Chemical structure generation device, chemical structure generation program, and chemical structure generation method

Abstract

To provide a chemical structure generation device, a method and a program for efficiently generating a molecular structure suitable for being used as learning data of machine learning.SOLUTION: In a chemical structure generation device 1, a generation function 100a generates a product list composed of one or more compounds based on a reactant list composed of one or more compounds and a chemical reaction list. A control function 100b provides the product list to a generation unit as a new reactant list, updates a database having at least one of the reactant list and the product list, and causes the generation function 100a to generate a new product list based on the new reactant list and the chemical reaction list.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a chemical structure generation device, a chemical structure generation program, and a chemical structure generation method. [Background technology]

[0002] There has been interest in computer simulation-based machine learning to design molecules with useful properties. For example, in a training phase, a trained model is generated using a training dataset through machine learning, such as deep learning, and then in an execution phase, the trained model is used to predict physical properties or search for molecules with desired properties.

[0003] Here, since a large amount of learning data (training data) is required for machine learning such as deep learning, it is important to efficiently generate high-quality learning data. From this perspective, a chemical structure generator that automatically generates molecular structures suitable for use as learning data for machine learning is important.

[0004] However, the number of possible molecular structures relative to the number of atoms increases exponentially as the number of atoms increases, resulting in a combinatorial explosion, making it practically impossible to comprehensively generate candidate molecular structures. Additionally, the ratio of available molecules, i.e., the number of molecules that actually exist stably and are commercially available, to the number of possible molecular structures rapidly decreases as the number of atoms increases. In machine learning, using training data containing a large number of molecular structures that are not commercially available is extremely inefficient in terms of time and cost. Therefore, a chemical structure generator that can efficiently generate only commercially available and valuable molecules is desired. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] International Publication No. 9501606 [Patent Document 2] U.S. Patent Application Publication No. 2018 / 096100 [Patent Document 3] Japanese Patent Application Laid-Open No. 2001-058962 [Patent Document 4] International Publication No. 9736252 Summary of the Invention [Problem to be solved by the invention]

[0006] An object of the present invention is to efficiently generate molecular structures suitable for use as learning data for machine learning. [Means for solving the problem]

[0007] In order to solve the above-mentioned problems and achieve the object, the chemical structure generator according to the present invention includes a generator and a controller. The generator generates a product list consisting of one or more compounds based on a reactant list consisting of one or more compounds and a chemical reaction list. The controller provides the product list to the generator as a new reactant list, updates a database containing at least one of the reactant list and the product list, and causes the generator to generate a new product list based on the new reactant list and chemical reaction list.

[0008] In addition, the chemical structure generation program according to the present invention is a program that causes a computer to execute the following processes: generating a product list consisting of one or more compounds based on a reactant list consisting of one or more compounds and a chemical reaction list; providing the product list as a new reactant list; updating a database that contains at least one of the reactant list and the product list; and generating a new product list based on the new reactant list and the chemical reaction list.

[0009] Furthermore, the chemical structure generation method according to the present invention is a chemical structure generation method executed by a chemical structure generation device, which generates a product list consisting of one or more compounds based on a reactant list consisting of one or more compounds and a chemical reaction list, provides the product list as a new reactant list, updates a database containing at least one of the reactant list and the product list, and generates a new product list based on the new reactant list and chemical reaction list. [Effects of the Invention]

[0010] According to the present invention, molecular structures suitable for use as learning data for machine learning can be efficiently generated. [Brief explanation of the drawings]

[0011] [Figure 1] FIG. 1 is a diagram showing an outline of a chemical structure generating apparatus according to an embodiment of the present invention. [Figure 2] FIG. 2 is a diagram showing an example of a chemical reaction list in the chemical structure generation device according to the first embodiment. [Figure 3] FIG. 3 is a diagram showing an example of a prohibited reaction list in the chemical structure generation device according to the first embodiment. [Figure 4] FIG. 4 is a flowchart showing the flow of processing performed by the chemical structure generation device according to the first embodiment. [Figure 5] FIG. 5 is a diagram illustrating the processing performed by the chemical structure generation device according to the first embodiment. [Figure 6] FIG. 6 is a flowchart illustrating in more detail the processing flow of step S200 in FIG. [Figure 7] FIG. 7 is a diagram illustrating the processing performed by the chemical structure generation device according to the first embodiment. [Figure 8] FIG. 8 is a flowchart showing the flow of processing performed by the chemical structure generation apparatus according to the second embodiment. [Figure 9]FIG. 9 is a flowchart illustrating in more detail the flow of the process in step S200A in FIG. DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, a chemical structure generating device according to an embodiment of the present invention will be described with reference to the accompanying drawings.

[0013] (First embodiment) FIG. 1 is a diagram illustrating an outline of a chemical structure generation device 1 according to an embodiment of the present invention. The chemical structure generation device 1 includes a processing circuit 100, a storage unit 120, a database 110, an input device 130, and a display 140. The processing circuit 100 includes a generation function 100a and a control function 100b. The database 110 includes a reactant list 110a, which is a list of reactants for a chemical reaction, a product list 110b, which is a list of products for the chemical reaction, a chemical reaction list 110c, a prohibited reaction list 110d, and a numerical information list 110e. The chemical structure generation device 1 is a device for generating chemical structures that serve as the basis for learning data used in machine learning, as the reactant list 110a or the product list 110b.

[0014] The processing circuit 100 is a processor (e.g., a CPU (Central Processing Unit), a GPU (Graphical Processing Unit), an ASIC (Application Specific Integrated Circuit), a PLD (Programmable Logic Device), etc.) that realizes functions such as a generating function 100a and a control function 100b by reading and executing a program from a storage unit 120. That is, the processing functions of the generating function 100a, the control function 100b, etc. are stored in the storage unit 120 in the form of a program executable by the processor. As a result, the processing circuit 100 in a state in which the program has been read out has the functions such as the generating function 100a and the control function 100b. That is, the generating function 100a and the control function 100b are examples of a generating unit and a control unit, respectively. Details of the processing of each function of the processing circuit 100 will be described later.

[0015] The database 110 is a database of various types referenced by the processing circuit 100 when the chemical structure generation device 1 generates new compounds, and is composed of, for example, a reactant list 110a, a product list 110b, a chemical reaction list 110c, a prohibited reaction list 110d, a numerical information list 110e, etc. These databases are stored in the storage unit 120 as needed.

[0016] The chemical reaction list 110c, the prohibited reaction list 110d, and the numerical information list 110e are data used when the processing circuit 100 performs the process of generating chemical structures. In contrast, the reactant list 110a or the product list 110b is data that serves as the basis for learning data used in subsequent machine learning, and generating these lists is one of the purposes of the process performed by the chemical structure generation device 1.

[0017] The reactant list 110a is a list of reactants and the like that are the basis of a chemical reaction, and is composed of compounds and the like that represent reactants. More precisely, the reactant list 110a is composed of one or more elements, and the elements are elements that represent operations that do not select simple substances, compounds, or reactants, or elements that represent intramolecular reactions. Here, the chemical structure generation device 1 according to the embodiment starts from an initial compound and sequentially updates the reactant list. Therefore, the reactant list 110a includes a reactant list X0 in the initial state (0th generation (p=0)), a reactant list X1 of the first generation, a reactant list X2 of the second generation, ..., a reactant list X3 of the pth generation. p Here, the reactant list X of the k+1th generation k+1 is the k-th generation reactant list X k or the k+1 generation reactant list X k+1 is the k-th generation reactant list X k may not be included entirely.

[0018] Also, the reactant list for each generation X k optionally containing a compound x representing a reactant k1 …x kn In addition to the special compound x, which represents a special operation null , x intra These special compounds will be described later. If these special compounds are not included as elements, the k-th generation reactant list X k The elements of compound x represent the kth generation reactants. k1、 x k2 …, x kn Conversely, if these special compounds are included as elements, the k-th generation reactant list X k The elements of, for example, x k1、 x k2 …, x kn , x null , x intra This becomes:

[0019] The product list 110b is a list of products of a chemical reaction, and is composed of compounds and the like that represent products. More precisely, the product list 110b is composed of one or more elements, and the elements are elements that represent operations that do not select simple substances, compounds, or reactants, or elements that represent intramolecular reactions. Here, similar to the reactant list 110a, the product list 110b is composed of a product list Y0 of the 0th generation (p=0), a product list Y1 of the 1st generation, a product list Y2 of the 2nd generation, ..., a product list Y3 of the pth generation. p It consists of:

[0020] Note that in some embodiments, the k+1 generation reactant list X k+1 The elements of are the product list Y k is equal to the elements of

[0021] The chemical reaction list 110c is a list of possible chemical reactions, which is a list of partial structures of reactants associated with partial structures of products produced by chemical reactions using the reactants. The chemical reaction list 110c also includes information on the cost of the chemical reactions, as needed.

[0022] FIG. 2 shows an example of a chemical reaction list 110c for two or one reactant and one corresponding product.

[0023] 2 describes a case where the structure "X1-CO-X2" is generated from the partial structure "X1-COOH" and the partial structure "X2-H." In this case, the list {{"X1-CO-X2"}, {"X1-COOH", "X2-H"}}, in which the partial structure list of the reactant {"X1-COOH", "X2-H"} and the partial structure list of the product {"X1-CO-X2"} generated by the chemical reaction are associated with each other, is an example of the chemical reaction list 110c.

[0024] Furthermore, the second line in FIG. 2 describes the case where the structure "X1-CHOH-X2" is generated from the partial structure "X1-CO-X2." In such a case, the list {{"X1-CHOH-X2"}, {"X1-CO-X2"}}, in which the partial structure list of the reactant {"X1-CO-X2"} and the partial structure list of the product ("X1-CHOH-X2") generated by the chemical reaction are associated, is an example of the chemical reaction list 110c. Furthermore, in this chemical reaction, H2 can also be considered as a partial structure and expressed as {{"X1-CHOH-X2"}, {"X1-CO-X2", "H2"}}.

[0025] 2, for example, a case where the structure "X1-CHX3-X2" is generated from the partial structure "X1-CHOH-X2" and the partial structure "X3-H" is described. In such a case, a list {{"X1-CHX3-X2"}, {"X1-CHOH-X2", "X3-H"}} in which the partial structure list of the reactant {"X1-CHOH-X2", "X3-H"} is associated with the partial structure list of the product ("X1-CHX3-X2") generated by the chemical reaction is an example of the chemical reaction list 110c.

[0026] 2 illustrates a case where there are two or one reactants and one corresponding product, but the embodiment is not limited to this, and there may be three or more reactants. For example, when structure D is generated from partial structure A, partial structure B, and partial structure C, an example of chemical reaction list 110c is a list {{D},{A,B,C}} in which the partial structure list {A,B,C} of the reactants corresponds to the partial structure list {D} of the product generated by the chemical reaction.

[0027] Furthermore, in the above example, the case where there is one product has been described, but the embodiment is not limited to this, and there may be two or more products. For example, consider a case where the structure "X1-COOH" and the structure "X2-H" are generated from the partial structure "X1-CO-X2" as the reverse reaction of the first line in FIG. 2. In this case, a list {{"X1-COOH", "X2-H"}, {"X1-CO-X2"}} in which the partial structure list {"X1-CO-X2"} of the reactant and the partial structure list {"X1-COOH", "X2-H"} of the product generated by the chemical reaction are associated with each other is an example of the chemical reaction list 110c.

[0028] The above content can be expressed using a formula, for example, as follows: k and substructure y l From the new structure y kl If ' is generated, the list {{y kl '}, {y k , y l}} is the chemical reaction list 110c. q1, y q2… y qn From the m new structures y q1q2…qn;1 , y q1q2…qn;2、 …, y q1q2…qn;m If a list {{y q1q2…qn;1 , y q1q2…qn;2、 …, y q1q2…qn;m},{y q1, y q2… y qn}} becomes the chemical reaction list 110c.

[0029] Note that in some embodiments, unlike the reactant list 110a and the product list 110b, the elements of the chemical reaction list 110c are not updated for each generation, but are fixed across generations.

[0030] The prohibited reaction list 110d is a list of prohibited chemical reactions, in which a list of partial structures of reactants is associated with a list of partial structures of products produced by chemical reactions carried out using the reactants, along with information that the reactions are prohibited chemical reactions.

[0031] For example, the substructure y k and substructure y l From the new structure y kl If the new structure produced has no commercial value and is not suitable for production, the list {{y kl '},{y k , y l}} is the prohibited reaction list 110d. In general, n substructures y q1, y q2… y qn From the m new structures y q1q2…qn;1 , y q1q2…qn;2、 …, y q1q2…qn;m If the product is not suitable as a product due to reasons such as lack of commercial value, q1q2…qn;1 , y q1q2…qn;2、 …, y q1q2…qn;m}, {y q1, y q2… y qn}} is the prohibited reaction list 110d. Whether or not a compound has commercial value is determined for each field based on, for example, cost, stability, toxicity, and the like.

[0032] An example of the prohibited reaction list 110d is shown in Fig. 3. Fig. 3 is a diagram illustrating processing related to the prohibited reaction list 110d.

[0033] The upper part of Figure 3 shows the case where the structure "X1-CO-X2" is generated from the partial structure "X1-COOH" and the partial structure "X2-H." Since there is no particular reason to restrict the generation of such a structure, the generation of this structure is allowed.

[0034] On the other hand, in the lower part of Figure 3, the structure "X1-COOH" and the partial structure "X2-CO-CH2-CO-X3" are used to derive the structure "X xThe list 110d describes the case where "X2-CO-CHX1-CO-X3" occurs. If the generated structure is chemically unstable and is judged to have no commercial value in a certain field, for example, the generation of such a structure is prohibited, and the generation of this structure is set to "Deny." In other words, the list {{"X2-CO-CHX1-CO-X3"}, {"X1-COOH", "X2-CO-CH2-CO-X3"}} becomes the prohibited reaction list 110d.

[0035] The numerical information list 110e is a list of numerical information that indicates the cost of each compound in the reactant list 110a, etc. The numerical information list 110e will be described in detail in the second embodiment.

[0036] The memory unit 120 is a storage area such as RAM (Random Access Memory), ROM (Read Only Memory), flash memory, or hard disk, and stores various programs executed by the processing circuit 100, the execution results of the various programs, a database 110, etc.

[0037] The input device 130 is a device for performing various operations by a user of the chemical structure generation device 1. The input device 130 is configured, for example, with a mouse, a keyboard, a touch panel, or hardware keys.

[0038] The display 140 displays various types of information. For example, the display 140 displays the processing results of the CPU, a GUI (Graphical User Interface) for receiving various operations from the user, and the like. The display 140 is configured as, for example, a liquid crystal display, an organic EL (Electro Luminescence) display, or a cathode ray tube display. Note that the input device 130 and the display 140 may be configured as one unit, for example, in the form of a touch panel.

[0039] Note that the embodiments are not limited to the above examples. Depending on the embodiment, the storage unit 120, the input device 130, the display 140, etc. may not be essential components of the chemical structure generator 1, i.e., the compound generator 1 may not include the storage unit 120, the input device 130, the display 140, etc. As another example, the storage unit 120, the input device 130, the display 140, etc. may be located outside the compound generator 1, for example, via a network, and data may be exchanged between the storage unit 120, the input device 130, the display 140, etc. and the processing circuit 100.

[0040] Next, the background of the embodiment will be described.

[0041] There has been interest in computer simulation-based machine learning to design molecules with useful properties. For example, in a training phase, a trained model is generated using a training dataset through machine learning, such as deep learning, and then in an execution phase, the trained model is used to predict physical properties or search for molecules with desired properties.

[0042] Here, since a large amount of training data is required for machine learning such as deep learning, it is important to efficiently generate high-quality training data. From this perspective, a chemical structure generator that automatically generates molecular structures suitable for use as training data for machine learning is important.

[0043] However, the number of possible molecular structures relative to the number of atoms increases exponentially as the number of atoms increases, resulting in a combinatorial explosion, making it practically impossible to comprehensively generate candidate molecular structures. Additionally, the ratio of available molecules, i.e., the number of molecules that actually exist stably and are commercially available, to the number of possible molecular structures rapidly decreases as the number of atoms increases. In machine learning, using training data containing a large number of molecular structures that are not commercially available is extremely inefficient in terms of time and cost. Therefore, a chemical structure generator that can efficiently generate only commercially available and valuable molecules is desired.

[0044] In view of this background, the processing circuit 100 included in the chemical structure generation device 1 according to the embodiment includes a generation function 100a and a control function 110b. The processing circuit 100 generates, via the generation function 100a, a product list 110b, which is a list of products consisting of one or more compounds, based on a reactant list 110a, which is a list of reactants consisting of one or more compounds, and a chemical reaction list 110c, which is a list of chemical reactions. The processing circuit 100 also provides, via the control function 100b, the product list 110b as a new reactant list to a program associated with the generation function 100a, updates the database 110, which contains at least one of the reactant list 110a and the product list 110b, and causes the program associated with the generation function 100a to generate a new product list based on the new reactant list and the chemical reaction list 110c.

[0045] As a result, the chemical structure generation device 1 according to the embodiment can efficiently generate high-quality training data suitable as training data for machine learning, and by using trained models generated using this training data, the efficiency of molecular design can be improved.

[0046] The details of this process will be explained using FIGS.

[0047] FIG. 4 is a flowchart showing the flow of processing performed by the chemical structure generation device according to the first embodiment.

[0048] First, the processing circuit 100 starts generating a compound list using the generating function 100a (step S100). The processing circuit 100 generates a p-th generation reactant list X0 by increasing p using the generating function 100a, with p=0 in the initial state, based on the 0th generation reactant list X0. p is generated as the compound list.

[0049] Next, the processing circuit 100 generates the reactant list X of the pth generation, which is the reactant list 110a, by the generating function 100a. p and the chemical reaction list 110c, a product list Y p The product list 110b is generated based on the reactant list 110a, which is made up of one or more compounds, and the chemical reaction list 110c (step S200). That is, the processing circuit 100 generates the product list 110b, which is made up of one or more compounds, using the generation function 100a. The detailed processing of step S200 will be described again with reference to FIG. 6.

[0050] Next, the processing circuit 100 generates the product list Y of the pth generation by the control function 100b. p into a new reactant list (the reactant list X of the p+1th generation) p+1 ) to the program related to the generating function 100a (step S300). That is, the processing circuit 100 uses the control function 100b to generate the product list Y p into a new reactant list X p+1 is given to the program relating to the generating function 100a.

[0051] In addition to executing this processing, the processing circuit 100 also uses the control function 100b to update the database 110, which has at least one of the reactant list 110a and the product list 110b. Here, updating the database 110 refers to an operation of changing the elements of the database 110 by adding new elements to the database 110, for example, an operation of inputting some elements of the product list 110b into the database 110 as new elements of the new reactant list 110a. For example, the processing circuit 100 uses the control function 100b to update the reactant list 110a in step S300.

[0052] At least a portion of the database 110 updated in this manner is used as training data in, for example, machine learning. As an example, the processing circuit 100 generates training data for machine learning based on the updated reactant list 110a using a training data generation function (not shown). As an example, the processing circuit 100 generates supervised data in which chemical structures are associated with the physical property values ​​of the chemical structures by associating the physical property values ​​of each element of the updated reactant list 110a using the training data generation function (not shown). This supervised data may be used as training data for machine learning. Machine learning is performed using such training data to generate a trained model, which makes it possible to predict the physical property values ​​of unknown chemical structures, thereby enabling, for example, the search for chemical structures with desirable physical property values. As another example, the processing circuit 100 generates unsupervised data based on the updated reactant list 110a using the training data generation function (not shown), and the unsupervised data may be used as training data for machine learning.

[0053] Next, the processing circuit 100 increments the value of p by 1 using the control function 100b (step S400). Here, if the value of p is equal to or greater than a preset threshold value p max If it is greater than the threshold value p (step S500: Yes), the process is completed.max If the value of p exceeds p, the processing circuit 100 having the control function 100b determines that the end condition of the product list creation process is satisfied, and ends the process. max If p is smaller than p (No in step S500), the processing circuit 150 repeats the processes of steps S200 to S400. That is, the processing circuit 100 causes the program associated with the generating function 100a to generate a new product list, using the control function 100b, based on the new reactant list in step S300 and the chemical reaction list in the chemical reaction list 110c. Note that the termination condition for the product list creation process is not limited to the above-described embodiment, in which the process is terminated when the number of generations p of repetitions exceeds a certain number of repetitions. The processing circuit 100 may terminate the product list creation process in accordance with another termination condition, such as when the number of elements in the generated product list reaches a certain value.

[0054] An example of compounds sequentially generated by this process is shown in Figure 5. Figure 5 is a diagram illustrating the process performed by the chemical structure generation device according to the first embodiment.

[0055] Compound 10a and compound 10b are examples of compounds included in the 0th generation reactant list X0 at the initial state, i.e., p=0. At p=0, in step S200, the processing circuit 100 generates compound 11 as a 0th generation product list Y0 based on compound 10a and compound 10b using the generation function 100a. At p=0, in step S300, the processing circuit 100 provides the 0th generation product list Y0 as a 1st generation reactant list X1 to the program related to the generation function 100a using the control function 100b. That is, compound 11 is included in the 1st generation reactant list X1. Furthermore, the processing circuit 100 having the control function 100b stores these product lists generated by the generation function 100a in the memory unit 120. Subsequently, in step S400, the processing circuit 100 increases the value of p by 1 to p=1 using the control function 100b.

[0056] Next, at p=1, in step S200, the processing circuit 100 generates compound 12a based on compound 10a and compound 11, and compound 12b based on compound 11, as a first-generation product list Y1, using the generation function 100a. At p=1, in step S300, the processing circuit 100 provides the first-generation product list Y1 to the program associated with the generation function 100a as a second-generation reactant list X2, using the control function 100b. That is, compounds 12a and 12b are included in the second-generation reactant list X1. Next, at step S400, the processing circuit 100 increases the value of p by 1 to p=2 using the control function 100b.

[0057] Similarly, the processing circuit 100 generates compound 13a based on compound 10b and compound 12b using the generation function 100a, and generates compound 14 based on compound 10a and compound 13a. In this way, the chemical structure generation device 1 can efficiently generate various compounds in sequence. Note that in FIG. 5, the chemical reaction that generates compound 13b from compound 12a and compound 11 corresponds to the chemical reaction included in the prohibited reaction list 110d, so in this case, the processing circuit 100 does not add the generated compound 13b to the product list.

[0058] Next, the process of step S200 in Fig. 4 will be described in detail with reference to Fig. 6. That is, the flowchart in Fig. 6 is a flowchart that explains the flow of the process of step S200 in Fig. 4 in more detail.

[0059] As mentioned above, the reactant list X for each generation k optionally containing a compound x representing a reactant k1 …x kn In addition to the special compound x, which represents a special operation null , x intra These special compounds will be explained later. First, let us consider the reactant list X of each generation. k However, the case where these special compounds are not included will be explained.

[0060] In addition, in the following, to avoid complicating the explanation, in FIG. 6, a case will be explained in which the chemical reaction used to generate the product list is a chemical reaction that produces one product from two reactants.

[0061] First, in step S210, the processing circuit 100 generates the product list Y of the pth generation in the product list 110b by the generation function 100a. p Here, the processing circuit 100 starts generating a 0th generation reactant list X0, a 1st generation reactant list X1, a 2nd generation reactant list X2, ..., a pth generation reactant list X p Based on this, the product list Y for each p generation is p The processing circuit 100 starts generating the product list Y p Immediately after starting the generation of p is an empty set. When the process is performed up to the p0th generation, the generated product list Y0, Y1...Y p0 In other words, the processing circuit 100 generates a product list Y0, Y1, ... Y p0 The data to be used as learning data for machine learning is generated as the union of the above.

[0062] First, in step S220, the processing circuit 100 generates a 0th generation reactant list X0, a 1st generation reactant list X1, a 2nd generation reactant list X2, ..., a pth generation reactant list X p Select an element from the reactant list of either m Extract as.

[0063] For example, the 0th generation reactant list X0={x 01、 x 02}, first generation reactant list X1={x 11 , x 12} as x 01 ="CH3OH", x02 ="CH4", x 11 ="CH3COOH", x 12 ="C2H5COOH", p=1. In this case, for example, in step S220, the processing circuit 100 selects one element from the 0th generation reactant list X0 by the generation function 100a, and 02 = "CH4" as the first element x m As another example, in step S220, the processing circuit 100 selects one element from the first-generation reactant list X1 using the generation function 100a, and extracts the element x 12 = "C2H5COOH" as the first element x m Extract as.

[0064] Next, in step S230, the processing circuit 100 generates the p-th generation reactant list X by the generating function 100a. p Select an element from the second element x n For example, if p=1, in step S230, the processing circuit 100 selects one element from the first-generation reactant list X1 by the generation function 100a, and extracts the element x 11 = "CH3COOH" as the second element x m In this way, in steps S220 and S230, the processing circuit 100 extracts one or more compounds from the reactant list using the generating function 100a.

[0065] The processing circuit 100 generates various first elements x by the generating function 100a. m and the second element x n The processes of steps S220 to S280 are repeated for each of the above.

[0066] In this iterative process, the processing circuit 100 iterates over all the first elements x m and the second element x n The processing of steps S220 to S280 may be performed for the combination of the first element x m and the second element x nAlternatively, the processes of steps S220 to S280 may be performed for only some of the possible combinations.

[0067] Subsequently, in steps S240 and S250, the processing circuit 100 selects, from the selected elements, a partial structure in which a chemical reaction will occur, using the generating function 100a.

[0068] In step S240, the processing circuit 100 generates the first element x extracted in step S220 by the generating function 100a. m From the first substructure y m For example, the first element x extracted in step S240 is m is "C2H5COOH", this structure matches the structure of "R-COOH", so in step S240, the processing circuit 100 generates y m the first partial structure y = "COOH" m As another example of extracting a substructure, the first element x extracted in step S240 is extracted as follows: m is "C2H5COOH", this structure matches the structure of the "RH" type, so in step S240, the processing circuit 100 generates y m ="H" as the first partial structure y m Extract.

[0069] Similarly, in step S250, the processing circuit 100 generates the second element x extracted in step S240 by the generating function 100a. n From the second substructure y n Extract.

[0070] The processing circuit 100 generates various first partial structures y m and a second partial structure y n The processes of steps S240 to S280 are repeated for each of the above.

[0071] Next, the processing circuit 100 reads the chemical reaction list 110c from the storage unit 120 using the generation function 100a, and determines whether the structures extracted in steps S240 and S250 are included in the chemical reactions listed in the chemical reaction list 110c. m and the second substructure y m If the chemical reaction list does not match (No in step S260), a new first partial structure y m and a second partial structure y n The processes of steps S240 to S280 are repeated for the first partial structure y m and the second substructure y n If the first partial structure y matches the chemical reaction list (Yes in step S260), the process proceeds to step S265. m is "COOH" and the second partial structure y n is "H", these partial structures match the chemical reaction list {"X1-CO-X2", {"X1-COOH", "X2-H"}. Therefore, in such a case, the processing circuit 100 determines that the generation function 100a can generate a new compound through the chemical reaction, and the process proceeds to step S265.

[0072] In this way, the processing circuit 100 uses the generating function 100a to extract from the chemical reaction list 110c chemical reactions in which one or more of the compounds extracted in steps S220 and S230 are reactants.

[0073] Next, in step S265, the processing circuit 100 reads the prohibited reaction list 110d from the storage unit 120 using the generating function 100a, and generates the first partial structure y m and the second partial structure y extracted in step S250 n Determine whether the first substructure y matches the prohibition rule. m and the second substructure y m If the combination of the partial structures y and y coincides with the prohibition rule (Yes in step S265), the processing circuit 100 does not perform any new processing for this combination of partial structures, and instead creates a new first partial structure ym and a second partial structure y n For example, the processing of steps S240 to S280 is repeated for the first partial structure y m is "-COOH" and the second partial structure y n is "-CO-CH2-CO-", these partial structures match the prohibited reaction list {"X2-CO-CHX1-CO-X3", {"X1-COOH", "X2-CO-CHX1-CO-X3"}. Therefore, in such a case, the processing circuit 100 determines that the new compound generated by the chemical reaction by the generation function 100a has no commercial value, and the processing circuit 100 does not perform new processing for this combination of partial structures, and generates a new first partial structure y m and a second partial structure y n The processes of steps S240 to S280 are repeated for the first partial structure y m and the second substructure y n If the result does not match the prohibition rule (No in step S265), the process proceeds to step S270, and the obtained product is added to the product list Y p will be added.

[0074] In this way, the processing circuit 100 generates the product list Y by the generating function 100a, further based on the prohibition rules. p In the above example, the prohibition rule is defined on a chemical reaction basis, that is, the prohibition rule is a rule that determines chemical reactions to be excluded from the chemical reaction list in step S260. However, the embodiment is not limited to this. That is, the prohibition rule may be defined on a product or reactant basis. In other words, the prohibition rule is defined as a rule that determines chemical reactions to be excluded from the product list Y p Products excluded from the product list Y p The embodiment is not limited to the above example, and the processing circuit 100 generates the product list Y by the generation function 100a based on the permission rule that determines the chemical reactions to be included in the chemical reaction list, rather than the prohibition rule that determines the chemical reactions to be excluded from the chemical reaction list. pmay be determined.

[0075] Subsequently, in step S270, the processing circuit 100 generates the product y based on the chemical reaction list using the generation function 100a. mn For example, the first element x m is "C2H5COOH" and the second element x n is "CH4" and the first partial structure y n is "-COOH" and the second partial structure y m is "-H" and the chemical reaction list is {"X1-CO-X 2」、 In the case of {"X1-COOH", "X2-H"}}, the processing circuit 100 generates the product y mn would be "C2H5-CO-CH3".

[0076] Subsequently, in step S275, the processing circuit 100 generates the product y generated in step S270 by the generating function 100a. mn , i.e., the product list Y p The processing circuit 100 determines whether the candidate compound y is included in the list of products generated previously, i.e., whether it is a compound that has already been generated in a previous process. mn is included in the list of previously generated products, that is, is determined to be a compound that has already been generated in a previous process (step S275 Yes), the processing circuit 100 mn For the pth generation product list Y p and a new first partial structure y m and a second partial structure y n The processes of steps S240 to S280 are repeated for each of the above.

[0077] Meanwhile, the processing circuit 100 generates the product y generated in step S270 by the generating function 100a. mnis not a compound already produced in the previous process (No in step S275), the process proceeds to step S280, and the processing circuit 100 mn Let Y be the product list of the pth generation. p In this way, the processing circuit 100 generates the product y of the chemical reaction extracted in step S260 by the generation function 100a. nm The product produced in step S270 is referred to as a product Y consisting of one or more compounds. p Add to list.

[0078] The processing circuit 100 generates the first element x given by the generating function 100a. m and the second element x n For the first partial structure y m and a second partial structure y n While changing the first partial structure y m and a second partial structure y n The processing circuit 100 repeats the processing of steps S240 to S280 for various first elements x m and the second element x n The processes of steps S220 to S280 are repeated for each of the above.

[0079] It should be noted that the embodiment is not limited to the above example.

[0080] 6, the chemical reaction used to produce the compound is a chemical reaction that produces one product from two reactants (the number of dimensions of the reaction is two), but the embodiment is not limited to this. As an example, the chemical reaction used to produce the compound may be a chemical reaction that produces one product from three reactants (the number of dimensions of the reaction is three), or may be a chemical reaction that produces two products from one reactant, or may be a chemical reaction that produces one product from one reactant that is different from the reactant.

[0081] For example, a case will be described in which the chemical reaction used to generate a compound is a chemical reaction that generates one product from three reactants. In this case, instead of steps S220 to S230, the processing circuit 100 generates a list of reactants X0 to X1 of 0th to pth generations by the generation function 100a. p The first element x from one of the reactant lists m1 and extract the reactant list X0 to X for generations 0 to p. p The second element x from one of the reactant lists m2 Extract the p-th generation reactant list X p to the third element x m3 Also, instead of steps S240 to S250, the processing circuit 100 extracts the first element x by using the generating function 100a. m1 to the first substructure y m1 Extract the second element x m2 to the second substructure y m2 Extract the third element x m3 to the third substructure y m3 The processing circuit 100 extracts the first element x given by the generating function 100a. m1 , the second element x m2 , the third element x m3 For the combination of various first substructures y m1 , the second substructure y m2 , the third substructure y m3 , and repeat these operations for various first elements x m1 , the second element x m2 , the third element x m3 This is done for the combination of:

[0082] Also, for example, a case will be described in which the chemical reaction used to generate a compound is a chemical reaction that generates two products from one reactant. In this case, instead of steps S220 to S230, the processing circuit 100 generates a p-th generation reactant list X p to element x m Also, instead of steps S240 to S250, the processing circuit 100 extracts the element xm to substructure y m In step S260, the processing circuit 100 extracts the partial structure y m A chemical reaction that matches is obtained from the chemical reaction list 110c, and after checking for prohibition rules in step S265, in step S270, two products y m:1、 y m;2 In step S275, it is determined whether each of these products is a compound that has already been generated. If the product is not a compound that has already been generated, in step S280, the product is added to the p-th generation product list Y p will be added.

[0083] These processes can be easily generalized to cases where the chemical reactions used to produce the compounds are those that produce r products from q reactants.

[0084] The chemical reaction used to produce a compound may be a chemical reaction that produces one product from one reactant, such as a chemical reaction that produces a product "R-Br" from a reactant "R-OH." The chemical reaction can be expressed, for example, as {{"R-Br"}, {"R-OH"}}.

[0085] In the embodiment, the processing circuit 100 having the generating function 100a generates the reactant lists X0 to X1 of the 0th to pth generations in step S220. p to the first element x m In step S230, the p-th generation reactant list X p to the second element x n Although the case where the reactant lists X0 to X1 of the 0th generation to the pth generation are extracted has been described, the embodiment is not limited to this. For example, the processing circuit 100 having the generating function 100a also extracts the reactant lists X0 to X1 of the 0th generation to the pth generation in step S230. p to the second element x n may be extracted.

[0086] Next, in FIG. 6, reactant list Xp However, special compound x null The case where a special compound x is included as an element will be explained. null is a special element that indicates an operation that does not select a compound, and the special compound x null is selected, the compound is treated as not being selected.

[0087] For example, reactant list X p is a normal compound with n elements x1, x2…x n and special compounds x null Let us consider a list of n+1 elements consisting of the special compound x. null Compounds other than x k , x l , x m When three reactants are extracted, the processing circuit 100 generates x k , x l , x m Extract the three-dimensional response, which is the three x null is extracted once, and x null Compounds other than x k , x l When two compounds are extracted, one compound is treated as not being selected, and the processing circuit 100 generates a signal by generating function 100a that the reactant is x k , x l Extract the two-dimensional response, which is two of x null is extracted twice, and x null Compounds other than x k When one compound is extracted, the two compounds are treated as not being selected, and the processing circuit 100 generates a signal by generating function 100a that the reactant is x k In this way, we extract one of the one-dimensional reactions from the reactant list X p Let x1, x2…x be the n compounds n and special compounds x nullBy extracting d elements from the list, allowing duplication, the processing circuit 100 can count all chemical reactions with the number of reactants ranging from 1 to d, i.e., the reaction dimension is d or less, without omission, using the generating function 100a. That is, the reactant list X p The operation that does not select a compound is null The processing circuit 100 generates the reaction unit list X by the generating function 100a. p Instead of extracting the compound, null By extracting the element x null The compounds obtained by chemical reactions involving fewer reactants than when not extracted are listed in the product list Y p can be added to.

[0088] Also, reactant list X p may include as an element an operation of converting the structure of a certain compound to obtain a compound different from the original compound. Examples of "an operation of converting the structure of a certain compound to obtain a compound different from the original compound" include operations such as intramolecular reaction, isomerization, and functional group substitution.

[0089] Such an operation will be explained using an intramolecular reaction as an example with reference to Fig. 7. Fig. 7 is a diagram illustrating the processing performed by the chemical structure generation device according to the first embodiment.

[0090] First, we will explain the case of a normal chemical reaction that is not an intramolecular reaction. p Consider the case where compound x0 containing the formula is compound 15 shown in Figure 7. In the case of a normal chemical reaction that is not an intramolecular reaction, two molecules of compound 15 come together to produce compound 16, as shown in the middle of Figure 7. In this case, the reactant of the chemical reaction is represented as {"compound 15", "compound 15"}, the product of the chemical reaction is represented as "compound 16", and the chemical reaction is represented as {{"compound 16"}, {{"compound 15", "compound 15"}.

[0091] Next, let us consider the case where an intramolecular reaction is included. When an intramolecular reaction is present, as shown in the lower part of FIG. 7, an intramolecular reaction occurs in one molecule of compound 15, resulting in the generation of compound 17. In this case, the reactant of the chemical reaction can be expressed as "compound 15", the product of the chemical reaction as "compound 17", and the chemical reaction as {"compound 17", "compound 15"}. However, as will be described below, the processing circuit 100 having the generating function 100a can express the operation x that performs the intramolecular reaction as follows: intra reactant list X p That is, the processing circuit 100 having the generating function 100a can treat the reactants of the reaction as elements included in the operation x intra Using the special element corresponding to {"Compound 15", x intra}, the product of the chemical reaction is "Compound 17", the chemical reaction is {{"Compound 17"}, {"Compound 15", x intra 6. The processing circuit 100 having the generating function 100a performs the operation x intra reactant list X p By treating a compound in the reactant list as an element in the product list Y p As shown in this example, by treating an operation on a molecule as an element included in a set of molecule candidates, the processing circuit 100 having the generation function 100a can systematically generate molecule candidates obtained by an operation such as an intramolecular reaction.

[0092] In step S270, the processing circuit 100 generates the product y nm is usually some chemical structure (compound), but the embodiment is not limited thereto. For example, in step S270, the processing circuit 100 generates a special compound x by the generation function 100a. null ya x intra etc., product y nm For example, the processing circuit 100 may generate all elements selected as reactants by the generating function 100a as x nullIf so, in step S270, x null The processing circuit 100 also generates x as a product. intra If so, in step S270, x intra If only special compounds are selected as reactants, the processes of steps S240, S250, S260, S265, S275, etc. in FIG. 6 do not need to be performed.

[0093] As described above, in the chemical structure generator 1 according to the first embodiment, the processing circuit 100 generates a product list consisting of one or more compounds based on a reactant list consisting of one or more compounds and a chemical reaction list using the generation function 100a, and provides the product list as a new reactant list using the control function 100b, causing the generator to generate a new product list based on the new reactant list and chemical reaction list. This makes it possible to efficiently generate molecular structures that are suitable for use as learning data for machine learning, have commercial value, and can actually be manufactured.

[0094] (Second embodiment) In the second embodiment, a chemical structure generation device will be described that generates commercially valuable molecular structures more efficiently by assigning numerical information to each candidate compound and controlling it, as a method for generating commercially valuable molecular structures more efficiently.

[0095] In the second embodiment, the chemical structure generator 1 assigns numerical information to each candidate compound based on the numerical information list 110e and performs control based on the numerical information. Here, the numerical information assigned to each compound is, for example, information indicating the cost required to generate the compound.

[0096] Typically, the numerical information is given for each compound. For example, the p-th generation reactant list X p But compounds x1, x 2、 …x n Compounds x1, x 2、 …, x n The costs of generating w1 and w 2、 …, w n In this case, the list of numerical information W corresponding to the reactant list of the pth generation is p The elements of are w1, w 2、 …w n This becomes:

[0097] The processing circuit 100 having the generating function 100a integrates the reactant list 110a and the numerical information list 110e to generate a single compound list X' in which the numerical information is set as an attribute for each compound. p ={x1(w1), x2(w2), … x n (w n )} can also be generated.

[0098] The processing circuit 100 calculates the numerical information for each compound using the generating function 100a. Specifically, the processing circuit 100 having the generating function 100a provides the known numerical information as numerical information for compounds whose numerical information is known, and for the numerical information of other compounds, recursively defines the numerical information based on the numerical information of the reactants of the compound based on the synthetic route of the compound, until it arrives at a compound whose numerical information is known.

[0099] The flow of this process will be specifically described using Figures 8 and 9. Figures 8 and 9 are flowcharts showing the flow of the process performed by the chemical structure generation apparatus according to the second embodiment. Figure 8 is a flowchart showing the overall flow of the process, and Figure 9 is a flowchart explaining the process of step S200A in Figure 8. Note that in Figures 8 and 9, repeated explanations of the processes already explained in the first embodiment will be omitted.

[0100] First, the processing circuit 100 starts generating a compound list using the generating function 100a (step S100A). The processing circuit 100 generates a p-th generation reactant list X0 by increasing p based on the 0th generation reactant list X0 and the numerical information list W0 for the reactants, with p=0 in the initial state. p and p-generation reactant list X p Numerical information list W corresponding to each reactant p As an example, the processing circuit 100 having the generating function 100a generates a reactant list X0={x 1、 x2, …x n} and a list of information (cost information) such as numerical information W0 = {w1, w 2、…、 w n} is initially generated, and based on these, a reactant list X0'={x1(w1), x2(w2), ..., x n (w n )} is initially generated, and the generation of the compound list begins.

[0101] Next, the processing circuit 100 generates the reactant list X of the pth generation, which is the reactant list 110a, by the generating function 100a. p and a chemical reaction list 110c, which is a list of chemical reactions, to generate a product list Y p As part of this process, the processing circuit 100 generates the product list Y p Numerical information list W corresponding to each product of p That is, the processing circuit 100 having the generation function 100a calculates the product list Y p In step S200A, the process is performed in detail with reference to FIG.

[0102] Next, the processing circuit 100 generates the product list Y of the pth generation by the control function 100b. pand a list of numerical information W corresponding to each of the products p , respectively, into a new reactant list (the reactant list X p+1 ) and new numerical information list W p+1 In this way, the processing circuit 100 having the control function 100b generates the product list Y p to a new reactant list (the reactant list X of the p+1th generation p+1 ) is given to the program related to the generating function 100a.

[0103] Next, the processing circuit 100 increments the value of p by 1 through the control function 100b (step S400). Next, the processing circuit 100 increments the value of p by 1 through the control function 100b. p In the second embodiment, the processing circuit 100 having the generating function 100a determines whether the numerical information list W is an empty set. p As p increases, a new reactant list X p Therefore, the processing circuit 100 having the control function 100b generates a new reactant list X p The p for which p is an empty set can be used as the end point of processing for p.

[0104] That is, a new reactant list X p If is an empty set (step S500A Yes), the process ends and a new reactant list X p is not an empty set (No in step S500A), the process returns to step S200A. By setting such a processing stop condition, for example, compounds whose costs are less than a given threshold and which have high commercial value can be efficiently generated.

[0105] Next, the details of the process of step S200A will be described with reference to Fig. 9. In Fig. 9, the numerical information list W corresponding to each reactant is pis information representing the cost of each of the reactants as a product.

[0106] As with FIG. 6, FIG. 9 illustrates a case where the chemical reaction used to generate the product list is a chemical reaction that produces one product from two reactants, but as already mentioned, the embodiment can be similarly applied to cases where the number of reactants or products is different.

[0107] First, in step S210A, the processing circuit 100 generates the p-th generation product list Y p The processing circuit 100 starts generating the product list Y p Immediately after starting the generation of p is the empty set.

[0108] As in the first embodiment, in step S220A, the processing circuit 100 generates, by the generation function 100a, a 0th generation reactant list X0, a 1st generation reactant list X1, a 2nd generation reactant list X2, ..., a pth generation reactant list X p Select an element from the reactant list of either m Extract as.

[0109] Next, in step S230A, the processing circuit 100 generates the p-th generation reactant list X by the generating function 100a. p Select an element from the second element x n In this way, in steps S220 and S230, the processing circuit 100 extracts one or more compounds from the reactant list using the generating function 100a.

[0110] The processing circuit 100 generates various first elements x by the generating function 100a. m and the second element x n The processes of steps S220A to S280A are repeated for each of the above.

[0111] Subsequently, in step S231, the processing circuit 100 generates the first element x selected in step S220A by the generating function 100a. m and the second element x selected in step S230A. n For a reaction including the above, the cost of the reactants (sum of raw material costs) is calculated. Here, the cost of the product is the sum of the cost of the reactants (sum of raw material costs) and the cost involved in carrying out the chemical reaction. If the sum of the raw material costs exceeds the allowable cost, the cost of the product will exceed the maximum allowable cost, regardless of the cost involved in the chemical reaction. Therefore, the processing circuit 100 having the generating function 100a performs pruning processing in step S232 before entering the loop related to the substructure.

[0112] That is, the cost of the reactants, e.g., the first element x selected in step S220A m Costs related to m and the second element x selected in step S230A. n Costs related to n The sum of m +w n exceeds the allowable cost (No in step S232), the processing circuit 100 having the generating function 100a performs the following: m and the second element x n is determined to be an inappropriate element, and a new first element x m and the second element x n The processes of steps S220A to S280A are repeated for the reactant cost, for example, the first element x selected in step S220A. m Costs related to m and the second element x selected in step S230A. n Costs related to n The sum of m +w n However, if the cost does not exceed the allowable cost (Yes in step S232), the process proceeds to step S240.

[0113] Subsequently, in steps S240 and S250, the processing circuit 100 selects a substructure in which a chemical reaction will occur from the selected elements by the generating function 100a. For example, in step S240, the processing circuit 100 selects a substructure in which a chemical reaction will occur from the first element x extracted in step S220 by the generating function 100a. m From the first substructure y m Similarly, in step S240, the processing circuit 100 extracts the second element x extracted in step S230 by the generating function 100a. n From the second substructure y n Extract.

[0114] The processing circuit 100 generates various first partial structures y m and a second partial structure y n The processes of steps S240 to S280A are repeated for each of the above.

[0115] Next, the processing circuit 100 reads the chemical reaction list 110c from the storage unit 120 using the generation function 100a, and determines whether the structures extracted in steps S240 and S250 are included in the chemical reactions listed in the chemical reaction list 110c. m and the second substructure y m If the chemical reaction list does not match (No in step S260), a new first partial structure y m and a second partial structure y n The processes of steps S240 to S280A are repeated for the first partial structure y m and the second substructure y n If the list of chemical reactions matches the list of chemical reactions (Yes in step S260), the process proceeds to step S270.

[0116] In this way, the processing circuit 100 uses the generating function 100a to extract from the chemical reaction list 110c chemical reactions in which one or more of the compounds extracted in steps S220A and S230A are reactants.

[0117] Subsequently, in step S270, the processing circuit 100 generates the product y based on the chemical reaction list using the generation function 100a. mn Generate.

[0118] Subsequently, in step S275, the processing circuit 100 generates the product y generated in step S270 by the generating function 100a. mn , i.e., the product list Y p The processing circuit 100 determines whether the candidate compound y is included in the list of products generated previously, i.e., whether it is a compound that has already been generated in a previous process. mn is included in the list of previously generated products, that is, is determined to be a compound that has already been generated in a previous process (step S275 Yes), the processing circuit 100 mn For the pth generation product list Y p and a new first partial structure y m and a second partial structure y n The processes of steps S240 to S280 are repeated for each of the above.

[0119] Meanwhile, the processing circuit 100 generates the product y generated in step S270 by the generating function 100a. mn However, if it is determined that the compound is not a compound that has already been generated in the previous process (No in step S275), the process proceeds to step S271A.

[0120] In step S271A, the processing circuit 100 having the generating function 100a calculates the product y generated in step S270. mn Regarding the generation cost w mn Here, the generation cost w mn is the sum of the reactant costs w m +w n and the cost of the chemical reaction z mn That is, the processing circuit 100 having the generating function 100a generates the product list Y pCandidate compound y mn The corresponding numerical information is calculated as the sum of the costs of the reactants that are the raw materials for the candidate compounds, w m +w n and candidate compound y from the reactants mn A chemical reaction that produces y mn、 {x m、 x n Cost z related to}} mn Here, the processing circuit 100 having the generating function 100a generates the cost z mn from the chemical reaction list 110c. The chemical structure generation device 1 stores the chemical reaction and the cost of the chemical reaction in the chemical reaction list 110c in the storage unit 120 in association with each other.

[0121] Subsequently, in step S272, the processing circuit 100 uses the generating function 100a to generate the product y calculated in step S271A. mm The generation cost of w mn However, the maximum allowable cost c max Determine whether the generation cost w mn However, the maximum allowable cost c max If it is not equal to or less than the above (No in step S272), mn is the product list Y p is not added to the first substructure y m and a second partial structure y n The processes of steps S240 to S280A are repeated for the generation cost w mn However, the maximum allowable cost c max If it is equal to or less than this (step S272 Yes), the process proceeds to step S280A.

[0122] In this way, the processing circuit 100 having the generating function 100a generates the product list Y p Candidate compound y mn When generating the candidate compound y mn The corresponding numerical information is the generation cost w mn is calculated in step S271A, and the calculated numerical information, the generation cost w mnIn step S272, a candidate compound y mn , the product list Y p Decide whether to include it in the

[0123] That is, reactant list X p Each of the reactants included in Y is associated with information representing the cost of the reactant, and the processing circuit 100 having the generating function 100a generates a product list Y p Candidate compound y mn The cost of mn is calculated in step S271A, and the calculated product list Y p Candidate compound y mn The cost of mn is set to the maximum allowable cost c max Compared to the cost mn The maximum allowable cost c is the threshold max (No in step S272), the candidate compound y mn , the product list Y p Not included.

[0124] In step S280A, processing circuit 100 calculates the product y mn Let Y be the product list of the pth generation. p Add to the product y mn The generation cost of w mn The p-th generation numerical information list W p In this way, the processing circuit 100 generates a product Y consisting of one or more compounds by the generation function 100a. p Add to list.

[0125] The processing circuit 100 generates the first element x given by the generating function 100a. m and the second element x n For the first partial structure y m and a second partial structure y n While changing the first partial structure y m and a second partial structure y nThe processing circuit 100 repeats the processing of steps S240 to S280A for various first elements x m and the second element x n The processes of steps S220A to S280A are repeated for each of the above.

[0126] The embodiments are not limited to the above examples.

[0127] In the embodiment, a numerical information list W corresponding to each compound is p is the information representing the cost of the compound, the embodiment is not limited to this. For example, p may be a score determined for each compound. As an example, the processing circuit 100 having the generating function 100a determines the overall commercial usefulness of each compound based on various factors, such as not only the price of the raw materials but also the technical difficulty of handling, ease of procuring the raw materials, and market demand, and determines a score for each compound, and compounds with a high score or above a certain threshold may be preferentially added to the list of reactants of the next generation. Here, the score determined for each compound may be such that a higher score indicates a more desirable compound, or conversely, such that a lower score indicates a more desirable compound. In addition, the numerical information list W corresponding to each compound may be p For example, information on the physical properties of the compound, such as molecular weight, vapor pressure, boiling point, melting point, dipole moment, oil-water partition coefficient (for example, logP), etc. may be used.

[0128] Also, in the embodiment, the processing circuit 100 having the generating function 100a generates the product list Y p When generating candidate compounds for the product list Y, numerical information corresponding to each candidate compound is further calculated, and the candidate compounds are sorted into a product list Y based on the calculated numerical information. p However, the embodiment is not limited to this. For example, the processing circuit 100 having the generation function 100a determines whether or not to include the product list Y pWhen generating candidate compounds for the product list Y, additional information other than numerical information corresponding to each candidate compound is calculated, and the candidate compounds are sorted based on the calculated information. p Here, examples of information that is not numerical information include information on whether a compound contains a specific substituent or information on a hazardous material classification according to an official organization. In such a case, for example, the processing circuit 100 having the generation function 100a may determine whether or not a compound that is a candidate for the product list contains a specific substituent or is classified into a specific hazardous material classification, and add the compound to the product list Y. p Exclude from.

[0129] In the flowchart of FIG. 9, the first element x m and the second element x n For the iterative process of m and the second element x n The selection of each compound is shown in the numerical information list W p The processing circuit 100 having the generating function 100a sequentially performs the first element x in accordance with the priority based on the numerical information of each compound, for example, in order of lowest cost or highest score. m and the second element x n Search for the product list Y p Candidate for y mn may be generated.

[0130] Furthermore, in the embodiments, the cost of a compound is not limited to information indicating only the simple price of the compound, but may also include various costs other than price, such as the difficulty of safety management, the ease of procurement of raw materials, and their stability.

[0131] Furthermore, the numerical information in the embodiments is not limited to numerical information in the form of a scalar quantity, but may be numerical information in the form of a vector quantity or a tensor quantity, as necessary.

[0132] In the embodiment, the processing circuit 100 generates a p-th generation product list Y based on the 0th generation reactant list X0.p In other words, in the embodiment, one reactant list Y is generated based on one initial list called the 0th generation reactant list X0. p However, the embodiment is not limited to this, and a plurality of reactant lists corresponding to the plurality of initial lists may be generated based on the plurality of initial lists.

[0133] For example, consider the case where a certain compound (this compound itself is not necessarily the subject of calculation by this compound generator) requires two or more raw material groups (e.g., group A, group B, group C, etc.), and compound lists for each raw material group (a compound list for group A, a compound list for group B, a compound list for group C, etc.) are generated within an allowable cost.

[0134] At this time, for example, the processing circuit 100 generates a 0th generation reactant list X for the compounds in group A by the generation function 100a. 0、A Based on this, a p-th generation compound list Y for compounds in group A is created. p、A and generating a 0th generation reactant list X0 for the compounds of group B, B Based on this, a p-th generation compound list Y for compounds in group B is created. p , B and a list of reactants X for generation 0 of the compounds in group C 0、C Based on this, a p-th generation compound list Y for compounds in group C is created. p , to generate c.

[0135] First, consider the first case where an allowable cost is set for each ingredient group. In this case, the processing circuit 100 uses the generation function 100a to calculate the allowable cost set for each ingredient group as the maximum allowable cost c maxFor example, consider the case where a compound list is generated for two raw material groups, group A and group B, where the total allowable cost c is given and the ratio of the allowable cost of group A to the allowable cost of group B is set to 8:2. In this case, the processing circuit 100 determines that the allowable cost for group A is c max =0.8c, and for group B, the allowable cost is c max =0.2c and perform each process.

[0136] Next, consider the second case where the allowable cost is set for the entire raw material group, rather than for each raw material group individually. In this case, the processing circuit 100 uses the generation function 100a to perform processing for each raw material group in parallel or alternately, and then, in step S272 of FIG. 9, determines whether the production cost of the product is less than or equal to the maximum allowable cost while referring to information on the allowable costs of the compound lists generated for the other raw material groups. As an example, in step S272 of generating the compound list for group A, if the sum of the production cost of the currently generated product and the minimum production cost in the compound list for group B exceeds the overall allowable cost, the processing circuit 100 uses the generation function 100a to not add the product to the product list.

[0137] By performing the above processing, the processing circuit 100 can simultaneously generate compound lists for two or more raw material groups in parallel.

[0138] In this way, in the second embodiment, the processing circuit 100 assigns numerical information to each candidate compound and controls the generation of chemical structures, thereby enabling more efficient generation of commercially valuable molecular structures suitable for use as training data for machine learning.

[0139] As described above, according to the embodiment of the present invention, molecular structures suitable for use as learning data for machine learning can be efficiently generated.

[0140] It should be noted that the embodiments of the present invention are presented merely as examples, and various substitutions, modifications, etc. can be made.

[0141] With respect to the above-described embodiments, the following supplementary notes are disclosed as optional features of the invention.

[0142] (Addendum) A data processing system comprising: a database having at least one of a reactant list consisting of one or more compounds and a product list consisting of one or more compounds; and a chemical structure generator, The chemical structure generator comprises: a generator that generates the product list based on the reactant list and the chemical reaction list; and a control unit that provides the product list to the generating unit as a new reactant list, updates the database, and causes the generating unit to generate a new product list based on the new reactant list and the chemical reaction list. [Explanation of symbols]

[0143] 100 Processing Circuit 100a generation function 100b Control Function 110 databases 110a Reactant List 110b Product List 110c Chemical Reaction List 110d Prohibited Reaction List 110e Numerical Information List 120 Storage section 130 Input Device 140 Display

Claims

1. a generator for generating a product list consisting of one or more compounds based on a reactant list consisting of a list of reactants and a possible chemical reaction list in which a list of partial structures of the reactants is associated with a list of partial structures of products produced by chemical reactions carried out using the reactants; a control unit that provides the product list to the generating unit as a new reactant list, updates a database that has at least one of the reactant list and the product list, and causes the generating unit to generate a new product list based on the new reactant list and the chemical reaction list; The reactant list includes a specific compound representing a specific operation as an element; The special operation is an operation that does not select a compound, or an operation that transforms the structure of a certain compound into a compound different from the original compound.

2. the reactant list includes, as an element, the special compound representing an operation in which no compound is selected; 2. The chemical structure generation device according to claim 1, wherein the generation unit extracts the elements from the reactant list instead of extracting compounds, thereby adding to the product list compounds obtained by chemical reactions involving a smaller number of reactants than when the elements are not extracted.

3. 2. The chemical structure generator according to claim 1, wherein the reactant list includes, as an element, an operation for converting the structure of a compound into a compound different from the compound.

4. 2. The chemical structure generation device according to claim 1, wherein the generation unit adds a compound obtained by performing an intramolecular reaction on any compound in the reactant list to the product list by treating an operation of performing an intramolecular reaction as an element included in the reactant list.

5. Each of the reactants is associated with information; 5. The chemical structure generation device according to claim 1, wherein, when generating candidate compounds for the product list, the generation unit further calculates information corresponding to each of the candidate compounds, and determines whether or not to include the candidate compound in the product list based on the calculated information.

6. 6. The chemical structure generation apparatus according to claim 5, wherein the information associated with each of the reactants is numerical information.

7. the numerical information associated with each of the reactants is information representing the cost of each of the reactants as a product, 7. The chemical structure generation device according to claim 6, wherein the generating unit generates the numerical information corresponding to each of the candidate compounds using a sum of a cost of the reactant that is a raw material of the candidate compound and a cost associated with a chemical reaction that produces the candidate compound from the reactant.

8. Numerical information is associated with each of the reactants; 5. The chemical structure generation device according to claim 1, wherein said generating unit generates candidates for said product list in accordance with a priority order based on said numerical information.

9. Each of the reactants is associated with information representing a cost of the reactant; the generating unit calculates costs of candidate compounds for the product list based on information representing costs of the reactants, and compares the calculated costs of candidate compounds for the product list with a preset threshold; 7. The chemical structure generator of claim 5, wherein if the cost exceeds the threshold, the candidate compound is not included in the product list.

10. a process of generating a product list consisting of one or more compounds based on a reactant list consisting of a list of reactants, the reactant list consisting of one or more compounds, and a possible chemical reaction list, the chemical reaction list including a list of partial structures of the reactants and a list of partial structures of products generated by chemical reactions using the reactants associated with each other; providing the product list as a new reactant list; updating a database containing at least one of the reactant list and the product list; generating a new product list based on the new reactant list and the chemical reaction list; on the computer, The reactant list includes a specific compound representing a specific operation as an element; A chemical structure generation program, wherein the special operation is an operation that does not select a compound, or an operation that transforms the structure of a certain compound to produce a compound different from the original compound.

11. A chemical structure generation method performed by a chemical structure generator, comprising: generating a product list consisting of one or more compounds based on a reactant list consisting of a list of reactants and a possible chemical reaction list in which a list of partial structures of the reactants is associated with a list of partial structures of products produced by chemical reactions carried out using the reactants; providing said product list as a new reactant list; updating a database containing at least one of the reactant list and the product list; generating a new product list based on the new reactant list and the chemical reaction list; The reactant list includes a specific compound representing a specific operation as an element; The method for generating a chemical structure, wherein the special operation is an operation that does not select a compound, or an operation that converts the structure of one compound into a compound different from the original compound.

Images