A universal cleaning system for molecular dynamics simulation data

Abstract

The present invention provides a general cleaning system for molecular dynamics simulation data, which relates to the field of data processing technology. The system comprises a data acquisition module for acquiring molecular dynamics simulation data; a parameter acquisition module for acquiring cleaning parameter information; a data cleaning module for cleaning molecular dynamics simulation data based on the cleaning parameter information to obtain cleaning results; and a result display module for displaying the cleaning results and related information. The present invention solves the problems of complex result analysis and high difficulty in data output of molecular dynamics simulation data.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a general cleaning system for molecular dynamics simulation data. Background Art

[0002] Molecular dynamics simulation is a key technology for exploring the behavior of molecular systems at the atomic level. It enables us to have a deeper understanding of molecular structure, dynamics and reaction mechanisms by simulating the motion trajectories of atoms and molecules. This simulation generates a large amount of data. Due to the diversity of data formats and the complexity of results, large-scale batch scientific data processing is required to ensure that valuable information can be extracted from it. As the reactions that need to be explored become more and more complex, the models established become larger and larger, and the scale and accuracy of simulations continue to grow, the amount of data generated has increased significantly, making data processing more important. By effectively processing and analyzing these data, researchers can convert complex simulation results into multi-scale data for analysis, which is of great significance for exploring reaction mechanisms and understanding material properties. However, the current lack of a universally applicable cleaning method for various molecular dynamics simulation data has greatly hindered the development of molecular dynamics scientific research. Summary of the invention

[0003] In view of the above-mentioned deficiencies in the prior art, the present invention provides a universal cleaning system for molecular dynamics simulation data, which solves the problems of complex result analysis and high difficulty in data output of molecular dynamics simulation data.

[0004] In order to achieve the above-mentioned invention object, the technical solution adopted by the present invention is: a general cleaning system for molecular dynamics simulation data, comprising:

[0005] A data acquisition module, used to acquire molecular dynamics simulation data;

[0006] A parameter acquisition module is used to obtain cleaning parameter information;

[0007] A data cleaning module, including a log data cleaning unit, a trajectory data cleaning unit, a key-level data cleaning unit, a species data cleaning unit and a block data cleaning unit; used to clean the molecular dynamics simulation data based on the cleaning parameter information to obtain a cleaning result; wherein the species data cleaning unit includes a species data cleaning subunit, a centroid data cleaning subunit and a transition state data cleaning subunit;

[0008] The result display module is used to display the cleaning results and related information.

[0009] Furthermore, the parameter acquisition module includes:

[0010] A trajectory time acquisition unit, used to acquire simulation step information; based on the simulation step information and trajectory interval information, obtain trajectory time information;

[0011] A track interval adjustment unit, used to obtain track interval information; based on the track interval information, obtain corresponding track time point information through calculation;

[0012] A duration adjustment unit, used to adjust the trajectory time, ignore the time information of the preprocessing step, and obtain ignored duration information;

[0013] The size adjustment unit is used to convert the supercell size to obtain the single cell size information; wherein the trajectory time point information, trajectory time information, ignored duration information and single cell size information belong to cleaning parameter information.

[0014] Furthermore, the data cleaning module includes:

[0015] A log data cleaning unit, used for calculating the unit cell parameters of the log data, adjusting the unit cell size, and obtaining the log data cleaning result;

[0016] A trajectory data cleaning unit, used for adjusting the key length of the trajectory data based on the trajectory data to obtain a trajectory data cleaning result;

[0017] A key-level data cleaning unit, used to clean the key-level data based on a preset key-level threshold and a key-length threshold, and obtain a key-level data cleaning result;

[0018] A species data cleaning unit is used to calculate the molecular weight of the species data based on the species data, sort the elements, and obtain the species data cleaning result;

[0019] The block data cleaning unit is used to segment the block data to obtain the block data cleaning results; wherein the log data, bond level data, trajectory data, species data, center of mass data, transition state data and block data belong to the molecular dynamics simulation data.

[0020] Furthermore, the log data cleaning unit is used to:

[0021] Based on the unit cell information, simulation step information and track interval information, the unit cell size data is calculated to obtain the log data cleaning result; among them, the unit cell information, simulation step information and track interval information belong to the cleaning parameter information, and the log data cleaning result includes the single unit cell adjustment parameter, single unit cell adjustment volume and adjustment time:

[0022] cella=cella0 / supercella;

[0023] cellB = cellB0 / supercellB;

[0024] cellc=cellc0 / supercellc;

[0025] volume=volume0 / (supercella×supercellb×supercellc);

[0026] time=frame×timestep×0.001;

[0027] Among them, cella represents the adjustment parameter of the single cell in the a direction, cella0 represents the log data of the cell in the a direction, supercella represents the expansion multiple of the cell in the a direction, cellb represents the adjustment parameter of the single cell in the b direction, cellb0 represents the log data of the cell in the b direction, supercellb represents the expansion multiple of the cell in the b direction, cellc represents the adjustment parameter of the single cell in the c direction, cellc0 represents the log data of the cell in the c direction, supercellc represents the expansion multiple of the cell in the c direction, volume represents the adjusted volume of the single cell, volume0 represents the log data of the volume of the single cell, time represents the adjustment time, frame represents the number of trajectories, and timestep represents the step size.

[0028] Furthermore, the trajectory data cleaning unit is used to:

[0029] Converting the trajectory data into one-dimensional data; wherein the trajectory data includes the simulation box length, initial origin coordinates, tilt value and bonding atom coordinates of each trajectory;

[0030] Using the trajectory data, the key length of the trajectory data is adjusted to obtain the adjusted key length:

[0031]

[0032] (x1-x2) 2 =((x1-x2)-(y1-y2)*tan(xy)-(z1-z2)*tan(xz)) 2 ;

[0033] (y1-y2) 2 =((y1-y2)-(z1-z2)*tan(yz)) 2 ;

[0034] Wherein, l represents the adjusted bond length, x1 represents the abscissa of the first bonding atom, x2 represents the abscissa of the second bonding atom, y1 represents the ordinate of the first bonding atom, y2 represents the ordinate of the second bonding atom, z1 represents the height coordinate of the first bonding atom, z2 represents the height coordinate of the second bonding atom, xy represents the first tilt value, and yz represents the second tilt value;

[0035] Based on the adjusted bond length, the unit cell side length is periodically judged, and the data exceeding the boundary is adjusted to obtain the trajectory data cleaning result.

[0036] Furthermore, the key-level data cleaning unit is used to:

[0037] Construct bond-level data templates using atom types, numbering dictionaries, atom species, atomic weights, and element names;

[0038] Using the list data fusion function, adding a chemical bond column to the bond level data template;

[0039] By adding a threshold function to the key-level data template, the key-level data is input into the key-level data template for cleaning, thereby obtaining a key-level data cleaning result.

[0040] Furthermore, the species data cleaning subunit is used to:

[0041] Based on the element sorting rules, the species data are sorted to obtain the standardized molecular formula;

[0042] Use the molecular weight calculation function to sort the standardized molecular formulas and obtain the species data cleaning results:

[0043] Weight = α × M A +β×M B +γ×M C ;

[0044] Among them, Weight represents molecular weight, M A Represents the atomic weight of element A, M B Indicates the atomic weight of element B, M C represents the atomic weight of element C, α represents the amount of element A, β represents the amount of element B, and γ represents the amount of element C.

[0045] Furthermore, the centroid data cleaning subunit is used for:

[0046] By extracting the trajectory number, species number, atomic number, molecular formula, average charge and center of mass coordinates in the center of mass data, a standardized molecular formula was constructed to obtain the center of mass data cleaning results.

[0047] Furthermore, the transition state data cleaning subunit is used for:

[0048] Use regular expressions to extract the minimized energy information of transition state data, calculate the energy difference before and after relaxation, and obtain a list of relaxation numbers and energy differences;

[0049] The relaxation numbers and energy difference lists are normalized to obtain transition state data cleaning results.

[0050] Furthermore, the block data cleaning unit is used to:

[0051] Divide the block data into small units and obtain the amount of each small unit;

[0052] Perform keyword matching on the small unit content to extract first block keyword data and second block keyword data; wherein the first block keyword data includes the track number and the block number, and the second block keyword data includes the block number, block parameter, block atom number and target parameter;

[0053] By using data fusion, the first block keyword data and the second block keyword data are fused to obtain the block data cleaning result.

[0054] The beneficial effects of the present invention are: based on the data acquisition module, parameter acquisition module, data cleaning module and result display module of the molecular dynamics simulation data cleaning algorithm, the molecular dynamics simulation data can be efficiently cleaned. In this way: (1) a variety of data can be processed uniformly, greatly improving the analysis efficiency; (2) a variety of new data such as single cell parameters, chemical bond lengths and types, species molecular weights, species centroids, time parameters, etc. can be obtained through data cleaning; (3) parameters such as supercell size, bond order and bond length thresholds, species element sorting, model periodicity, etc. can be customized to achieve single crystal data acquisition, customized bond formation or bond breaking, bond length correction, centroid correction, etc. Multiple data optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] This specification will be further described in the form of exemplary embodiments, which will be described in detail by the accompanying drawings. These embodiments are not restrictive, and in these embodiments, the same number represents the same structure, wherein:

[0056] Figure 1 It is a module schematic diagram of a universal cleaning system for molecular dynamics simulation data according to some embodiments of this specification. DETAILED DESCRIPTION

[0057] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0058] Example

[0059] Figure 1 It is a module schematic diagram of a universal cleaning system for molecular dynamics simulation data according to some embodiments of this specification.

[0060] In some embodiments, the universal molecular dynamics simulation data cleaning system may include a data acquisition module, a parameter acquisition module, a data cleaning module and a result display module.

[0061] The data acquisition module is used to acquire molecular dynamics simulation data.

[0062] Molecular dynamics simulation data refers to all types of data generated by the molecular dynamics model during the simulation process. For example, it can include log data, bond level data, trajectory data, species data, center of mass data, transition state data, and block data.

[0063] Log data is the log file data of the molecular dynamics model during the simulation process.

[0064] Bonds data is data reflecting the number and type of atoms.

[0065] Dump data is data that reflects the running track of each element during the simulation process. For example, the dump data may include the simulation box length, initial origin coordinates, tilt values, and bonding atom coordinates of each track.

[0066] Species data refers to data reflecting different molecular species.

[0067] Center of mass (POS) data is data related to the center of mass position of different species of molecules. For example, center of mass data can include track number, species number, number of atoms, molecular formula, average charge and center of mass coordinates.

[0068] Transition state (NEB) data is data reflecting energy changes.

[0069] Block data is data reflecting the parameters of small unit atomic groups.

[0070] In some embodiments, the data acquisition module can use molecular dynamics simulation software, such as LAMMPS, to simulate the molecular dynamics model and obtain molecular dynamics simulation data.

[0071] The parameter acquisition module is used to obtain cleaning parameter information.

[0072] The cleaning parameter information is preset parameter information before data cleaning. For example, the cleaning parameter information may include cleaning step information, cleaning track interval information, cleaning ignore time information, cleaning cell size information, and the like.

[0073] In some embodiments, the cleaning parameter information may be determined by simulating preset parameters through expert input.

[0074] In some embodiments, the parameter acquisition module can also be used to set the read data file and encoding format.

[0075] In some embodiments, the parameter acquisition module may use preset coding information to match the current data file and correctly read the data file.

[0076] In some embodiments, the parameter acquisition module may include a step length adjustment unit, a track interval adjustment unit, a duration adjustment unit, and a size adjustment unit.

[0077] In some embodiments, the parameter acquisition module may further include an encoding adjustment unit for matching the data file encoding format and correctly extracting data information.

[0078] The trajectory time acquisition unit is used to acquire simulation step information; based on the simulation step information and trajectory interval information, the trajectory time information is obtained.

[0079] The simulation step size information is the size of the "step" from one iteration step to the next.

[0080] The track time information is the time information corresponding to each track.

[0081] In some embodiments, the trajectory time acquisition unit can find the best balance point, trajectory time point information, by testing different simulation step sizes.

[0082] In this way, the accuracy of the simulation can be guaranteed while maintaining acceptable computational efficiency.

[0083] The track interval adjustment unit is used to obtain track interval information; based on the track interval information, the corresponding track time point information is obtained through calculation.

[0084] The trajectory interval information is the frequency of output trajectory data set during molecular simulation calculations.

[0085] The trajectory time point information is the time point information corresponding to the trajectory output.

[0086] In this way, the accuracy of the trajectory interval information can be better guaranteed, and the time point information corresponding to the trajectory can be correctly output, which is helpful for the analysis and drawing of the cleaned data.

[0087] The duration adjustment unit is used to adjust the trajectory time, ignore the time information of the preprocessing step, and obtain the ignored duration information.

[0088] Ignoring duration information means that the duration of the trajectory consumed by the preprocessing step is not considered during simulation.

[0089] In some embodiments, the duration adjustment unit may adjust the output trajectory time to ignore the time occupied by steps such as static relaxation and room temperature relaxation during the molecular simulation process, thereby obtaining the ignored duration.

[0090] In some embodiments, the duration adjustment unit can make the simulated time output more accurate.

[0091] The size adjustment unit is used to convert the supercell size to obtain the single cell size information.

[0092] The single cell size information is the cell size converted based on the supercell size and the cell magnification.

[0093] In some embodiments, the size adjustment unit can obtain the size of a single unit cell, which is convenient for experimenters to compare with single unit cell data in a crystal library or literature.

[0094] The data cleaning module is used to clean the molecular dynamics simulation data based on the cleaning coding information and the cleaning parameter information to obtain a cleaning result.

[0095] The cleaning results are molecular dynamics simulation data that are neater, more intuitive, and reflect more details. For example, the cleaning results can include log data cleaning results, bond level data cleaning results, trajectory data cleaning results, species data cleaning results, centroid data cleaning results, transition state data cleaning results, and block data cleaning results.

[0096] In some embodiments, the data cleaning module may include a log data cleaning unit, a key-level data cleaning unit, a trajectory data cleaning unit, a species data cleaning unit, and a block data cleaning unit.

[0097] The log data cleaning unit is used to calculate the unit cell parameters of the log data, adjust the unit cell size, and obtain the log data cleaning result.

[0098] In some embodiments, the log data cleaning unit may clean the log data and calculate and obtain the single cell parameters based on the user-defined super cell size parameters.

[0099] The log data cleaning result is a result reflecting parameters such as unit cell volume and size after cleaning. For example, the log data cleaning result may include unit cell adjustment parameters, unit cell adjustment volume and adjustment time.

[0100] In some embodiments, the log data cleaning unit is used to calculate the unit cell size data based on the unit cell information, the simulation step information and the track interval information to obtain the log data cleaning result.

[0101] In some embodiments, the log data cleaning unit can convert the log data into a CSV file, fuzzy match the file according to the file name, return the initial file path, traverse all lines of the file, extract the starting and ending lines of the target data, and convert it into Pandas list data. When there are multiple groups of data, the target data group is numbered and extracted according to the extraction order, and whether the data is complete is determined according to the log format, and the judgment result is fed back in the log data cleaning result.

[0102] The Pandas list organizes the extracted log data into a structured data table to facilitate subsequent data analysis and processing.

[0103] In some embodiments, the log data cleaning unit can perform preliminary processing on the log data according to preset cell information, simulation step information and trajectory interval information, such as calculating time, single cell parameters, single cell adjustment volume, etc. according to the trajectory.

[0104] In some embodiments, the expression of the log data cleaning result may be:

[0105] cella=cella0 / supercella;

[0106] cellB = cellB0 / supercellB;

[0107] cellc=cellc0 / supercellc;

[0108] volume=volume0 / (supercella×supercellb×supercellc);

[0109] time=frame×timestep×0.001;

[0110] Among them, cella represents the adjustment parameter of the single cell in the a direction, cella0 represents the log data of the cell in the a direction, supercella represents the expansion multiple of the cell in the a direction, cellb represents the adjustment parameter of the single cell in the b direction, cellb0 represents the log data of the cell in the b direction, supercellb represents the expansion multiple of the cell in the b direction, cellc represents the adjustment parameter of the single cell in the c direction, cellc0 represents the log data of the cell in the c direction, supercellc represents the expansion multiple of the cell in the c direction, volume represents the adjusted volume of the single cell, volume0 represents the log data of the volume of the single cell, time represents the adjustment time, frame represents the number of trajectories, and timestep represents the step size.

[0111] The key-level data cleaning unit is used to clean the key-level data based on a preset key-level threshold and a key-length threshold to obtain a key-level data cleaning result.

[0112] In some embodiments, the bond-level data cleaning unit can be used to clean the bond-level data, and combine the trajectory data to obtain new data on chemical bond types and bond lengths, perform bond length correction based on custom periodicity, and filter the bond-level data based on custom bond-level thresholds and bond length thresholds.

[0113] The bond-level data cleaning result is the result of normalizing the data format after cleaning, adding chemical bond columns, bond-level threshold columns, bond-length threshold columns, and bond-level data.

[0114] In some embodiments, the bond-level data cleaning unit can be used to: construct a bond-level data template using atom type, numbering dictionary, atom type, atomic weight and element name; add a chemical bond column to the bond-level data template using a list data fusion function; and input the bond-level data into the bond-level data template for cleaning by adding a bond-level threshold and a bond length threshold function to the bond-level data template to obtain a bond-level data cleaning result.

[0115] The bond-level data template is a template for organizing and classifying the elements of the bond-level data. For example, the bond-level data template may include types such as atom type, number dictionary, atom type, atomic weight, and element name.

[0116] In some embodiments, the key-level data cleaning unit can use cleaning encoding information (such as UTF-8) and regular expressions to extract the atom number and type of the key-level data, define chemical bond columns and time columns, and store the key-level data into a Pandas list.

[0117] In some embodiments, the bond-level data cleaning unit can use the first frame trajectory of the file to establish an atom type and number dictionary; use the molecular dynamics model file to establish a list containing atom type, atomic weight, and element name information; and add a chemical bond column through the Pandas list data fusion function of the bond-level data, including two element names of the chemical bond (such as CN, NN, etc.).

[0118] In some embodiments, the bond-level data cleaning unit can add a threshold function, define the bond-level threshold and bond-length threshold of all chemical bonds, add a bond-level threshold column to the cleaned bond-level data file through data fusion, and add time column information based on the trajectory column to obtain the bond-level data cleaning result.

[0119] In some embodiments, the paired atom number and bond order threshold for adding a single atom row may be 0 and 6, respectively.

[0120] The trajectory data cleaning unit is used to adjust the key length of the trajectory data based on the trajectory data to obtain a trajectory data cleaning result.

[0121] In some embodiments, the trajectory data cleaning unit may be used to clean the trajectory data and obtain newly added data of element types.

[0122] The trajectory data cleaning result is the result after cleaning that reflects the relationship between the coordinates of bonding atoms and time.

[0123] In some embodiments, the trajectory data cleaning unit can be used to: convert the trajectory data into one-dimensional data; use the trajectory data to adjust the bond length of the trajectory data to obtain the bond length of the bonding atom; based on the bond length of the bonding atom, perform periodic judgment on the unit cell side length, adjust the data that exceeds the boundary, and obtain the trajectory data cleaning result.

[0124] The bond length of the bonding atom is the original bond length of the bonding atom in the trajectory file.

[0125] In some embodiments, the expression for the bond length of bonding atoms may be:

[0126]

[0127] (x1-x2) 2 =((x1-x2)-(y1-y2)*tan(xy)-(z1-z2)*tan(xz)) 2 ;

[0128] (y1-y2) 2 =((y1-y2)-(z1-z2)*tan(yz)) 2 ;

[0129] Among them, l represents the bond length of the bonding atoms, x1 represents the horizontal coordinate of the first bonding atom, x2 represents the horizontal coordinate of the second bonding atom, y1 represents the vertical coordinate of the first bonding atom, y2 represents the vertical coordinate of the second bonding atom, z1 represents the height coordinate of the first bonding atom, z2 represents the height coordinate of the second bonding atom, xy represents the first tilt value, and yz represents the second tilt value.

[0130] In some embodiments, the trajectory data cleaning unit may split the trajectory data into multiple rows, traverse all frames to read trajectory numbers, simulation box information and atom information, add trajectory columns and time columns; and use the molecular dynamics model file to add element name columns.

[0131] In some embodiments, the trajectory data cleaning unit can extract the simulation box length, initial origin coordinates and tilt value of each trajectory in the trajectory data, and convert it into one-dimensional data (cell data); based on the one-dimensional data, the coordinates of the bonding atoms are extracted, and the corresponding bonding atom bond lengths are calculated.

[0132] In some embodiments, the trajectory data cleaning unit can perform periodic processing on the ultra-long chemical bonds to obtain trajectory data cleaning results. For example, the trajectory data cleaning unit can compare the distances of the unit cell bonding atom pairs in three directions (x, y, z) and the corresponding bonding atom bond lengths. When the bonding atoms are periodic in the corresponding directions and the distance difference between the two bonding atoms exceeds half of the corresponding bond length, the bonding atom distance parameter is adjusted to l-|x1-x2|.

[0133] The species data cleaning unit is used to calculate the molecular weight of the species data based on the species data, sort the elements, and obtain the species data cleaning result.

[0134] In some embodiments, the species data cleaning unit can be used to clean the species data, calculate the molecular weight of the species, sort the species molecular formula based on a custom element order, and perform centroid correction based on a custom periodicity.

[0135] The species data cleaning result is the result reflecting the information related to the number of species after cleaning. For example, the species data cleaning result may include the number of species, the number of molecules, the centroid data, and the transition state data.

[0136] In some embodiments, the species data cleaning unit may include a species data cleaning subunit, a centroid data cleaning subunit, and a transition state data cleaning subunit.

[0137] In some embodiments, the species data cleaning subunit can be used to: sort the species data based on the element sorting rule to obtain the standardized molecular formula; use the molecular weight calculation function to sort the standardized molecular formula to obtain the species data cleaning result.

[0138] The standardized molecular formula is the molecular formula after the molecular formulas are sorted according to preset rules.

[0139] In some embodiments, the species data cleaning subunit may define an element sorting rule to sort the elements in the molecular formula to obtain a standardized molecular formula. For example, the element sorting rule may be:

[0140] rule = {B:1,A:2,C:3};

[0141] Among them, rule represents the element sorting rule, A, B, and C represent the corresponding elements respectively, and the numbers 1, 2, and 3 represent the corresponding priorities.

[0142] In some embodiments, the species data cleaning subunit can establish a molecular weight calculation function, calculate the molecular weight corresponding to each molecular formula, and store the result in a molecular weight column. At the same time, the time column information is added based on the trajectory column to obtain the species data cleaning result.

[0143] In some embodiments, the expression of molecular weight in the species data cleaning result can be:

[0144] Weight = α × M A +β×M B +γ×M C ;

[0145] Among them, Weight represents molecular weight, M A Represents the atomic weight of element A, M B Indicates the atomic weight of element B, M C represents the atomic weight of element C, α represents the amount of element A, β represents the amount of element B, and γ represents the amount of element C.

[0146] In some embodiments, researchers can use custom functions to select and calculate corresponding data indicators based on specific research objectives and required functions. For example, researchers can define their own analysis functions, such as fitting equations for pressure-volume relationships, second-order or third-order Birch-Murnaghan solid-state equations, etc.

[0147] The centroid data cleaning subunit is used to perform standardization processing on the centroid data to obtain the centroid data cleaning result.

[0148] The centroid data cleaning result reflects the data related to the centroid coordinates after cleaning.

[0149] In some embodiments, the centroid data cleaning subunit can be used to extract the trajectory number, species number, atomic number, molecular formula, average charge and centroid coordinates in the centroid data, construct a standardized molecular formula, and obtain the centroid data cleaning result.

[0150] In some embodiments, the centroid data cleaning subunit can convert the centroid data into a multi-line list, traverse the list to match keywords, extract the number of trajectories, as well as the species number, number of atoms, molecular formula, average charge and center of mass coordinates, create a Pandas list, add molecular weight and time columns, and standardize the molecular formula to obtain the centroid data cleaning result.

[0151] The transition state data cleaning subunit is used to calculate the energy difference of the transition state data and obtain the transition state data cleaning result.

[0152] The transition state data cleaning results reflect the energy difference related results after cleaning.

[0153] In some embodiments, the transition state data cleaning subunit can be used to: extract the minimized energy information of the transition state data using a regular expression, calculate the energy difference before and after relaxation, and obtain a relaxation number and an energy difference list; normalize the relaxation number and the energy difference list to obtain a transition state data cleaning result.

[0154] The relaxation number is the sequence number of relaxation times.

[0155] The energy difference list is a list of energy differences reflecting each relaxation.

[0156] In some embodiments, the transition state data cleaning subunit can match the data file name under the target path through a regular expression, traverse all files to extract the minimized energy information through keyword matching, calculate the energy difference before and after relaxation, generate a relaxation number and energy difference list, and calculate the normalized value to obtain the transition state data cleaning result.

[0157] The block data cleaning unit is used to segment the block data to obtain the block data cleaning result.

[0158] The block data cleaning result reflects the information of each small unit atomic group after cleaning.

[0159] In some embodiments, the block data cleaning unit can be used to: divide the block data into small units and obtain the quantity of each small unit; perform keyword matching on the content of the small unit to extract the first block keyword data and the second block keyword data; use data fusion to fuse the first block keyword data and the second block keyword data to obtain the block data cleaning result.

[0160] The quantity of a small unit is data reflecting the temperature and speed information of the small unit.

[0161] In some embodiments, the block data cleaning unit may obtain the amount of the small unit by counting the parameter data of each atom in each small unit after division.

[0162] The first block keyword data is data reflecting the overall operation of the small unit. For example, the first block keyword data may include the number of trajectories and the number of blocks.

[0163] The second block keyword data is data reflecting detailed parameters of each small unit. For example, the second block keyword data may include a block number, a block parameter, a block atom number, and a target parameter.

[0164] In some embodiments, the block data cleaning unit can match the block data file under the target path through a regular expression and split the file content by line. The number of trajectories and the number of blocks are obtained by matching keywords respectively, and the first block keyword data is extracted by traversing the file; the second block keyword data is obtained by matching the block number, block parameter, block atom number, and target parameter through keywords.

[0165] In some embodiments, the block data cleaning unit may obtain the block data cleaning result by fusing the first block keyword data and the second block keyword data, storing the data into a Pandas list, and adding a time column.

[0166] The result display module is used to display the cleaning results and related information.

[0167] The relevant information is the process data related to the cleaning results. For example, the relevant information may include analysis progress, error information, result data format, etc.

[0168] In some embodiments, the result display module can display the obtained cleaning results and related information as output information in the display column.

[0169] In some embodiments of the present specification, the molecular dynamics simulation data universal cleaning system completes the cleaning of molecular dynamics simulation data by using a data acquisition module, a parameter acquisition module, a data cleaning module, and a result display module. In this way, (1) multiple data files can be processed uniformly, greatly improving the analysis efficiency; (2) users can comprehensively analyze the changes in unit cell parameters, density, chemical reactions, chemical bonds, and species types within the same time period based on the energy change results in the thermodynamics file, simplify the complexity of result analysis and reduce the difficulty of data output, thereby obtaining a more comprehensive analysis conclusion; (3) add custom function functions to increase the scope of application of the molecular dynamics simulation data universal cleaning system.

Claims

1. A general cleaning system for molecular dynamics simulation data, characterized in that: include: A data acquisition module, used to acquire molecular dynamics simulation data; A parameter acquisition module is used to obtain cleaning parameter information; A data cleaning module, including a log data cleaning unit, a trajectory data cleaning unit, a key-level data cleaning unit, a species data cleaning unit and a block data cleaning unit; used to clean the molecular dynamics simulation data based on the cleaning parameter information to obtain a cleaning result; wherein the species data cleaning unit includes a species data cleaning subunit, a centroid data cleaning subunit and a transition state data cleaning subunit; wherein the data cleaning module includes: A log data cleaning unit, used for calculating the unit cell parameters of the log data, adjusting the unit cell size, and obtaining the log data cleaning result; A trajectory data cleaning unit, used for adjusting the key length of the trajectory data based on the trajectory data to obtain a trajectory data cleaning result; A key-level data cleaning unit, used to clean the key-level data based on a preset key-level threshold and a key-length threshold, and obtain a key-level data cleaning result; A species data cleaning unit is used to calculate the molecular weight of the species data based on the species data, sort the elements, and obtain the species data cleaning result; A block data cleaning unit is used to segment the block data to obtain a block data cleaning result; wherein the log data, bond level data, trajectory data, species data, center of mass data, transition state data and block data belong to the molecular dynamics simulation data; The result display module is used to display the cleaning results and related information.

2. The molecular dynamics simulation data universal cleaning system according to claim 1, characterized in that: The parameter acquisition module includes: A trajectory time acquisition unit, used to acquire simulation step information; based on the simulation step information and trajectory interval information, obtain trajectory time information; A track interval adjustment unit, used to obtain track interval information; based on the track interval information, obtain corresponding track time point information through calculation; A duration adjustment unit, used to adjust the trajectory time, ignore the time information of the preprocessing step, and obtain ignored duration information; The size adjustment unit is used to convert the supercell size to obtain the single cell size information; wherein the trajectory time point information, trajectory time information, ignored duration information and single cell size information belong to cleaning parameter information.

3. The molecular dynamics simulation data universal cleaning system according to claim 1, characterized in that: The log data cleaning unit is used for: Based on the unit cell information, simulation step information and track interval information, the single unit cell size data is calculated to obtain the log data cleaning result; among them, the unit cell information, simulation step information and track interval information belong to the cleaning parameter information, and the log data cleaning result includes the single unit cell adjustment parameter, single unit cell adjustment volume and adjustment time: cella=cella0 / supercella; cellB = cellB0 / supercellB; cellc=cellc0 / supercellc; volume=volume0 / (supercella×supercellb×supercellc); time=frame×timestep×0.001; Among them, cella represents the adjustment parameter of the single cell in the a direction, cella0 represents the log data of the cell in the a direction, supercella represents the expansion multiple of the cell in the a direction, cellb represents the adjustment parameter of the single cell in the b direction, cellb0 represents the log data of the cell in the b direction, supercellb represents the expansion multiple of the cell in the b direction, cellc represents the adjustment parameter of the single cell in the c direction, cellc0 represents the log data of the cell in the c direction, supercellc represents the expansion multiple of the cell in the c direction, volume represents the adjusted volume of the single cell, volume0 represents the log data of the volume of the single cell, time represents the adjustment time, frame represents the number of trajectories, and timestep represents the step size.

4. The molecular dynamics simulation data universal cleaning system according to claim 1, characterized in that: The trajectory data cleaning unit is used for: Converting the trajectory data into one-dimensional data; wherein the trajectory data includes the simulation box length, initial origin coordinates, tilt value and bonding atom coordinates of each trajectory; Using the trajectory data, the key length of the trajectory data is adjusted to obtain the adjusted key length: (x1-x2) 2 =((x1-x2)-(y1-y2)*tan(xy)-(z1-z2)*tan(xz)) 2 4 (y1-y2) 2 =((y1-y2)-(z1-z2)*tan(yz)) 2 ; Wherein, l represents the adjusted bond length, x1 represents the abscissa of the first bonding atom, x2 represents the abscissa of the second bonding atom, y1 represents the ordinate of the first bonding atom, y2 represents the ordinate of the second bonding atom, z1 represents the height coordinate of the first bonding atom, z2 represents the height coordinate of the second bonding atom, xy represents the first tilt value, and yz represents the second tilt value; Based on the adjusted bond length, the unit cell side length is periodically judged, and the data exceeding the boundary is adjusted to obtain the trajectory data cleaning result.

5. The molecular dynamics simulation data universal cleaning system according to claim 1, characterized in that: The key-level data cleaning unit is used for: Construct bond-level data templates using atom types, numbering dictionaries, atom species, atomic weights, and element names; Using the list data fusion function, adding a chemical bond column to the bond level data template; By adding a threshold function to the key-level data template, the key-level data is input into the key-level data template for cleaning, thereby obtaining a key-level data cleaning result.

6. The molecular dynamics simulation data universal cleaning system according to claim 1, characterized in that: The species data cleaning subunit is used for: Based on the element sorting rules, the species data are sorted to obtain the standardized molecular formula; Use the molecular weight calculation function to sort the standardized molecular formulas and obtain the species data cleaning results: Weight=α×M A +β×M B +γ×M C ; Among them, Weight represents molecular weight, M A Represents the atomic weight of element A, M B Indicates the atomic weight of element B, M C represents the atomic weight of element C, α represents the amount of element A, β represents the amount of element B, and γ represents the amount of element C.

7. The molecular dynamics simulation data universal cleaning system according to claim 1, characterized in that: The centroid data cleaning subunit is used for: By extracting the trajectory number, species number, atomic number, molecular formula, average charge and center of mass coordinates in the center of mass data, a standardized molecular formula was constructed to obtain the center of mass data cleaning results.

8. The molecular dynamics simulation data universal cleaning system according to claim 1, characterized in that: The transition state data cleaning subunit is used for: Use regular expressions to extract the minimized energy information of transition state data, calculate the energy difference before and after relaxation, and obtain a list of relaxation numbers and energy differences; The relaxation numbers and energy difference lists are normalized to obtain transition state data cleaning results.

9. The molecular dynamics simulation data universal cleaning system according to claim 1, characterized in that: The block data cleaning unit is used for: Divide the block data into small units and obtain the amount of each small unit; Perform keyword matching on the small unit content to extract first block keyword data and second block keyword data; wherein the first block keyword data includes the track number and the block number, and the second block keyword data includes the block number, block parameter, block atom number and target parameter; By using data fusion, the first block keyword data and the second block keyword data are fused to obtain the block data cleaning result.

Images