Program, method and apparatus for analyzing reaction path network

Abstract

In the analysis of a conventional reaction pathway network, a shortest-path-based analysis method is used, but in the analysis using a network including several hundred equilibrium structures (EQs), it is important to systematically consider not only a main reaction pathway but also a side reaction pathway. However, this operation takes a long time and it is not easy to define an important partial network. Further, an important region differs depending on a target system, and a method of performing visualization by user definition is required.SOLUTION: The present invention provides an analysis program or the like that interactively visualizes and displays a partial network of a region by extracting the region that is kinetically reached based on an initial structure, a reaction temperature, a reaction time, and a proportion of a competing side reaction input from a user in a graph network in which a reaction path network is coarse-grained with respect to a molecular structure.SELECTED DRAWING: Figure 14

Description

[Technical Field]

[0001] In quantum chemistry, this invention relates to analyzing reaction mechanisms by calculating equilibrium structures (EQ structures, or simply EQs) on a potential energy surface (PES) and the reaction paths corresponding to the elementary processes connecting two EQs. This invention relates to an analysis program, method, and device that displays a graph network representing a reaction path network in which countless reaction paths are connected, coarse-grained with respect to molecular structures, and then receives user inputs such as the initial structure, reaction temperature, reaction time, and the proportion of competing side reactions, thereby interactively visualizing and displaying a subgraph network (hereinafter referred to as a "subnetwork") for the region kinetically reached.

[0002] Kinetics is a theory that can be used in quantum chemical calculations to obtain pruning information on reaction yields and which chemical elementary processes are progressing and which EQs can be reached, depending on all reaction barriers on the network. [Background technology]

[0003] In recent years, chemical reactions have been analyzed based on reaction path networks, in which thousands of reaction paths connect a large number of EQs in a mesh-like fashion. The resulting reaction path network can be analyzed by visualizing it as a graph network, where EQs are represented as nodes and reaction paths are represented as edges.

[0004] Another analytical method is to use a coarse-grained graph network (hereinafter referred to as "coarse-grained network"), which clusters EQs with similar molecular structures into a single group, regards that group as a single EQ, and expresses the reaction pathways connecting the groups as edges. In this case, the nodes on the graph network are associated with the energy of the EQ, and the edges are associated with the electronic energy or free energy (hereinafter referred to as "energy") of the highest structure along the reaction pathway. [Prior art documents] [Non-patent literature]

[0005] [Non-Patent Document 1] Yosuke Sumiya, Yu Harabuchi, Yuuya Nagata, and Satoshi Maeda, JACS Au, 2022, 2, 5, 1181-1188URL https: / / pubs.acs.org / doi / 10.1021 / jacsau.2c00157 (retrieved August 14, 2024) [Non-patent document 2] Mikael Kuwahara, Yu Harabuchi, Satoshi Maeda, Jun Fujima and Keisuke Takahashi, Digital Discovery, Digital Discovery, 2023, 2, 1104-1111URL https: / / pubs.rsc.org / en / content / articlelanding / 2023 / dd / d3dd00026e (Retrieved August 14, 2024) [Non-patent document 3] GRRM: Global Reaction Route Mapping, Quantum Chemistry Research Institute URL https: / / iqce.jp / GRRM / index.shtml (searched on August 14, 2024) [Non-patent document 4] SCAN: Searching Chemical Actions and Networks URL: https: / / scan.sci.hokudai.ac.jp (Retrieved August 14, 2024) Summary of the Invention [Problem to be solved by the invention]

[0006] Analysis of reaction pathway networks is based on shortest paths, but when analyzing networks containing hundreds of EQs, it is important to systematically consider not only the main reaction pathway but also the pathways for side reactions. However, this process is time-consuming, and defining important sub-networks is not easy. Furthermore, important regions vary depending on the system under study, and a method for user-defined visualization is required.

[0007] Therefore, the present invention provides an analysis program, method, and device that, in a graph network that is a coarse-grained reaction path network for a molecular structure, extracts a region that is kinetically reached based on the initial structure, reaction temperature, reaction time, and proportion of competing side reactions input by the user, and interactively visualizes and displays a partial network of that region. [Means for solving the problem]

[0008] As one embodiment of the reaction pathway network analysis program according to the present invention, the analysis program comprises: A program for analyzing a reaction path network expressed as a visualized graph network, in which each of a plurality of equilibrium structures (EQs) on a potential energy surface (PES) is a node, and reaction paths connecting the EQs are edges connecting the nodes, the program comprising: A step of extracting a partial network from a coarse-grained network constructed by clustering nodes of a graph network representing the reaction path network based on the identity or similarity of EQ structures, by calculating EQ for a region kinetically reached from an initial EQ corresponding to the initial structure based on a given initial structure, reaction temperature, and reaction time. The present invention is characterized in that the following is executed.

[0009] In a preferred embodiment of the reaction pathway network analysis program according to the present invention, an edge connecting nodes of a plurality of nodes included in the partial network is an edge on the PES whose relative energy from the initial EQ0 is lower than the sum of a threshold ΔE of the reaction barrier and a threshold ΔF for a side reaction pathway.

[0010] In a preferred embodiment of the reaction pathway network analysis program according to the present invention, the reaction barrier threshold ΔE is determined based on the reaction temperature and the reaction time, The threshold ΔF for the side reaction pathway is determined based on the reaction temperature, the reaction time, and the ratio of a given competing side reaction.

[0011] The threshold value ΔE of the reaction barrier and the threshold value ΔF for the side reaction pathway can be directly input by the user.

[0012] In a preferred embodiment of the reaction pathway network analysis program according to the present invention, the coarse-grained network is constructed by clustering a plurality of EQs corresponding to nodes of a graph network representing the reaction pathway network into one group for each of the same or similar molecular structures, and then treating each of the plurality of clustered groups as a single EQ.

[0013] In a preferred embodiment of the reaction pathway network analysis program according to the present invention, the coarse-grained network includes, as nodes, a plurality of EQs corresponding to each of the plurality of groups after the clustering, and a plurality of reaction pathways as edges connecting the nodes, Each of the plurality of reaction paths is characterized by connecting the EQs of the plurality of EQs.

[0014] In a preferred embodiment of the reaction pathway network analysis program according to the present invention, the energy of the EQ of the coarse-grained network is set to the minimum energy among the energies of the multiple EQs of the reaction pathway network that are combined into the EQ, The energy of the reaction pathway connecting the EQs of the plurality of EQs included in the coarse-grained network is the smallest energy among the energies of the plurality of reaction pathways connecting the plurality of EQs of the reaction pathway network grouped together in the EQ at one end of the reaction pathway and the plurality of EQs of the reaction pathway network grouped together in the EQ at the other end.

[0015] A preferred embodiment of the reaction pathway network analysis program according to the present invention is characterized in that it further comprises, before the step of extracting the partial network, a step of constructing a coarse-grained network by clustering nodes of a graph network representing the reaction pathway network based on identity or similarity of EQ structure.

[0016] A preferred embodiment of the reaction pathway network analysis program according to the present invention is characterized by further comprising a step of displaying the partial network on a user interface (display screen) of the computer.

[0017] As one embodiment of the reaction pathway network analysis method according to the present invention, the reaction pathway network analysis method is a method for processing a graph network representing a reaction pathway network in which each of a plurality of equilibrium structures (EQs) on a potential energy surface (PES) is defined as a node, and reaction pathways connecting the EQs are defined as edges connecting the nodes, The method is characterized by including a step of extracting a partial network from a coarse-grained network constructed by clustering nodes of a graph network representing the reaction pathway network based on the identity or similarity of EQ structures, by calculating an EQ for a region kinetically reached from an initial EQ0 corresponding to the initial structure based on a given initial structure, reaction temperature, and reaction time.

[0018] A preferred embodiment of the reaction path network analysis method according to the present invention is the same as a preferred embodiment of the reaction path network analysis program according to the present invention. Also, a reaction path network analysis device according to the present invention is characterized in that it is configured to execute any of the embodiments of the reaction path network analysis method according to the present invention. [Effects of the Invention]

[0019] The analysis program, method, and apparatus according to the present invention construct a coarse-grained network by clustering nodes in a graph network representing a reaction path network based on the identity or similarity of their EQ structures, and extract a partial network for the region that will be reached kinetically based on the initial structure, reaction temperature, reaction time, and proportion of competing side reactions input by the user. This allows for interactive visualization and display, thereby efficiently analyzing a partial region containing several dozen edges necessary for chemical reaction analysis from a reaction path network containing thousands of edges. Kinetics is a theory that, in quantum chemical calculations, can obtain pruning information on reaction yields and which elementary processes are progressing and which EQs can be reached based on all reaction barriers on the network.

[0020] Furthermore, while conventional reaction path network analysis requires specifying a start point and a goal point to visualize the route within the network, the present invention allows for extraction of a partial network by inputting node information for only the reaction's starting point, thereby streamlining the process. Furthermore, since the present invention does not require specification of a goal point, it can also be used in cases where the goal product is unknown. Furthermore, a program according to one embodiment of the present invention receives user input such as an initial structure, reaction temperature, reaction time, and the proportion of competing side reactions, and automatically calculates the definition of the partial network, including the EQ to be reached and the edges between them, and visualizes them. [Brief explanation of the drawings]

[0021] [Figure 1] FIG. 1 is a diagram showing one embodiment of a system that implements a reaction pathway network analysis program or a reaction pathway network analysis method of the present invention. [Figure 2] FIG. 1 is a schematic diagram showing the relationship between the potential energy surface (PES) used in quantum chemical calculations and the reaction path network. [Figure 3] FIG. 1 is a diagram showing an example of a visualized reaction pathway network. [Figure 4] FIG. 1 is a diagram showing an example of data of nodes included in a reaction pathway network. [Figure 5] FIG. 10 is a diagram showing an example of edge data included in a reaction pathway network. [Figure 6] 1 is a flowchart showing the flow of processing for constructing a coarse-grained network from a reaction pathway network in the reaction pathway network analysis program or reaction pathway network analysis method of the present invention. [Figure 7] FIG. 1 is a diagram showing an example of a screen display of a plurality of EQs having the same or similar molecular structures in a reaction pathway network. [Figure 8] FIG. 1 is a diagram showing an example of EQs after clustering, which correspond to one group, in which a plurality of EQs having the same or similar molecular structures are clustered into one group. [Figure 9] FIG. 1 is a diagram showing an example of a visualized coarse-grained network and an example of a partial network extracted from the visualized coarse-grained network. [Figure 10] 10 is a flowchart showing the flow of processing for extracting a partial network from a coarse-grained network. [Figure 11] 10 is a flowchart showing the flow of processing for extracting a partial network from a coarse-grained network. [Figure 12] FIG. 1 is a diagram showing the threshold ΔE of the reaction barrier defined based on the transition state theory from the reaction temperature and reaction time. [Figure 13]FIG. 1 illustrates a pruning scheme based on reaction temperature, reaction time, and the proportion of competing side reactions. [Figure 14] FIG. 2 illustrates an example of a user interface according to an embodiment of the present invention. [Figure 15] FIG. 2 illustrates an example of a user interface according to an embodiment of the present invention. [Figure 16] FIG. 2 illustrates an example of a user interface according to an embodiment of the present invention. [Figure 17] FIG. 2 illustrates an example of a user interface according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0022] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will now be described with reference to the accompanying drawings. The individual embodiments of the present invention are not independent and can be appropriately combined with each other for implementation.

[0023] 1 shows one embodiment of a system that implements a reaction path network analysis program and a reaction path network analysis method of the present invention (hereinafter also referred to as "the program, etc. of the present invention"). The system that implements the program, etc. of the present invention includes a terminal device 100 and a server device 200. The terminal device 100 and the server device 200 are connected via a network N.

[0024] The terminal device 100 and the server device 200 have the hardware configuration of a general computer (information processing device), and include, for example, hardware resources such as a CPU (Central Processing Unit), memory consisting of ROM (Read Only Memory) and RAM (Random Access Memory), a bus, an input / output interface, an input unit, an output unit, a storage unit, and a communication unit.

[0025] The CPU executes various processes according to a program recorded in the memory or a program loaded from the storage unit to the memory. For example, the CPU can execute the program of the present invention to cause the computer to function as a reaction path network analyzer. It is also possible to implement at least some of the functions of the analyzer in hardware form using an application specific integrated circuit (ASIC) or the like.

[0026] The memory also stores data necessary for the CPU to execute various processes as appropriate. The CPU and memory are interconnected via a bus. An input / output interface is also connected to this bus. An input unit, an output unit, a storage unit, and a communication unit are connected to the input / output interface. The input unit is composed of a keyboard, various buttons, a touch panel, a microphone, etc., and inputs various information in response to instructions from users of the terminal device 100 and the server device 200. The output unit is composed of a display, a speaker, etc., and outputs various data such as text, images, videos, and audio. The storage unit is composed of semiconductor memory such as DRAM (Dynamic Random Access Memory) or a hard disk, and stores various data. The communication unit realizes communication with other devices.

[0027] The terminal device 100 can visualize and display the analysis results of the reaction path network on the screen (user interface) of a display device such as a display, which is an output unit. In the embodiment shown in Fig. 1, the server device 200 executes the analysis of the reaction path network, and the terminal device 100 can acquire the analysis results of the reaction path network from the server device 200 via the network N. The present invention is not limited to the embodiment shown in Fig. 1, and processing related to the analysis of the reaction path network can also be executed in the terminal device 100 or the server device 200 on a stand-alone basis, without using the network N.

[0028] A reaction pathway network analysis device according to one embodiment of the present invention includes at least one processor and a memory that stores a reaction pathway network analysis program that performs a process of constructing a coarse-grained network by clustering a reaction pathway network as shown in Fig. 6 (described later), and a process of extracting a partial network from the coarse-grained network as shown in Figs. 10 and 11 (described later). The at least one processor of the reaction pathway network analysis device executes the reaction pathway network analysis program, thereby executing each step of the reaction pathway network analysis method (each step of the process shown in Figs. 6 and 10 to 11).

[0029] Figure 2 shows the relationship between the potential energy surface (PES) used in quantum chemical calculations and the reaction path network. Figure 2(a) shows an example of a potential energy surface (PES) that shows the energy distribution of a certain chemical reaction. The PES shown in Figure 2(a) is an example of a two-degree-of-freedom system, and the energy value calculated from the three-dimensional structure of the molecule using the Schrödinger equation represents the altitude of that coordinate. In the PES, the point (EQ0, EQ1) where the energy minimum is reached corresponds to the equilibrium structure (EQ).

[0030] In the PES shown in Figure 2(a), the lines connecting EQ0 and EQ1 correspond to the reaction pathways (transition state structures: TS) corresponding to the elementary steps connecting the two EQs. The reaction pathways can be determined by calculating the intrinsic reaction coordinate (IRC) from the transition state (TS).

[0031] Figure 2(b) shows the change in the energy value of the reaction path (for example, the TS connecting EQ0 and EQ1) in the PES shown in Figure 2(a). In the example shown in Figure 2(b), there is a reaction path with an energy value of 100 between EQ0, which has an energy value of 30, and EQ1, which has an energy value of 0.

[0032] Figure 2(c) shows an example of two nodes of a reaction pathway network and an edge connecting the nodes. Using the example shown in Figure 2(b), the reaction pathway network is represented as a node with an EQ (EQ0) with an energy value of 30 and an EQ (EQ1) with an energy value of 0, and the reaction pathway connecting the EQs (between EQ0 and EQ1) is represented as an edge connecting the nodes.

[0033] An example of a visualized reaction path network is shown in Figure 3. As shown in Figure 3, the reaction path network represents multiple equilibrium structures (EQs) on a potential energy surface (PES) as nodes, and the reaction paths connecting the EQs as edges connecting the nodes, making it possible to represent a state in which a large number of EQs are connected in a mesh-like pattern by an enormous number of reaction paths, exceeding several thousand.

[0034] 4 and 5 show examples of data on nodes and edges included in a reaction path network. The data shown in Fig. 4 and 5 is output data from an automated chemical reaction path search program (GRRM (registered trademark) program) (Non-Patent Document 3) used in the field of quantum chemistry calculations. This program uses the Schrodinger equation

number

[0035] The data shown in Figure 4 is EQ list data, in which the three-dimensional coordinates of atoms included in the molecular structure of each of multiple EQs are recorded in the format of (atom, X coordinate, Y coordinate, Z coordinate), and the energy value of each EQ is also recorded. In the example shown in Figure 4, the EQ list data lists the three-dimensional coordinates of atoms included in the molecular structure of EQ0 after "# Geometry of EQ 0, ...", followed by a list of the three-dimensional coordinates of atoms included in the molecular structures of EQ1 and EQ2 as "# Geometry of EQ 1, ...", "# Geometry of EQ 2, ...". The list also records the energy values ​​of each of EQ0, EQ1, and EQ2 (shown in Figure 4 as the "Energy" item).

[0036] The data shown in Figure 5 is a list of reaction paths (transition state structures (TS)) (also displayed as "Path Top" (PT)). The three-dimensional coordinates of atoms included in each molecular structure of multiple reaction paths (TS) are recorded in the format of (atom, X coordinate, Y coordinate, Z coordinate), and the energy value of each TS is also recorded. In the example shown in Figure 5, the TS list data lists the three-dimensional coordinates of atoms included in the molecular structure of TS0 after "# Geometry of TS0, ...", followed by a list of the three-dimensional coordinates of atoms included in the molecular structures of TS1 and TS2, similarly listed as "# Geometry of TS1, ...", "# Geometry of TS2, ...". The list also records the energy values ​​of TS0, TS1, and TS2 (shown as the "Energy" item in Figure 5). Also, TS0 is a reaction path connecting EQ0 and EQ1 in the EQ list (Fig. 4), and information about this reaction path is recorded as "CONNECTION: 0-1" in the last line of the "# Geometry of TS 0, ..." item in Fig. 5. Similarly, TS1 is a reaction path connecting EQ2 and EQ3, and TS2 is a reaction path connecting EQ4 and EQ5.

[0037] In the reaction path network analysis program and reaction path network analysis method of the present invention, a reaction path network can be obtained based on data from an automated chemical reaction path search program (GRRM® program) (see FIGS. 4 and 5 ). The reaction path network is represented by a visualized graph network in which each of a plurality of equilibrium structures (EQs) on a potential energy surface (PES) is a node, and reaction paths connecting the EQs are edges connecting the nodes. In a system implemented with the program of the present invention shown in FIG. 1 , for example, a graph network representing the reaction path network is constructed from reaction path network data (EQ list, TS list) obtained by running the automated chemical reaction path search program on a server device 200. A terminal device 100 accesses the server device 200 via a network N, and visualizes the graph network constructed by the server device 200 in a user interface (e.g., a browser screen) displayed on the screen of a display, which is the output unit of the terminal device 100.

[0038] 6 is a flowchart showing the flow of processing for constructing a coarse-grained network from a reaction path network in the program of the present invention. In the processing for constructing a coarse-grained network, a graph is defined for all EQ structures by defining chemical bonds as edges and atoms as nodes. For example, the format of the graph network for the reaction path network of the present invention can be determined using the same format as the network data output by the GRRM (registered trademark) program.

[0039] 6, the program of the present invention generates a graph for all EQ structures based on the data format of the output data of the GRRM (registered trademark) program, with chemical bonds as edges and atoms as nodes (step S11). At this time, information on atomic species is assigned to the nodes.

[0040] Next, EQ structures with identical or similar graphs are clustered and considered as a single group (cluster). The program of the present invention clusters each EQ structure with identical or similar graphs to form a single group (step S12). In this case, the lowest energy within the group is taken as the group's energy. In other words, the energy of an EQ included in the coarse-grained network corresponds to the smallest energy among the energies of the multiple EQs in the reaction path network before clustering that are grouped into that EQ. In this way, the program of the present invention can cluster multiple nodes in a graph network representing a reaction path network into a single group based on the identity or similarity of their molecular structures. This single group is also referred to as the clustered EQ to distinguish it from the EQ in the reaction path network before coarse-graining.

[0041] Finally, the lowest-energy reaction path connecting each group (EQ after clustering) is determined as the reaction path between the groups (EQ after clustering) (step S13). In other words, the energy associated with a reaction path connecting EQs of multiple EQs included in the coarse-grained network is the smallest energy associated with multiple reaction paths connecting multiple EQs of the reaction path network grouped in the EQ at one end of the reaction path to multiple EQs of the reaction path network grouped in the EQ at the other end of the reaction path. This reaction path between EQs is also called a clustered reaction path (TS after clustering) to distinguish it from the reaction path (TS) of the reaction path network before coarse-graining. In this way, the coarse-grained network is constructed by clustering multiple EQs (see FIG. 7) for identical or similar molecular structures for each EQ corresponding to a node in the graph network representing the reaction path network, and then defining each of the multiple groups as a clustered EQ (see FIG. 8).

[0042] Figure 7 shows an example of a screen display of multiple EQs with the same or similar molecular structures in a reaction path network. In the example shown in Figure 7, the presence or absence of bonds is determined based on a completely geometric distance, such as 1.2 times the atomic covalent bond radius, and then the molecular structure can be displayed in two dimensions. Molecular structures with EQs that have the same bonding pattern are grouped together.

[0043] Figure 8 shows an example of a clustered EQ corresponding to a group in which multiple EQs having the same or similar molecular structures are clustered. Multiple EQs (such as "G0 EQ104," "G0 EQ220," "G0 EQ51," and "G0 EQ97" in Figure 7) each contain the same molecular structure, and as shown in Figure 8, these multiple EQs are grouped together to generate a clustered EQ ("G0 EQ104 -161.3" in the left column of Figure 8). The right column of Figure 8 shows multiple EQs (such as "EQ104," "EQ220," "EQ51," and "EQ97") before clustering that correspond to one group and are included in the clustered EQ.

[0044] By the process of constructing a coarse-grained network by clustering the reaction path network shown in Fig. 6, for example, the reaction path network shown in Fig. 3 is coarse-grained to generate the coarse-grained network shown in Fig. 9(a), and by the process of extracting a partial network from the coarse-grained network (see Figs. 10 and 11), which will be described later, the partial network shown in Fig. 9(b) is generated. Fig. 9(a) shows an example of a visualized coarse-grained network, and Fig. 9(b) shows an example of a partial network extracted from the visualized coarse-grained network.

[0045] Figures 10 and 11 are flowcharts showing the process flow for extracting a partial network from a coarse-grained network. Here, the groups of the grouped coarse-grained network are called EQs (nodes), and the lowest energy paths connecting the groups are called reaction paths (edges). The initial EQ input by the user is called EQ0. Also, the threshold of the reaction barrier defined based on the transition state theory from the reaction temperature and reaction time input by the user is called ΔE. The threshold for the side reaction path defined from the reaction temperature, reaction time, and the ratio of competing side reactions input by the user (reaction yield of the side reaction) is called ΔF. Furthermore, EQ i and EQ j EDGE is a reaction pathway that directly connects ij A list of EQs to be displayed is created by repeating the process of extracting sub-networks from the coarse-grained network as shown in Figures 10 and 11.

[0046] Figure 12 shows the threshold ΔE of the reaction barrier (PT) defined based on the transition state theory from a given reaction temperature and reaction time. The threshold ΔE of the reaction barrier as shown in Figure 12 can be determined from the following formula using the given reaction temperature and reaction time.

number

[0047] Figure 13 shows a mechanism for pruning based on the reaction temperature, reaction time, and the proportion of competing side reactions. In the example of Figure 13(a), when there are two reaction paths to the left and right of EQ0 (start), the reaction leading to EQ2, which has a low reaction barrier (barrier (energy): 100), is more likely to occur, but the reaction leading to EQ1, which has a high reaction barrier (barrier (energy): 130), also progresses slightly. The program of the present invention prunes the reaction path where this reaction is less likely to occur, but the degree of pruning can be adjusted based on the proportion of competing side reactions entered by the user.

[0048] A computer that executes the program of the present invention, or a reaction path network analysis device that executes the reaction path network analysis method of the present invention, can receive input from a user via a user interface such as those shown in Figures 14 to 17. Figures 14 to 17 show an example of a user interface (display screen) according to one embodiment of the present invention. In the example shown in Figure 14, the item "Temp" at the top of the display screen is [K] ", "T sec ", "Side % ", "EQ init " are items that accept various input values ​​such as reaction temperature, reaction time, proportion of competing reactions, and initial structure. In this example, the user specifies the reaction temperature as 300K, the reaction time as 3600 seconds, the proportion of competing reactions as 0.0001%, and the initial structure as the EQ after clustering with the identification number 59. Based on these specified input values, the barrier (energy) difference ("E" in Figure 14) is calculated. lim " is calculated to be 34.46, and it is possible to avoid pruning depending on the threshold ΔE of the reaction barrier and the threshold ΔF for the side reaction pathway.

[0049] Figure 13(b) shows the energy difference (E lim " is 0. In this case, the probability of taking the path leading to EQ1 and the probability of taking the path leading to EQ2 are each 50%, making the choice of the path the reaction takes equally probable. The threshold ΔF for the side reaction path can be determined based on a given reaction temperature, a given reaction time, and the proportion of a given competing side reaction.

[0050] 10 and 11 again, in the first half of the process of extracting a partial network from the coarse-grained network shown in Fig. 10, first, EQ0 input by the user is added to the reachable display EQ list (step S21).

[0051] Next, for the EQ in the displayed EQ list, it is determined whether there is a reaction path (edge) connected to that EQ that has the smallest reaction barrier (the difference in energy between the EQ and the reaction path) and whose relative energy from EQ0 is lower than the threshold value ΔE+ΔF (step S22). If there is one ("Yes" in step S22), it is determined whether there is an EQ that has a reaction path that satisfies the above conditions. i The energy barrier of the path with the lowest barrier is selected as E MIN (Step S23). If there is no reaction path that satisfies the above condition ("No" in Step S22), proceed to Step S31 (see FIG. 11). The edges connecting the nodes of the multiple nodes included in the partial network are edges on the PES whose relative energy from the initial EQ0 is lower than the sum of the threshold ΔE of the reaction barrier and the threshold ΔF for the side reaction path.

[0052] And EQ i Among the reaction pathways leading from E MIN +ΔF and the energy difference between EQ0 and EDGE is smaller than ΔE+ΔF. ij , EDGE ik , EDGE il EQ connected to j , EQ k , EQ l , ... are listed (step S24). j , EQ k , EQ l , …, EQ i If the energy is higher than (YES in step S25), j , EQ k , EQ l , ... are added to the display EQ list (step S26), and the process returns to step S22. j , EQ k , EQ l , …, EQ i If there is anything with higher energy than (YES in step S25), j , EQ k , EQ l, ...all are added to the display EQ list (step S26) and the process returns to step S22. EQ j , EQ k , EQ l ,...all of these are EQ i If it is lower than the threshold, the process proceeds to step S27 (see FIG. 11).

[0053] In the latter half of the process of extracting sub-networks from the coarse-grained network shown in Figure 11, EQ i and EQ j , EQ k , EQ l EDGE: All reaction paths connecting... ij , EDGE ik , EDGE il , remove..., EQ i All EQs connected to j , EQ k , EQ l , ... are added to the coarse-grained network. k and EQ j EDGE connects kj EDGE ij , EDGE ki The higher energy one is EDGE kj is defined as the energy of the newly added reaction path (step S27).

[0054] EQ k and EQ j Directly connecting reaction paths kj Before adding the new coarse-grained network (step S27), kj (Original EDGE kj If there is a .EDGE file (YES in step S28), kj and EDGE kj The lower of the energy of the newly added reaction paths is called EDGE kj (Step S29), and then return to Step S22 (see FIG. 10). kj (Original EDGE kj) does not exist ("No" in step S28), k and EQ j EDGE connects kj EDGE kj The energy of the newly added reaction path is assigned as information (step S30), and the process returns to step 22 (see FIG. 10). i EQ connected to j , EQ k , EQ l , …, EQ k and EQ j The same process is performed for all other combinations.

[0055] Finally, the EQs in the display EQ list are displayed, and among the edges connecting these EQs, those whose relative energy from EQ0 is lower than the value of ΔE+ΔF are displayed on a display device such as a computer display (step S31), and the process ends. In this way, a partial network is extracted from the coarse-grained network by extracting a partial network for a region kinetically reached from the initial EQ0 corresponding to the initial structure based on a given initial structure, reaction temperature, and reaction time. The partial network extracted from the coarse-grained network by the process shown in Figures 10 and 11 is, for example, a graph network such as those shown in Figures 14 to 17.

[0056] In the example of the user interface shown in FIGS. 14 to 17, the above-mentioned “Temp” is displayed at the top of the display screen. [K] ", "T sec ", "Side % In addition to items that accept various input values ​​such as "(reaction temperature, reaction time, ratio of competing reactions)," TSmax " and "E lim " can also accept input values ​​directly from the user. TSmax " corresponds to the threshold ΔE, and "E lim " corresponds to the threshold ΔF. The reaction temperature, reaction time, and the ratio of competing reactions ("Temp [K] ", "T sec ", "Side% ) is "E TSmax " and "E lim In the example user interface shown in Figures 14 to 17, TSmax " and "E lim The user can also directly input ". The present invention is not limited to this example, and the input values ​​of the reaction temperature, reaction time, ratio of competing reactions, and initial structure can be provided by the user, or can be received by various means such as another terminal device, a server device, etc.

[0057] Figures 15 to 17 show examples of extracting sub-networks from the same coarse-grained network, and the reaction time ("T sec "), the proportion of competing reactions ("Side % ”), the structure of the partial network extracted from the coarse-grained network changes. In the example shown in Figure 15, the user has specified a reaction temperature of 300K, a reaction time of 0.001 seconds, a proportion of competing reactions of 0.001%, and an initial structure of the clustered EQ with identification number 59.

[0058] A computer that executes the program of the present invention, or a reaction path network analysis device that executes the reaction path network analysis method of the present invention, can visualize a partial network extracted from the coarse-grained reaction path network based on various input values ​​(reaction temperature: 300K, reaction time: 0.001 seconds, proportion of competing reactions: 0.001%, initial structure: EQ with identification number 59), as shown in FIG. 15. In this case, "E TSmax " (threshold ΔE) and "E lim " (threshold ΔF) is calculated based on the various input values, and "E TSmax " (threshold ΔE) is 56.47, and "E lim ” (threshold ΔF) is 28.72.

[0059] By setting a longer reaction time, it is possible to expand the partial network extracted from the coarse-grained network. In the example shown in Figure 16, by setting the reaction time to 3600 seconds, it is possible to visualize edges and nodes (reaction paths and EQs) that were not visualized in the partial network shown in Figure 15. At this time, the "E" calculated based on the various input values ​​is TSmax " (threshold ΔE) is 94.12, and "E lim ” (threshold ΔF) is 28.72.

[0060] In addition, by increasing the proportion of competing reactions, the partial network extracted from the coarse-grained network can be made smaller. Figure 17 shows an example in which the proportion of competing reactions is set to 1% in the example shown in Figure 16, with the reaction temperature remaining at 300K and the reaction time at 3600 seconds. In this case, the "E" calculated based on the various input values ​​is TSmax " (threshold ΔE) is 94.12, and "E lim ” (threshold ΔF) is 11.49. In the example shown in Figure 17, the edges and nodes (reaction paths and EQs) that were visualized in the partial network shown in Figure 16 are no longer visualized, and a partial network that is smaller than the partial network shown in Figure 16 is displayed.

[0061] In this way, a computer that executes the program of the present invention, or a reaction path network analysis device that executes the reaction path network analysis method of the present invention, can efficiently analyze a partial region containing about several tens of edges necessary for analyzing a chemical reaction from a reaction path network containing several thousand edges by visualization as shown in Figures 14 to 17. This not only improves the efficiency of analysis based on the reaction path network, but also prevents arbitrary oversights.

[0062] In addition, in typical reaction path network analyses, such as SCAN (Non-Patent Document 4), the visualization of routes within a network requires specifying a start point and a goal point. However, the programs of the present invention can extract a partial network by inputting only the node information of the reaction's starting point, thereby streamlining the process. In other words, by specifying only the start point and not the goal, the user's work time can be significantly reduced. Furthermore, since the programs of the present invention do not require the goal point to be specified, they can also be used in cases where the goal product is unknown. Furthermore, when the values ​​required for quantum chemical calculations are input, a system implementing the programs of the present invention can automatically calculate the definition of the partial network, including the EQs to be reached and the edges between them, and visualize them.

[0063] A system incorporating the program of the present invention can simultaneously display multiple routes that cannot be extracted by methods that display the shortest route. In addition, because the target edges are partial, the calculation cost is unlikely to increase, and calculations can be performed easily using a web browser. [Industrial Applicability]

[0064] The reaction path network analysis program and reaction path network analysis method according to the present invention can be used to support the analysis of chemical reactions based on a reaction path network. [Explanation of symbols]

[0065] 100 Terminal Device 200 Server device N Network

Claims

1. 1. A program for analyzing a reaction path network expressed as a visualizeable graph network, in which each of a plurality of equilibrium structures (EQs) on a potential energy surface (PES) is a node, and reaction paths connecting the EQs are edges connecting the nodes, the program comprising: From a coarse-grained network constructed by clustering nodes of a graph network representing the reaction path network based on the identity or similarity of EQ structures, an initial EQ corresponding to the initial structure is selected based on a given initial structure, reaction temperature, and reaction time. 0 Extracting a partial network by calculating the EQ for the region kinetically reached from A program to execute.

2. The edges connecting the nodes of the plurality of nodes included in the partial network are determined by the initial EQ on the PES. 0 The program according to claim 1, wherein the relative energy from the edge is lower than the sum of the threshold ΔE of the reaction barrier and the threshold ΔF for the side reaction pathway.

3. the reaction barrier threshold ΔE is determined based on the reaction temperature and the reaction time, The program according to claim 2 , wherein the threshold ΔF for the side reaction pathway is determined based on the reaction temperature, the reaction time, and a rate of a given competing side reaction.

4. The program according to claim 2 , wherein the threshold value ΔE of the reaction barrier and the threshold value ΔF for the side reaction pathway can be directly input by a user.

5. 2. The program according to claim 1, wherein the coarse-grained network is constructed by clustering a plurality of EQs for identical or similar molecular structures into one group for each EQ corresponding to a node of a graph network representing the reaction pathway network, and treating each of the plurality of clustered groups as a single EQ.

6. the coarse-grained network includes, as nodes, a plurality of EQs corresponding to each of the plurality of groups after the clustering, and a plurality of reaction paths as edges connecting the nodes; The program according to claim 5 , wherein each of the plurality of reaction paths connects the EQs of the plurality of EQs.

7. The energy of the EQ of the coarse-grained network is the minimum energy among the energies of the EQs of the reaction pathway network combined into the EQ; The program according to claim 5, wherein the energy of a reaction path connecting the EQs of the plurality of EQs included in the coarse-grained network is the minimum energy among the energies of a plurality of reaction paths connecting the plurality of EQs of the reaction path network summarized in the EQ at one end of the reaction path and the plurality of EQs of the reaction path network summarized in the EQ at the other end.

8. 2. The program according to claim 1, further comprising, before the step of extracting the partial network, a step of constructing a coarse-grained network by clustering nodes of the graph network representing the reaction pathway network based on identity or similarity of EQ structures.

9. The program according to claim 1 , further comprising the step of displaying the partial network on a user interface of the computer.

10. A method for analyzing a reaction path network, which processes a graph network representing a reaction path network in which each of a plurality of equilibrium structures (EQs) on a potential energy surface (PES) is a node, and reaction paths connecting the EQs are edges connecting the nodes, comprising: From a coarse-grained network constructed by clustering nodes of a graph network representing the reaction path network based on the identity or similarity of EQ structures, an initial EQ corresponding to the initial structure is selected based on a given initial structure, reaction temperature, and reaction time. 0 Extracting a partial network by calculating the EQ for the region kinetically reached from A method comprising:

11. The edges connecting the nodes of the plurality of nodes included in the partial network are determined by the initial EQ on the PES. 0 The method of claim 10, wherein the relative energy from is lower than the sum of the threshold ΔE of the reaction barrier and the threshold ΔF for the side reaction pathway.

12. the reaction barrier threshold ΔE is determined based on the reaction temperature and the reaction time, 12. The method of claim 11, wherein the threshold ΔF for the side reaction pathway is determined based on the reaction temperature, the reaction time, and the rate of a given competing side reaction.

13. 12. The method of claim 11, wherein the reaction barrier threshold ΔE and the threshold ΔF for the side reaction pathway can be directly input by a user.

14. The method of claim 10, wherein the coarse-grained network is constructed by clustering a plurality of EQs for identical or similar molecular structures into a single group for each EQ corresponding to a node of the graph network representing the reaction pathway network, and grouping the group into a single EQ.

15. the coarse-grained network includes, as nodes, a plurality of EQs corresponding to each of the plurality of groups after the clustering, and a plurality of reaction paths as edges connecting the nodes; The method of claim 14 , wherein each of the plurality of reaction pathways connects two of the plurality of EQs.

16. The energy of the EQ of the coarse-grained network is the minimum energy among the energies of the EQs of the reaction pathway network combined into the EQ; The method according to claim 14, wherein the energy of a reaction path connecting the EQs of the plurality of EQs included in the coarse-grained network is the minimum energy among the energies of a plurality of reaction paths connecting the plurality of EQs of the reaction path network summarized in the EQ at one end of the reaction path and the plurality of EQs of the reaction path network summarized in the EQ at the other end.

17. The method of claim 11 , further comprising, prior to the step of extracting the partial network, constructing a coarse-grained network by clustering nodes of the graph network representing the reaction pathway network based on identity or similarity of EQ structure.

18. 18. The method of any one of claims 10 to 17, further comprising displaying the sub-network on a user interface of a device performing the method.

19. An apparatus for analyzing a reaction pathway network configured to carry out the method of any one of claims 10 to 17.

20. 20. The apparatus of claim 19, further configured to display the sub-network on a user interface of the reaction pathway network analyzer.

Images