances.

[0117] It can be understood that different restricted atoms can be set with different new parameters σ' and ε', that is, the restricted atom 1 is σ'1 and ε'1, and the restricted atom type 2 is σ'2 and ε'2, so that the new interaction parameters between the restricted atoms 1 and 2 can be σ' = (σ'1 + σ'2) / 2.

[0118] The whole process is executed in parallel in each target computing unit, and after the van der Waals force between the confined atoms is calculated, it is accumulated into the short-range force (force vector) of atoms i and j, and the direction is opposite (the force of i to j is equal in size and opposite in direction to the force of j to i).

[0119] S350, according to the electrostatic parameters of the confined atoms in the data parameter information and the distance between the confined atoms, the electrostatic force between the confined atoms is calculated, and the electrostatic force is accumulated into the short-range force between the confined atoms.

[0120] It can be understood that the electrostatic force is essentially a long-range force (inversely proportional to the square of the distance), but in molecular simulation, in order to improve efficiency, the short-range part within the cutoff radius can be calculated.

[0121] Specifically, under the premise that the distance between the two confined atoms is less than the cutoff radius, according to the electrostatic parameters (such as atomic charges q1 and q2 in elec_param_table) of the confined atoms in the data parameter information and the distance (r ij ) between the confined atoms, the electrostatic force F elec between the confined atoms is calculated based on , wherein r is the Coulomb conversion factor, which is used to unify the charge unit and the distance unit, etc. to the force unit in simulation, to ensure dimensional consistency, erfc(·) is the complementary error function, which is used to smooth the truncation of the short-range force, exp(·) is the natural exponential function, a = λ / cutoff_radius, a is a coefficient related to the error parameter λ (used to control the decay speed of the correction term) and the cutoff radius cutoff_radius. is the dominant term, and the form is close to the traditional Coulomb force. is the short-range correction term, which ensures that the correction term can offset the sudden change of the dominant term at the cutoff radius, ensuring continuity.

[0122] The whole process is executed in parallel in each target computing unit, and after the van der Waals force between the confined atoms is calculated, it is accumulated into the short-range force (force vector) of atoms i and j, and the direction is opposite (the force of i to j is equal in size and opposite in direction to the force of j to i).

[0123] S360, return to execute the operation of calculating the van der Waals force between the confined atoms according to the additional information for calculating the additional repulsive force between the confined atoms in the data parameter information and the distance between the confined atoms, until the van der Waals force and the electrostatic force between each confined atom in the data parameter information are calculated, and the force between each confined atom is determined according to the calculation result.

[0124] Specifically, returning to perform the operation of calculating the van der Waals force between the restricted atoms according to the additional information for calculating the additional repulsive force between the restricted atoms in the data parameter information and the restricted atomic distance, until the van der Waals force and the electrostatic force between each restricted atom in the data parameter information are calculated and completed, and the total force between each restricted atom is obtained.

[0125] Further, on the basis of each of the above embodiments, after the van der Waals force and the electrostatic force between each restricted atom in the data parameter information are calculated and completed, the following operations can also be included:

[0126] By the global control unit, the first correction atom list and the corresponding first additional information for calculating the additional repulsive force between the correction atoms are obtained from the first correction task file.

[0127] By the global control unit, the first correction information and the corresponding first additional information in the first correction atom list are sent to the corresponding first calculation unit according to the data parameter information of each restricted atom.

[0128] By each first calculation unit, the first correction distance is calculated in parallel according to the received first correction information.

[0129] By each first calculation unit, the first correction force is calculated in parallel according to the first correction distance and the first additional information, and the first correction force is accumulated into the corresponding short-range force between the restricted atoms.

[0130] By the global control unit, the second correction atom list and the corresponding original additional information for calculating the original repulsive force between the correction atoms are obtained from the second correction task file.

[0131] By the global control unit, the second correction information and the corresponding original additional information in the second correction atom list are sent to the corresponding second calculation unit according to the data parameter information of each restricted atom.

[0132] By each second calculation unit, the second correction distance is calculated in parallel according to the received second correction information.

[0133] By each second calculation unit, the second correction force is calculated in parallel according to the second correction distance and the original additional information, and the second correction force is accumulated into the corresponding short-range force between the restricted atoms.

[0134] It can be understood that, in the above calculation process, in order to achieve the anti-aggregation effect, the calculation is performed on all the restricted atoms (including the restricted atoms in the same small molecule) based on the newly added parameter of the enhanced repulsion. However, the atoms in the same small molecule need to maintain structural stability, and if the parameter of the enhanced repulsion is used, the normal structure inside the molecule will be destroyed. Therefore, the above calculation needs to be corrected (corresponding to the first correction task) to ensure that the newly added parameter of the enhanced repulsion only acts on the restricted atoms between different small molecules.

[0135] In the embodiments of the present application, the first correction task can be understood as a task of removing the van der Waals force between the restricted atoms in the same small molecule calculated based on the newly added parameter. The first correction information can be understood as the data parameter information of the restricted atoms related to the first correction task. The first correction atom list can be understood as a list recording the restricted atoms inside the same small molecule related to the first correction task. The first additional information can be understood as the newly added parameters σ' and ε' matched by each restricted atom in the first correction atom list.

[0136] The second correction task can be understood as a task of supplementing the normal van der Waals force between the restricted atoms inside the small molecule calculated based on the original parameters. The second correction information can be understood as the data parameter information of the restricted atoms related to the second correction task. The second correction atom list can be understood as a list recording the restricted atoms inside the same small molecule related to the second correction task. The original additional information can be understood as the original parameters σ and ε matched by each restricted atom in the second correction atom list.

[0137] Specifically, the global control unit reads the first correction atom list and its corresponding first additional information for calculating the additional repulsive force between the correction atoms from the first correction task file. According to the data parameter information of each restricted atom (such as the position coordinates of the atom), the corresponding target calculation unit is determined as the first calculation unit, and the atom data participating in the designated position class calculation task in the associated storage unit (such as bond_information) of the global control unit is sent to the corresponding first calculation unit, that is, the first correction information and the corresponding first additional information in the first correction atom list are sent to the corresponding first calculation unit. Each first calculation unit calculates the first correction distance (r ij ) in parallel according to the received first correction information, and combines the first additional information to calculate the first correction force F correction in parallel based on The first correction force F correction is added to the short-range force (force vector) of atoms i and j, and the direction is opposite (the force of i on j is equal in size and opposite in direction to the force of j on i).

[0138] It can be understood that, on the basis of the first correction task, a second correction task should also be performed to restore the normal van der Waals interaction between the limited atoms in the small molecule.

[0139] Specifically, the global control unit reads the second correction atom list and its corresponding original additional information for correcting and restoring the original repulsion between atoms from the second correction task file. According to the data parameter information of each limited atom (such as the position coordinates of the atom), a corresponding target calculation unit is determined as a second calculation unit, and the atom data participating in the designated position class calculation task in the associated storage unit (such as bond_information) of the global control unit is sent to the corresponding second calculation unit, that is, the second correction information in the second correction atom list and the corresponding original additional information are sent to the corresponding second calculation unit. Each second calculation unit calculates the second correction distance (r ij ) in parallel according to the received second correction information, and combines the original additional information to calculate the second correction force F in parallel based on real , and adds the second correction force to the short-range force (force vector) of atoms i and j, and the direction is opposite (the force of i to j is equal in size and opposite in direction to the force of j to i).

[0140] The first correction is located by the global control unit to atom pairs in the same small molecule and distributes tasks, and the calculation unit calculates the reverse correction force to eliminate the wrong repulsive force caused by the use of new parameters, avoiding the destruction of the internal structure of the molecule; The second correction is obtained by the global control unit to obtain the original parameters of the atom pairs in the same molecule and distribute tasks, and the calculation unit restores the normal van der Waals force to ensure that the internal force of the molecule meets the physical law, realizes the separation of the force field between molecules (retaining the anti-aggregation force) and the internal molecule (only using the original force), balances the anti-aggregation target and the physical authenticity, so that the simulation results can support the simulation evaluation of molecular anti-aggregation and reflect the real behavior of molecules; At the same time, the design of global distribution and parallel calculation controls the time cost, ensures the efficient operation of large-scale system simulation, and improves the reliability and application value of molecular dynamics simulation in the field of drug design.

[0141] Further, on the basis of the above embodiments, after the second correction force is added to the corresponding long-range force between the limited atoms of the modified atoms, it can also include:

[0142] Through each target calculation unit, the real-space electrostatic force and the inverse-space electrostatic force between the limited atoms are calculated according to the electrostatic parameters of the limited atoms and the distance between the limited atoms.

[0143] Through each target calculation unit, the real-space electrostatic force and the inverse-space electrostatic force calculated are added to obtain the long-range electrostatic force between the limited atoms.

[0144] Specifically, the long-range electrostatic force between the confined atoms can be calculated based on a k-space Gaussian Split Ewald method (KGSE). Each target computing unit iterates through the confined atoms in the unit, calculates the distance (r ij ) between the confined atoms, combines the electrostatic parameters (q1 and q2) of the confined atoms, and calculates the real-space electrostatic force F reals between the confined atoms based on the KGSE method, where a is an Ewald parameter used to control the accuracy of the real-space truncation.

[0145] Each computing unit accumulates the real-space electrostatic force and the reciprocal-space electrostatic force to obtain the long-range electrostatic force between the confined atoms.

[0146] Finally, the total force between the confined atoms is the cumulative force of the short-range force and the long-range electrostatic force.

[0147] By processing the short-range and rapidly changing electrostatic interaction in the real space and the long-range and slowly changing interaction in the reciprocal space through the three-dimensional fast Fourier transform, the loss of long-range information caused by simple truncation is avoided, the problem of large computational load for large-scale systems is solved, the long-range electrostatic effect is captured to ensure that the molecular dynamic behavior conforms to the physical law and avoid non-physical phenomena, the parallel design is adapted to the three-dimensional grid hardware architecture, each computing unit independently processes local tasks and cooperatively participates in global calculation, the simulation capability can be extended to a large number of atomic complex systems, the limitations of small-scale simulation are broken, and the precision, efficiency, and application range of the anti-aggregation simulation of molecules are improved.

[0148] S370, through each target computing unit, according to the force between each confined atom and the data parameter information in the associated storage unit, parallel computing the new position description information of each confined atom as the anti-aggregation simulation result at the current time step, and appending the data parameter information in the associated storage unit.

[0149] S380, after updating the new target computing unit and the associated storage unit corresponding to each confined atom according to the new position description information, returning to perform the operation of parallel computing the van der Waals force and the electrostatic force between each confined atom according to the data parameter information in the associated storage unit until the simulation of the preset number of time steps is completed.

[0150] ​The technical scheme of the embodiment of the present application is characterized in that: the simulation result of the protein system molecular dynamics is analyzed, a plurality of restricted atoms needing anti-aggregation treatment are identified in the small molecule ligand meeting the aggregation condition, the corresponding target computing unit is allocated in the three-dimensional grid structure hardware based on the position information, and the additional information for calculating the additional repulsive force is added to the data parameters and stored in the associated storage unit; by calculating the van der Waals force and the electrostatic force only for the restricted atom pairs with a distance less than a preset cutoff radius, invalid operation is avoided, distributed splitting of tasks is realized in combination with parallel computing, and the short-range force calculation efficiency and resource utilization rate are improved; the additional information of the additional repulsive force is introduced in the van der Waals force calculation, the interaction between the atoms in the short range that may cause aggregation is inhibited, and the normal long-range interaction is not interfered, the accuracy and effectiveness of the anti-aggregation strategy are enhanced; by traversing and calculating all the atom pairs meeting the condition and adding the van der Waals force and the electrostatic force to the short-range force, it is ensured that the force calculation is not missed and the interaction contribution is comprehensively reflected, accurate mechanical basis is provided for position updating, and the integrity and accuracy of the simulation result are ensured; meanwhile, each target computing unit operates independently based on local data, cross-unit dependence is reduced, repeated calculation or conflict is avoided, and the cooperativeness and stability of parallel computing are strengthened; the force between each restricted atom is determined according to the calculation result, and then the new position information is obtained as the anti-aggregation simulation result of the current time step and is additionally stored, the target computing unit and the associated storage unit are dynamically updated according to the new position, and the calculation is repeated until the preset time step is completed, the efficiency and precision of the anti-aggregation simulation are improved, and stable simulation of a long time scale is ensured, which can effectively prevent the non-physical aggregation of the small molecule ligand in the simulation process and improve the accuracy of the small molecule ligand aggregation behavior simulation.

[0151] Embodiment four

[0152] Figure 4 A structural schematic diagram of a molecular anti-aggregation simulation device provided by the fourth embodiment of the present application is shown in FIG. 4. Figure 4 As shown in the figure, the device comprises a restricted identification module 410, a unit determination module 420, an additional information module 430, a force calculation module 440, an additional storage module 450 and a return execution module 460, wherein:

[0153] The restricted identification module 410 is configured to analyze the simulation result of the protein system molecular dynamics, and identify a plurality of restricted atoms needing anti-aggregation treatment in the small molecule ligand meeting the aggregation condition;

[0154] The unit determination module 420 is configured to determine the target computing unit corresponding to each restricted atom in the three-dimensional grid structure hardware according to the position description information in the data parameter information of each restricted atom extracted in the simulation result;

[0155] The additional information module 430 is configured to add additional information for calculating the additional repulsive force between the confined atoms to the data parameter information of each confined atom, and store the data parameter information into the associated storage unit of the target calculation unit;

[0156] The force calculation module 440 is configured to calculate the van der Waals force and the electrostatic force between the confined atoms in parallel according to the data parameter information in the associated storage unit by using the target calculation unit, and determine the force between the confined atoms according to the calculation result;

[0157] The additional storage module 450 is configured to calculate new position description information of each confined atom as the anti-aggregation simulation result at the current time step according to the force between the confined atoms and the data parameter information in the associated storage unit by using the target calculation unit in parallel, and additionally store the new position description information into the data parameter information in the associated storage unit;

[0158] The return execution module 460 is configured to update the new target calculation unit and the associated storage unit corresponding to each confined atom according to the new position description information, and return to execute the operation of calculating the van der Waals force and the electrostatic force between the confined atoms in parallel according to the data parameter information in the associated storage unit until the simulation of the preset number of time steps is completed.

[0159] The technical scheme of the embodiment of the present application is that, by analyzing the simulation result of the protein system molecular dynamics, a plurality of confined atoms needing anti-aggregation treatment are identified in the small molecule ligand meeting the aggregation condition, the corresponding target calculation unit is allocated in the three-dimensional grid structure hardware based on the position information of the confined atoms, and the additional information for calculating the additional repulsive force is added to the data parameter and stored in the associated storage unit. The target calculation unit calculates the van der Waals force, the electrostatic force and the force between the confined atoms in parallel, and then obtains the new position information as the anti-aggregation simulation result at the current time step and additionally stores the new position information. Meanwhile, the target calculation unit and the associated storage unit are dynamically updated according to the new position, and the calculation is repeated until the preset number of time steps is completed. By positioning the confined atoms needing anti-aggregation treatment and allocating the target calculation unit to the confined atoms, the hardware parallel computing capability can be fully utilized, the calculation and storage can be bound to reduce the delay, the simulation continuity can be ensured by dynamically adapting the change of the atomic position, and the efficiency and accuracy of the anti-aggregation simulation are improved. The long-time-scale stable simulation can be ensured, the non-physical aggregation of the small molecule ligand in the simulation process can be effectively prevented, and the accuracy of the simulation of the aggregation behavior of the small molecule ligand is improved.

[0160] On the basis of the above-mentioned embodiments, the return execution module 460 is specifically configured to:

[0161] The target computing unit selects, from the limited atoms, a cross-border atom that exceeds the coverage boundary of the target computing unit according to the new position description information, and determines a new target computing unit based on the new position description information of the cross-border atom;

[0162] The target computing unit of each cross-border atom constructs a target migration list based on the data parameter information of each cross-border atom and the corresponding new target computing unit, and sends each target migration list to the corresponding new target computing unit.

[0163] The new target computing unit appends the data parameter information of the cross-border atom in the received target migration list to the data parameter information of the associated storage unit.

[0164] Further, on the basis of each of the above embodiments, the molecular anti-aggregation simulation device can further include a global sending module, a global receiving module, and a global computing module, wherein:

[0165] The global sending module is configured to send, by each target computing unit, the data parameter information of the limited atoms of the associated storage unit to a global control unit in the three-dimensional grid structure hardware.

[0166] The global receiving module is configured to integrate the received data parameter information as a file output by the global control unit.

[0167] The global computing module is configured to calculate a system performance index based on the received data parameter information by the global control unit.

[0168] On the basis of each of the above embodiments, the stress calculation module 440 is specifically configured to:

[0169] When the distance between the limited atoms calculated according to the data parameter information is less than the preset cutoff radius, the van der Waals force between the limited atoms is calculated according to the additional information for calculating the additional repulsive force between the limited atoms in the data parameter information and the distance between the limited atoms, and the van der Waals force is accumulated in the short-range force between the limited atoms.

[0170] The electrostatic force between the limited atoms is calculated according to the electrostatic parameter of the limited atom in the data parameter information and the distance between the limited atoms, and the electrostatic force is accumulated in the short-range force between the limited atoms.

[0171] Return to the operation of calculating the van der Waals force between the limited atoms according to the additional information for calculating the additional repulsive force between the limited atoms in the data parameter information and the distance between the limited atoms, until the van der Waals force and the electrostatic force between each limited atom in the data parameter information are calculated.

[0172] Optionally, on the basis of each of the above embodiments, the stress calculation module 440 comprises: a first task unit, a first sending unit, a first receiving unit, a first calculation unit, a second task unit, a second sending unit, a second receiving unit and a second calculation unit, wherein:

[0173] The first task unit is configured to, after the van der Waals forces and the electrostatic forces between the limited atoms in the data parameter information are calculated, obtain, by the global control unit, a first modified atom list and corresponding first additional information for calculating additional repulsive forces between the modified atoms from a first modified task file;

[0174] The first sending unit is configured to, by the global control unit, send the first modified information in the first modified atom list and the corresponding first additional information to the corresponding first calculation unit according to the data parameter information of each limited atom;

[0175] The first receiving unit is configured to, by each first calculation unit, calculate a first modified distance in parallel according to the received first modified information;

[0176] The first calculation unit is configured to, by each first calculation unit, calculate a first modified force in parallel according to the first modified distance and the first additional information, and accumulate the first modified force into the corresponding short-range force between the limited atoms between the modified atoms;

[0177] The second task unit is configured to, by the global control unit, obtain a second modified atom list and corresponding original additional information for calculating original repulsive forces between the modified atoms from a second modified task file;

[0178] The second sending unit is configured to, by the global control unit, send the second modified information in the second modified atom list and the corresponding original additional information to the corresponding second calculation unit according to the data parameter information of each limited atom;

[0179] The second receiving unit is configured to, by each second calculation unit, calculate a second modified distance in parallel according to the received second modified information;

[0180] The second calculation unit is configured to, by each second calculation unit, calculate a second modified force in parallel according to the second modified distance and the original additional information, and accumulate the second modified force into the corresponding short-range force between the limited atoms between the modified atoms.

[0181] Optionally, on the basis of each of the above embodiments, the second calculation unit comprises: a spatial electrostatic force subunit and a long-range electrostatic force subunit, wherein:

[0182] a spatial electrostatic force subunit, configured to calculate, by each target calculation unit, real-space spatial electrostatic force and inverse-space spatial electrostatic force between the limited atoms according to the electrostatic parameters of the limited atoms and the limited interatomic distance after the second correction force is added to the corrected interatomic corresponding limited interatomic short-range force;

[0183] a long-range electrostatic force subunit, configured to add, by each target calculation unit, the calculated real-space spatial electrostatic force and inverse-space spatial electrostatic force to obtain long-range electrostatic force between the limited atoms.

[0184] On the basis of the above embodiments, the data parameter information can include: atomic information, group information, near-neighbor atomic index, van der Waals force parameter, to-be-sent information of target atoms in each storage unit, and electrostatic parameters.

[0185] The molecular anti-aggregation simulation device provided in the embodiments of the present application can perform the molecular anti-aggregation simulation method provided in any of the embodiments of the present application, and has the corresponding function modules and beneficial effects of the execution method.

[0186] In the technical solutions of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved in the technical solutions comply with relevant laws and regulations and do not violate public order and good customs.

[0187] Embodiment five

[0188] The three-dimensional grid structure hardware of the embodiments of the present application aims to build a parallel computing architecture adapted to molecular dynamics simulation, and cooperates with a three-dimensional grid point layout, a connected network and a functional unit to support large-scale and efficient molecular anti-aggregation simulation tasks, and can flexibly correspond to simulation requirements of different dimensions and scales.

[0189] Figure 5 A two-dimensional structural schematic diagram of three-dimensional grid structure hardware that can be used to implement the embodiments of the present application is shown, and the three-dimensional grid structure hardware includes:

[0190] a first dimension x a second dimension x a third dimension of grid points; each grid point includes a calculation unit and an associated storage unit; each calculation unit or associated storage unit is in communication connection with an adjacent calculation unit or associated storage unit; a calculation unit or associated storage unit located at the edge of the three-dimensional grid structure hardware is in communication connection with a calculation unit or associated storage unit located at the opposite edge of the three-dimensional grid structure hardware; and a global control unit in communication connection with each calculation unit; the global control unit is configured to issue parameters to the calculation units and coordinate calculation tasks;

[0191] Each associated storage unit is communicatively connected to a corresponding computing unit; wherein, the associated storage unit stores data parameter information as described in any embodiment of the present invention, and a computer program that can be executed by the at least one computing unit, the computer program being executed by the at least one computing unit to enable the at least one computing unit to execute the molecular anti-aggregation simulation method as described in any embodiment of the present invention.

[0192] Specifically, the 3D mesh structure hardware comprises grid points in a first dimension × a second dimension × a third dimension. Each grid point consists of a computing unit and an associated storage unit. The computing units or associated storage units of adjacent grid points communicate directly with each other, and the computing units or associated storage units of opposite edges communicate directly with each other, forming a closed communication network in 3D space.

[0193] In addition, the three-dimensional mesh structure hardware is equipped with a global control unit (the CU with position coordinates (0,0,0) can be selected as the global control unit). The global control unit is used to communicate with each computing unit and is responsible for issuing simulation parameters (such as simulation box size, cutoff radius, system temperature, initial time step and simulation step size) and coordinating parallel computing tasks (such as assigning different mesh points to process different atom calculations).

[0194] The associated storage unit is bound to the corresponding computing unit for communication. The associated storage unit is used to store the data and parameter information required for molecular simulation of its grid points, as well as the computer program for executing the molecular anti-aggregation simulation method, thereby supporting the corresponding computing unit to call data and run algorithms.

[0195] The computing unit executes the molecular anti-aggregation simulation method based on the computer program and data parameter information of the corresponding associated storage unit, and each computing unit can perform parallel computing, that is, different grid points can simultaneously execute the molecular anti-aggregation simulation method based on the data parameter information of their respective regions.

[0196] like Figure 5 As shown, taking a 3×3×3 grid layout as an example, the hardware is analyzed from a 2D perspective. The hardware is arranged in 3 rows and 3 columns of grid points. Each grid point consists of a computational unit and an associated storage unit. For example, grid point 11 contains both computational unit 11 and associated storage unit 11. The computational units or associated storage units of each grid point (connected vertically, horizontally, and vertically by border lines in the 2D diagram) can communicate, indicating connectivity between adjacent grid points, adjacent computational units, and adjacent storage units in the 3D structure. This connectivity includes the connectivity between edge computational units or associated storage units and the computational units or associated storage units on opposite edges. For example, edge grid point 11 can directly communicate with the opposite edge grid point 13, forming periodic boundary conditions.

[0197] It can be understood that the grid point layout dimension of the three-dimensional grid structure hardware can be extended to any first dimension x second dimension x third dimension.

[0198] It can be understood that, Figure 5 For example, each grid point contains one computing unit and one associated storage unit. In fact, each grid point can contain m computing units and n associated storage units. The three-dimensional grid structure hardware is connected through a high-speed interconnection network, supporting near-neighbor communication and global data broadcast. When the first dimension x second dimension x third dimension is X x Y x Z, the total number of computing units = m x X x Y x Z.

[0199] The components, their connections and relationships, and their functions shown in this paper are only examples and are not intended to limit the implementation of the invention described and / or claimed herein.

[0200] It can be understood that the computing unit in the three-dimensional grid structure hardware can be understood as a processor, and the associated storage unit can be understood as a storage connected in communication with the processor.

[0201] As shown in Figure 5 The three-dimensional grid structure hardware includes at least one processor and a memory connected in communication with the at least one processor, such as a read-only memory (ROM), a random access memory (RAM), etc., wherein the memory stores a computer program executable by the at least one processor, and the processor can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) or loaded into the random access memory (RAM) from the storage unit. In the RAM, various programs and data required for operation of the three-dimensional grid structure hardware can also be stored. The processor, ROM and RAM are connected in communication with each other.

[0202] The plurality of components in the three-dimensional grid structure hardware can be connected to an I / O interface, including: an input unit, such as a keyboard, a mouse, etc.; an output unit, such as various types of displays, speakers, etc.; a storage unit, such as a magnetic disk, an optical disk, etc.; and a communication unit, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit allows the three-dimensional grid structure hardware to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunications networks.

[0203] The processor can be various general and / or special purpose processing components with processing and computing capabilities. Some examples of the processor include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor performs various methods and processes described above, e.g., the molecular anti-aggregation simulation method, namely:

[0204] Analyzing the simulation results of the protein system molecular dynamics, identifying a plurality of constrained atoms that need to be subjected to anti-aggregation processing in the small molecule ligand that meets the aggregation condition;

[0205] According to the position description information in the data parameter information of each constrained atom extracted in the simulation results, determining the target computing unit corresponding to each constrained atom in the three-dimensional grid structure hardware;

[0206] After adding additional information for calculating the additional repulsive force between the constrained atoms in the data parameter information of each constrained atom, storing each data parameter information into the associated storage unit of the target computing unit;

[0207] Through each target computing unit, calculating the van der Waals force and electrostatic force between each constrained atom according to the data parameter information in the associated storage unit in parallel, and determining the force between each constrained atom according to the calculation result;

[0208] Through each target computing unit, calculating the new position description information of each constrained atom as the anti-aggregation simulation result at the current time step according to the force between each constrained atom and the data parameter information in the associated storage unit in parallel, and appending the data parameter information to the associated storage unit;

[0209] According to the new position description information, updating the new target computing unit and the associated storage unit corresponding to each constrained atom, and returning to perform the operation of calculating the van der Waals force and electrostatic force between each constrained atom according to the data parameter information in the associated storage unit in parallel until the simulation of the preset number of time steps is completed.

[0210] In some embodiments, the molecular anti-aggregation simulation method can be implemented as a computer program tangibly embodied in a computer readable storage medium, such as a storage unit. In some embodiments, part or all of the computer program can be loaded and / or installed on the three-dimensional grid structure hardware via the ROM and / or the communication unit. When the computer program is loaded into the RAM and executed by the processor, one or more steps of the molecular anti-aggregation simulation method described above can be performed. Alternatively, in other embodiments, the processor can be configured to perform the molecular anti-aggregation simulation method by any other appropriate means (e.g., by means of firmware).

[0211] Various implementations of the systems and techniques described above can be realized in digital electronic circuitry, integrated circuitry, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on a chip (SOC), a programmable logic device (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0212] Computer programs used to implement the methods of the application can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the computer program, when executed by the processor of the machine, implements the functions / acts specified in the flow diagrams and / or block diagrams. The computer program can be executed entirely on a machine, partially on a machine, partially on a machine as a standalone software package, partially on a machine and partially on a remote machine or entirely on a remote machine or server.

[0213] In the context of the present application, a computer-readable storage medium can be a tangible medium that can contain or store computer programs for use by or in connection with an instruction execution system, apparatus, or device. The computer-readable storage medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, the computer-readable storage medium can be a machine-readable signal medium. More specific examples of the machine-readable storage medium will include one or more lines of electrical connections, portable computer disks, hard disk drives, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or Flash memory), optical fibers, portable compact disc read-only memories (CD-ROMs), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0214] To provide for interaction with a user, the systems and techniques described here can be implemented on a three-dimensional grid structure hardware having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the three-dimensional grid structure hardware. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0215] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), blockchain network, and the Internet.

[0216] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. A server can be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system, to solve the defects of large management difficulty and weak business scalability in traditional physical host and VPS service.

[0217] It should be understood that the various forms of flow shown above can be re-ordered, added to, or deleted from without departing from the scope of the present disclosure. For example, the steps recited in the present disclosure can be executed in parallel, executed in sequence, or executed in different orders, as long as the desired results of the present disclosure are achieved, and the present disclosure is not limited herein.

[0218] The above detailed description does not limit the scope of the application. Various modifications, combinations, sub-combinations and alternatives can be made to the detailed embodiment within the scope of the application. Any modification, equivalent replacement and improvement made without departing from the spirit and principle of the application shall fall within the scope of the application.

Claims

1. A method of molecular anti-aggregation simulation, characterized in that, The simulation result of the protein system molecular dynamics is analyzed, and a plurality of restricted atoms needing anti-aggregation treatment are identified in the small molecule ligand meeting the aggregation condition. According to the position description information in the data parameter information of each restricted atom extracted in the simulation result, the target computing unit corresponding to each restricted atom is determined in the three-dimensional grid structure hardware. After adding additional information for calculating the additional repulsive force between the restricted atoms in the data parameter information of each restricted atom, the data parameter information is stored in the associated storage unit of the target computing unit. The van der Waals force and the electrostatic force between each restricted atom are calculated in parallel by each target computing unit according to the data parameter information in the associated storage unit, and the force between each restricted atom is determined according to the calculation result. The new position description information of each restricted atom is calculated as the anti-aggregation simulation result at the current time step by each target computing unit according to the force between each restricted atom and the data parameter information in the associated storage unit, and is appended to the data parameter information in the associated storage unit. According to the new position description information, the new target computing unit and the associated storage unit corresponding to each restricted atom are updated, and the operation of calculating the van der Waals force and the electrostatic force between each restricted atom according to the data parameter information in the associated storage unit is returned until the simulation of the preset number of time steps is completed. According to the new position description information, the new target computing unit and the associated storage unit corresponding to each restricted atom are updated, including:

2. The method of claim 1, wherein, By each target computing unit, the out-of-bound atoms exceeding the coverage boundary of the target computing unit are screened from the restricted atoms according to the new position description information, and the new target computing unit is determined based on the new position description information of the out-of-bound atoms. By the target computing unit of each out-of-bound atom, each target migration list is constructed based on the data parameter information of each out-of-bound atom and the corresponding new target computing unit, and each target migration list is sent to the corresponding new target computing unit. By each new target computing unit, the data parameter information of the out-of-bound atoms in the received target migration list is appended to the data parameter information in the associated storage unit. The molecular anti-aggregation simulation method further includes:

3. The method of claim 1, wherein, By each target computing unit, the data parameter information of the restricted atoms in the associated storage unit is sent to the global control unit in the three-dimensional grid structure hardware. By the global control unit, the received data parameter information is integrated as a file output. By the global control unit, the system performance index is calculated based on the received data parameter information. According to the data parameter information in the associated storage unit, the van der Waals force and the electrostatic force between each restricted atom are calculated in parallel, including:

4. The method of claim 1, wherein, When the distance between the restricted atoms calculated according to the data parameter information is less than the preset cutoff radius, the van der Waals force between the restricted atoms is calculated according to the additional information for calculating the additional repulsive force between the restricted atoms in the data parameter information and the distance between the restricted atoms, and the van der Waals force is accumulated in the short-range force between the restricted atoms. ​ The electrostatic force between the limited atoms is calculated according to the electrostatic parameters of the limited atoms in the data parameter information and the distance between the limited atoms, and the electrostatic force is accumulated into the short-range force between the limited atoms; The operation of calculating the van der Waals force between the limited atoms according to the additional information for calculating the additional repulsion force between the limited atoms in the data parameter information and the distance between the limited atoms is returned to execute until the van der Waals force and the electrostatic force between each limited atom in the data parameter information are traversed and calculated.

5. The method of claim 4, wherein, After the van der Waals force and the electrostatic force between each limited atom in the data parameter information are traversed and calculated, the method further comprises: The first correction atom list and the corresponding first additional information for calculating the additional repulsion force between the correction atoms are obtained from the first correction task file by the global control unit; The first correction information in the first correction atom list and the corresponding first additional information are sent to the corresponding first calculation unit by the global control unit according to the data parameter information of each limited atom; The first correction distance is calculated in parallel by each first calculation unit according to the received first correction information; The first correction force is calculated in parallel by each first calculation unit according to the first correction distance and the first additional information, and the first correction force is accumulated into the corresponding short-range force between the limited atoms between the correction atoms; The second correction atom list and the corresponding original additional information for calculating the original repulsion force between the correction atoms are obtained from the second correction task file by the global control unit; The second correction information in the second correction atom list and the corresponding original additional information are sent to the corresponding second calculation unit by the global control unit according to the data parameter information of each limited atom; The second correction distance is calculated in parallel by each second calculation unit according to the received second correction information; The second correction force is calculated in parallel by each second calculation unit according to the second correction distance and the original additional information, and the second correction force is accumulated into the corresponding short-range force between the limited atoms between the correction atoms.

6. The method of claim 5, wherein, After the second correction force is accumulated into the corresponding short-range force between the limited atoms between the correction atoms, the method further comprises: The real-space electrostatic force and the inverse-space electrostatic force between the limited atoms are calculated by each target calculation unit according to the electrostatic parameters of the limited atoms and the distance between the limited atoms; The real-space electrostatic force and the inverse-space electrostatic force calculated are accumulated by each target calculation unit to obtain the long-range electrostatic force between the limited atoms.

7. The method according to any one of claims 1 to 6, characterized in that, The data parameter information includes: atom information, group information, near-neighbor atom index, van der Waals force parameter, to-be-sent information of target atoms in each storage unit, and electrostatic parameters.

8. A three-dimensional mesh structure hardware, characterized by, The three-dimensional grid structure hardware includes: The first dimension x the second dimension x the third dimension grid points; each grid point includes a calculation unit and an associated storage unit; each calculation unit or associated storage unit is communicatively connected with adjacent calculation units or associated storage units; the calculation units or associated storage units located at the edges of the three-dimensional grid structure hardware are communicatively connected with the calculation units or associated storage units located at the opposite edges of the three-dimensional grid structure hardware; and a global control unit communicatively connected with each calculation unit; the global control unit is used to issue parameters to the calculation units and coordinate calculation tasks; The three-dimensional grid structure hardware includes: Each associated storage unit is in communication connection with a corresponding computing unit; wherein the associated storage unit stores the data parameter information of any one of claims 1-7, and a computer program executable by the at least one computing unit, the computer program being executed by the at least one computing unit to cause the at least one computing unit to perform the molecular anti-aggregation simulation method of any one of claims 1-7.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions for causing the processor to implement the molecular anti-aggregation simulation method of any one of claims 1-7 when executed.

10. A computer program product, characterised in that, The computer program product comprises a computer program which, when executed by the processor, implements the molecular anti-aggregation simulation method according to any one of claims 1-7.

Images