Retrosynthesis analysis system, retrosynthesis analysis apparatus, retrosynthesis analysis method, and retrosynthesis analysis program

Abstract

This invention provides a retrosynthesis analysis system, retrosynthesis analysis apparatus, retrosynthesis analysis method, and retrosynthesis analysis program that can efficiently explore the synthesis route of a target compound. [Solution] In a retrosynthesis analysis system in which a management server and a user terminal are connected via a network for communication, the management server 1 includes an input information receiving unit 101 that receives the target compound to be analyzed, a preprocessing unit 102 that uses the target compound and / or an intermediate compound of the target compound as a base compound and performs preprocessing to simplify a part of the molecular structure of the base compound to set up a provisional structure compound, and an analysis unit 103 that performs retrosynthesis analysis on the provisional structure compound and determines the synthesis route of the base compound.

Description

Technical Field

[0004] , , , ,

[0005] , , , ,

[0001] The present disclosure relates to a retrosynthetic analysis system, a retrosynthetic analysis apparatus, a retrosynthetic analysis method, and a retrosynthetic analysis program for analyzing a synthesis route of a compound.

Background Art

[0002] As shown in Patent Document 1, a method of formulating a synthesis route of a target compound by retrosynthetic analysis is known. In retrosynthetic analysis, starting from the target compound, the bonds contained in the molecular structure of the target compound are cleaved and decomposed into precursors. By repeating such a decomposition step into precursors until the generated precursors reach available compounds such as commercially available drugs, a multi-step synthesis route for obtaining the target compound is formulated. In such an information processing system for performing retrosynthetic analysis, a machine learning model such as a neural network that has learned known chemical reactions and compounds, and a path search algorithm based on reinforcement learning are used.

Prior Art Documents

Patent Documents

[0003] ​​​​​​​​​​​​​​​​​For example, retrosynthesis analysis techniques typically employ a one-step process (a process that repeatedly identifies the chemical reaction that generates a reaction site and the compound before the reaction for each reaction site), which has the limitation that only one reaction site change can be detected per step. In actual chemical reactions, multiple reaction sites may be generated in a single reaction (for example, multiple substituents with the same structure may be added to a precursor in a single reaction), but retrosynthesis analysis techniques using one-step processing cannot detect such chemical reactions, sometimes resulting in unnecessarily complex synthetic routes.

[0006] Furthermore, when attempting to determine the synthesis route of compounds with complex structures, such as those with large molecular weights or cyclic structures, it is necessary to consider various factors that can affect the chemical reaction, such as the reactivity of substituents, steric hindrance, and competitive reactions. However, conventional retrosynthesis analysis techniques tend to prioritize well-known or frequently performed chemical reactions, sometimes resulting in the output of chemical reactions that are incompatible with the structure of the target compound or chemical reactions that are impossible to perform.

[0007] The object of the exemplary embodiments of this disclosure is to provide a retrosynthesis analysis system, retrosynthesis analysis apparatus, retrosynthesis analysis method, and retrosynthesis analysis program that can efficiently explore the synthesis route of a target compound. [Means for solving the problem]

[0008] The retrosynthesis analysis system relating to one aspect of this disclosure is: An input information receiving unit that receives the target compound to be analyzed, A preprocessing unit that uses the target compound and / or an intermediate compound of the target compound as a base compound, performs a preprocessing to simplify a part of the molecular structure of the base compound, and sets up a provisional structure compound, The system includes an analysis unit that performs retrosynthesis analysis on the provisional structure compound and determines the synthesis pathway of the base compound.

[0009] The retrosynthesis analysis system possesses the above-mentioned features, which allows for more efficient exploration of synthetic pathways in retrosynthesis analysis. [Brief explanation of the drawing]

[0010] [Figure 1] Figure 1 is a diagram illustrating the configuration of a retrosynthesis analysis system according to one embodiment of the present disclosure. [Figure 2] Figure 2 is a block diagram showing the hardware configuration of the management server shown in Figure 1. [Figure 3] Figure 3 is a block diagram illustrating the software configuration of the management server shown in Figure 1. [Figure 4] Figure 4 is a diagram illustrating one example of the pretreatment process. [Figure 5] Figure 5 illustrates another example of the pretreatment process. [Figure 6] Figure 6 illustrates another example of the pretreatment process. [Figure 7] Figure 7 is a diagram that simulates the analysis results (comparative example) when retrosynthesis analysis is performed without pretreatment. [Figure 8] Figure 8 shows an example of the results of retrosynthesis analysis. [Figure 9] Figure 9 is a diagram that simulates the analysis results (comparative example) when retrosynthesis analysis is performed without pretreatment. [Figure 10] Figure 10 shows another example of the retrosynthesis analysis results. [Figure 11] Figure 11 is a diagram that simulates the analysis results (comparative example) when retrosynthesis analysis is performed without pretreatment. [Figure 12] Figure 12 shows another example of the retrosynthesis analysis results. [Figure 13] Figure 13 is a flowchart illustrating the retrosynthesis analysis method according to this embodiment. [Figure 14] Figure 14 is a flowchart illustrating the retrosynthesis analysis method according to this embodiment. [Modes for carrying out the invention]

[0011] A reverse synthesis analysis system according to an embodiment of the present disclosure will be described while referring to the drawings. In each of the attached drawings, the same or similar elements are given the same or similar reference numerals and names, and duplicate descriptions regarding the same or similar elements may be omitted in the description of the embodiment. Note that the content shown in each drawing is merely an example for explaining the present embodiment, and is only a schematic example shown for ease of explaining the present embodiment. The content of each drawing may be modified or changed within a range where no technical problems occur.

[0012] <System Overview> The reverse synthesis analysis system according to the present embodiment is an information processing system that receives an input of a target compound, which is the final product, from a user, performs a reverse synthesis analysis on the target compound, and formulates a synthesis route from a compound that is available to the user, such as a commercially available drug, to the target compound.

[0013] <Overview of Reverse Synthesis Analysis> The target compounds analyzed in this retrosynthesis analysis system are primarily organic compounds. In this system, retrosynthesis analysis is performed using a machine learning model constructed from a predetermined neural network such as a CNN (Convolutional Neural Network) or RNN (Recurrent Neural Network), and / or a search algorithm using reinforcement learning such as Monte Carlo tree search. The architecture of the neural network employed in this system is not particularly limited; other machine learning models, such as linear models or tree models, may also be used. Furthermore, the training data for the machine learning model used for retrosynthesis analysis is not particularly limited; for example, a compound database, a chemical reaction database, or supervised labeling training data designed based on these databases may be used. The machine learning model employed in this system may be configured to accept structural information of compounds as input and predict chemical reactions and reaction compounds (compounds used in the reaction). The machine learning model may be a single model or a combination of multiple models responsible for each prediction. While the prediction of chemical reactions and reaction compounds may be performed using classical computational processing without machine learning, this embodiment assumes the use of a machine learning model. Furthermore, the search algorithm employed in this system performs pathfinding based on a reward function such as the number of synthesis steps, using chemical reactions and reaction compounds predicted by a machine learning model. Note that the search algorithm may employ active learning methods other than Monte Carlo tree search, such as reinforcement learning or Bayesian optimization, or other classical search algorithms. The reward function for the search is also not particularly limited.

[0014] In the retrosynthetic analysis system of this embodiment, in retrosynthetic analysis, first, an input of a target target compound is received from the user. In retrosynthetic analysis, the determination of chemical reactions and reaction compounds by a machine learning model and a search algorithm is regarded as one step, and this step is repeated until all intermediate compounds used for the generation of the target compound reach available compounds such as commercially available drugs, thereby formulating a series of synthetic routes for synthesizing the target compound. The algorithm for retrosynthetic analysis is not particularly limited, and known algorithms can be adopted.

[0015] For example, the machine learning model for retrosynthetic analysis searches for retrons (the smallest substructures that can be synthesized by a predetermined chemical reaction) contained in the target compound by referring to a compound database, a chemical reaction database, etc., thereby determining the bond cleavage sites (i.e., chemical reactions) and the compounds separated by the bond cleavage (i.e., compounds before the reaction such as starting compounds, intermediate compounds (precursors), etc.). The molecular fragments generated by the bond cleavage are called synthons. Since a synthon is merely a fragment and cannot exist in reality, in retrosynthetic analysis, the reaction compound that serves as the source of the synthon during the reaction is determined from the compound database. In this embodiment, the compound (reactant) corresponding to the synthon determined by retrosynthetic analysis may be referred to as a "synthon compound".

[0016] From the perspective of outputting the optimal synthesis route, it is preferable to use a search algorithm constructed using reinforcement learning or the like in conjunction with a machine learning model. The search algorithm for retrosynthesis analysis refers to the output of the machine learning model to determine candidate chemical reactions for producing the target compound. For example, for each candidate chemical reaction and reaction compound of the target step output by the machine learning model, the search algorithm further determines candidate chemical reactions (i.e., steps prior to the target step in the synthesis route) for producing that reaction compound based on the machine learning model. By repeating this series of processes multiple times, candidate synthesis routes for the target compound are calculated. The search algorithm then evaluates these routes using a reward function and determines the chemical reactions and reaction compounds that can output a synthesis route leading to a higher reward as the chemical reactions at the target step. The search algorithm determines the overall synthesis route by repeating the above process.

[0017] After determining a one-step chemical reaction using the machine learning model and / or search algorithm described above, if one of the intermediate compounds (synthon compound or reaction compound) is difficult to obtain as a raw material, the next step of retrosynthesis analysis is performed using the intermediate compound as input data to the retrosynthesis analysis algorithm to search for a chemical reaction that produces the intermediate compound. In this embodiment, the target compound and intermediate compounds that serve as input data to the machine learning model in each step of the retrosynthesis analysis are referred to as "base compounds." Intermediate compounds that are available as raw materials are referred to as "registered compounds." Registered compounds may be compounds registered as commercially available drugs (commercial compounds), compounds that are not commercially available but are registered as being independently managed or owned by a user (user-registered compounds), or compounds from among the commercially available compounds and user-registered compounds that are in stock (stock compounds).

[0018] In retrosynthesis analysis, a "synthetic route search is successful" if all intermediate compounds determined at a predetermined step are registered compounds. Then, a "retrosynthetic phylogenetic tree" showing the series of chemical reactions from the target compound to the final step is output, with this step being the final step of retrosynthesis (the step that is the first chemical reaction during synthesis). If the final step (the step in which all intermediate compounds are registered compounds) cannot be found as described above, the retrosynthetic steps will be repeated, but if the number of steps becomes too large, it will not be a practical synthetic route. Therefore, in retrosynthesis analysis, predetermined upper limits (search limiting conditions) are set for the search time, number of search steps, cost, etc., as analysis conditions, and if the final step cannot be found before reaching these conditions, the retrosynthesis analysis calculation process is terminated as a "synthetic route search failure".

[0019] Retrosynthesis analysis requires the design of efficient synthetic routes with fewer steps. However, conventional retrosynthesis analysis systems identify the pre-reaction structure of each reaction site in the target compound's structure in a single step, which can result in the output of complex synthetic routes for target compounds that could be synthesized using simple routes. Furthermore, algorithms that prioritize well-known or frequently performed chemical reactions are employed when searching for chemical reactions. While setting such priority chemical reactions can be beneficial in streamlining chemical reaction searches, it also presents challenges such as outputting chemical reactions that do not fit the structure of the target compound or outputting chemical reactions that are impossible to perform.

[0020] In the retrosynthesis analysis system of this embodiment, a pretreatment is applied to the base compound to simplify its molecular structure, thereby enabling the planning of more efficient synthetic routes than conventional methods. The details of the retrosynthesis analysis system of this embodiment will be described below.

[0021] <Components of the retrosynthesis analysis system> As shown in Figure 1, the retrosynthesis analysis system of this embodiment comprises a management server 1 and one or more user terminals 2. The management server 1 and the user terminals 2 are connected to each other via a network NW. In this embodiment, the network NW is primarily assumed to be the internet, but the network NW is not limited to the internet and may be constructed using, for example, a public telephone network, a mobile phone network, a wireless communication network, Ethernet (registered trademark), etc. Note that the configuration of the retrosynthesis analysis system shown in Figure 1 is just one example, and one configuration may combine other configurations, or other configurations may be included.

[0022] The management server 1 is an information processing device (retrosynthesis analyzer) that performs processing related to retrosynthesis analysis, and may be managed by a provider of the retrosynthesis analysis system. The management server 1 may be a general-purpose computer such as a workstation or personal computer, or it may be logically implemented by cloud computing. As shown in Figure 2, the management server 1 includes at least a processor 10, memory 11, storage 12, a transceiver 13, an input / output unit 14, etc., which are electrically connected to each other via a bus 15. Note that the illustrated configuration is just an example, and the management server 1 may have a different configuration.

[0023] The processor 10 is a computing unit that controls the operation of the entire management server 1, controls the transmission and reception of data between each element, and performs information processing necessary for application execution and authentication processing. For example, the processor 10 is a CPU (Central Processing Unit) and / or a GPU (Graphics Processing Unit), and executes programs stored in the storage 12 and loaded into the memory 11 to perform various information processing tasks.

[0024] Memory 11 includes main memory composed of volatile storage devices such as DRAM (Dynamic Random Access Memory) and auxiliary memory composed of non-volatile storage devices such as flash memory and HDD (Hard Disk Drive). Memory 11 is used as a work area for the processor 10 and also stores the BIOS (Basic Input / Output System) executed when the management server 1 starts up, as well as various configuration information.

[0025] Storage 12 stores various programs, such as application programs. A database containing data used for each process may be built in storage 12. For example, the memory unit 120 described later may be provided in a part of the memory area of ​​memory 11 and / or storage 12.

[0026] The transmitting / receiving unit 13 is a communication interface for the management server 1 to communicate with user terminals 2, etc., via a communication network. The transmitting / receiving unit 13 may further include short-range communication interfaces such as Bluetooth® and BLE (Bluetooth Low Energy) and / or USB (Universal Serial Bus) terminals.

[0027] The input / output section 14 consists of information input devices such as keyboards and mice, and output devices such as displays.

[0028] Bus 15 is connected in common to all of the above elements and transmits, for example, address signals, data signals, and various control signals.

[0029] User terminal 2 is, for example, a terminal used by a user in an organization that conducts research and development of a target substance. In the retrosynthesis analysis system, the user operates user terminal 2 to perform various input operations, such as specifying the target compound and setting the conditions for retrosynthesis analysis. User terminal 2 may be, for example, a general-purpose computer such as a workstation or personal computer, or a mobile terminal such as a smartphone or tablet.

[0030] User terminal 2 also includes a processor, memory, storage, transceiver, input / output unit, etc., which are electrically connected to each other via a bus. The hardware configuration of user terminal 2 can be configured in the same way as that of management server 1 described above, and a detailed explanation of each element in the hardware configuration of user terminal 2 is omitted.

[0031] <Functions of Management Server 1> Figure 3 is a block diagram illustrating the functions implemented in the management server 1. In this embodiment, the management server 1 may include an input information receiving unit 101, a preprocessing unit 102, an analysis unit 103, and a display control unit 104. The storage unit 120 of the management server 1 may also include various databases such as a compound DB (database) 121, a chemical reaction DB 122, an analysis information storage unit 123, and a specific structure DB 124. Although the various functional units shown in Figure 3 are illustrated as functional units realized by the processor 10 of the management server 1, some or all of the various functional units may be realized by the processor of the user terminal 2.

[0032] Compound DB121 is a database that aggregates information on compounds and is referenced during retrosynthesis analysis. The data for each compound included in Compound DB121 is not necessarily limited. For example, in addition to basic information such as the substance name, structural formula, empirical formula, and CAS registry number of each compound, Compound DB121 may also include physical property data (melting point, boiling point, solubility, etc.), safety data (toxicity, flammability, handling precautions, etc.), and information on availability (manufacturer, product number, price, stock status, etc.). The empirical formula of each compound may be stored in a machine-readable format such as SMILES format, InChI format, or MOL file format.

[0033] Compounds registered in Compound DB121 may be managed in two categories. For example, the first category consists of compounds available as raw materials for the synthesis of commercially available drugs, etc. (hereinafter referred to as raw material compounds), and the second category consists of compounds produced by a predetermined synthesis, such as intermediate compounds and final compounds. Raw material compounds registered in the first category are not limited to commercially available compounds; compounds managed or owned independently by the user may also be registered. Information on raw material compounds registered in the first category may include information such as whether or not they are in stock, the quantity in stock, the price, storage location, source, and manufacturer, while information on compounds registered in the second category may include information such as the synthesis method and the manufacturing cost for synthesis.

[0034] Compound DB121 may be constructed using publicly available, known databases, or it may be constructed by adding information on compounds managed and held by the user to a public database, or a database independently set up (and updated as needed) by the user may be adopted as Compound DB121. Furthermore, Compound DB121 may be expanded by adding new target compounds entered by the user for retrosynthesis analysis, intermediate compounds output from retrosynthesis analysis, etc. (i.e., the information in Compound DB121 may be updated based on past retrosynthesis analysis results). Compound DB121 may also be updated based on public databases provided by reagent manufacturers, etc. When updating Compound DB121, information necessary for updates, such as inventory status, may be automatically collected from the reagent manufacturer's website and other public databases using technologies such as web scraping. Database updates such as adding, changing, or correcting compounds in Compound DB121 may be performed by the system provider.

[0035] Chemical Reaction DB122 is a database that aggregates information on chemical reactions, which is referenced during retrosynthesis analysis. The chemical reactions aggregated in Chemical Reaction DB are not limited to known chemical reactions, but may also include novel chemical reactions generated by information technologies such as machine learning models. The data for each chemical reaction recorded in Chemical Reaction DB122 is not necessarily limited. For example, Chemical Reaction DB122 may include information such as the name of the chemical reaction, the reaction equation, information identifying the reactants (starting compounds) and products (information identifying the compound name, molecular structure, etc.), reaction conditions (temperature, pressure, type of solvent, reaction time, type and amount of catalyst, etc.), yield, by-products, cost of the chemical reaction, and other information related to the reaction (e.g., production-related information such as the ratio of starting materials, order of addition, and post-treatment). In Chemical Reaction DB122, the reaction equation, reactants, products, etc. may also be stored in a machine-readable format.

[0036] Similar to the compound database (DB121), the chemical reaction database (DB122) may be constructed using publicly available databases. Users may add known chemical reactions to such public databases, or modify, delete, or correct the database contents to construct the chemical reaction database (DB122). Furthermore, the chemical reaction database (DB122) may be updated by collecting new information on chemical reactions from academic journal databases and other public databases, or by updating it based on past retrosynthesis analysis results. When collecting information on chemical reactions from public databases, techniques such as web scraping may be used. Updates to the chemical reaction database (DB122) may also be performed by the system provider.

[0037] The analysis information storage unit 123 stores information related to the results of retrosynthetic analyses performed in the past. This information may include, for example, the date and time the analysis was performed, information about the analysis requester (user) (user ID, affiliated organization, etc.), information indicating the target compound entered by the user, information about the conditions set during the analysis (maximum search time, maximum number of search steps, restrictions on usable raw materials, etc.), the details of the pretreatment performed in the preprocessing unit 102 (simplified parts, masked substituents, etc.), the synthesis route output by the analysis unit 103, evaluation indicators for the synthesis route (yield, cost, number of steps, etc.), the type of reaction prediction model (machine learning model) used in the retrosynthetic analysis and its parameters, the type of optimization algorithm (search algorithm) used for route searching and its parameters. Access to the retrosynthetic analysis results recorded in the analysis information storage unit 123 may be restricted so that only the user who requested the analysis can view them. The information related to retrosynthetic analysis results recorded in the analysis information storage unit 123 may be referenced by the analysis unit 103 when performing a new retrosynthetic analysis.

[0038] The specific structure DB124 stores information about molecular structures that may be subject to pretreatment, as described later. For example, the specific structure DB124 may store a list of substructures that are candidates for reaction-constraining substituents in a predetermined format (e.g., SMILES format, MOL file format, etc.). Candidate structures for reaction-constraining substituents may be set by collecting information on substituents designated by the user as reaction-constraining substituents from past retrosynthesis analysis results registered in the analysis information storage unit 123, or they may be set based on the knowledge of experts such as chemists.

[0039] The various data stored in the compound DB121, chemical reaction DB122, analysis information storage unit 123, and specific structure DB124 described above can be used as training data to build or enhance machine learning models that perform retrosynthesis analysis. Furthermore, the various databases included in the storage unit 120 do not necessarily need to be built within the management server 1, but may be located on an external database server. For example, the compound DB121 and chemical reaction DB122 may be configured to directly reference external databases provided by reagent manufacturers, etc. Also, the analysis information storage unit 123 may be configured to be managed by a dedicated storage server from the standpoint of ensuring security. In addition to the above databases, the storage unit 120 may also store user-related data such as user identification numbers, data related to various permissions, and other various data used in retrosynthesis analysis (for example, a database related to the cost of chemical synthesis).

[0040] The input information receiving unit 101 receives the specification of the target compound to be analyzed from the user. The format in which the user inputs the target compound is not particularly limited; for example, the input of the target compound may be accepted in a machine-readable format such as SMILES format or MOL file format, or it may also be accepted by substance name or structural formula. When the input information receiving unit 101 receives the input of the target compound in a machine-readable format such as SMILES format, it may generate the structural formula of the compound entered by the user and present it to the user to confirm whether there are any errors in the input. The input information receiving unit 101 may also accept input from the user regarding the conditions for retrosynthesis analysis. The items to be set as conditions for retrosynthesis analysis are not particularly limited; for example, conditions related to search time, number of search steps, cost, etc. These conditions may be specified by the user, or conditions predetermined by the system provider may be used.

[0041] Furthermore, the input information receiving unit 101 may also accept the designation of sites (substituents) to be pretreated, such as reaction-constraining substituents, as described later. The designation of sites to be pretreated may also be entered in a machine-readable format such as SMILES format, similar to the input of the target compound, or in the form of substituent names, structural formulas, etc.

[0042] The preprocessing unit 102 uses the target compound and / or an intermediate compound of the target compound as base compounds and performs a preprocessing to simplify a part of the molecular structure of the base compound, thereby setting up a provisional structure compound.

[0043] Here, "using the target compound and / or an intermediate compound of the target compound as the base compound" means that the pretreatment may be performed on the target compound before the start of the retrosynthetic analysis (before the first pathway search is performed), on the intermediate compound at an intermediate step in the synthetic pathway, or performed before the start and then again at any intermediate step. Furthermore, "simplifying part of the molecular structure" includes, for example, "deleting" part of the molecular structure, "masking" part of the molecular structure, or "converting" to a structure in which part of the molecular structure has been deleted or to a simpler structure. In the pretreatment, one of the following processes, "deletion," "masking," or "conversion," may be performed, or a combination of these processes may be performed. In this embodiment, the compound with the structure after pretreatment will be referred to as the "provisional structure compound." The pretreatment will be described in detail below based on the examples shown in Figures 4-6.

[0044] Figure 4 illustrates a pretreatment process in which part of the molecular structure is "removed". Figure 4(a) shows the molecular structure of 4,4'-(4-(tetrahydrofuran-2-yl)phenyl)azanediyl)diphenol, which is exemplified as Example 1. As shown enclosed by the dashed line in Figure 4(b), the compound of Example 1 has two substituents (substituent Ia and substituent Ib) of the same structure.

[0045] These two substituents with the same structure are substituents that can be formed (added) simultaneously in a single chemical reaction. The pretreatment unit 102 determines whether there are multiple substituents with the same structure within the molecular structure of the base compound. As shown in Figure 4(b), if it determines that there are multiple substituents with the same structure, it sets up a provisional structure compound by deleting at least one of the multiple substituents with the same structure. Figure 4(c) is an example of a provisional structure compound set up by pretreatment, where the provisional structure compound is set up by deleting substituent Ia while leaving substituent Ib (substituent Ib may also be deleted while substituent Ia is left). When at least one substituent with the same structure is deleted in this way, the pretreatment unit 102 may set up the provisional structure compound by substituting the deleted part with a hydrogen atom, or by substituting the deleted part with an atom other than a hydrogen atom (Figure 4(c) shows an example where it is replaced with a hydrogen atom).

[0046] Figure 4 illustrates the case where there are two substituents with the same structure, but the same pretreatment can be performed even when there are three or more substituents with the same structure, and the selection of substituents to be removed is not particularly limited. More specifically, one substituent may be kept from among multiple substituents with the same structure, and the others may be removed (i.e., at least one must be kept).

[0047] In addition, in molecular structures having multiple substituents of the same structure, symmetrical regions may arise within the molecular structure. For example, in the compound shown in Figure 4, region A enclosed by the dashed line in Figure 4(b) can be said to be a symmetrical region. The preprocessing unit 102 may identify a symmetrical region from the molecular structure of the base compound and set up a provisional structure compound by deleting at least one substituent included in that symmetrical region. More specifically, of the multiple substituents that create the symmetry in the symmetrical region, one substituent may be left and the others deleted. Even when simplifying the molecular structure of the base compound based on symmetry in this way, a provisional structure compound (Figure 4(c)) is set up in the same way as when searching for substituents of the same structure as described above.

[0048] As a method for identifying the parts to be deleted by summarizing structural identity and symmetry, for example, a method of referring to the surrounding structure may be adopted, as shown below. In this method, the surrounding structure up to radius r is used as the reference range, starting from an atom to which a structure such as a substituent has been attached, such as the nitrogen atom (N) in the compound in Figure 4(a). The "radius r" shown here is a value that indicates the range within the molecular structure by nodes and edges, with the vertices of atoms, etc. in the molecular structure being nodes and the spaces between vertices such as bonds being edges, and is expressed as an integer (note that radius r does not define a circle expressed in units of length such as angstroms). In other words, the "surrounding structure from the starting point to radius r" means the range from the starting atom to the node that is separated by the value of the edge indicated by radius r. For example, in the compound shown in Figure 4(a), if we set the radius r to "5" and the base point to the nitrogen atom, the range from the nitrogen atom along three edges to the OH group of substituent Ia, the OH group of substituent Ib, and the tetrahydro-2-furyl group (tetrahydrofuranyl group) corresponds to the "peripheral structure from the base point to radius r=5".

[0049] As described above, after defining the reference range, when comparing structures (substituents) attached to the base atom, the surrounding structure defined by radius r is referenced. If the structures being compared match, including the surrounding structure, the compared structures are determined to be subject to deletion in preprocessing. If radius r is set to "r=0", it will be determined whether the structures attached to the base are identical without considering the surrounding structure (i.e., r=0 is equivalent to deleting identical structures). On the other hand, if radius r is set to "r=∞ (or a positive integer that includes the terminal nodes)", it will be determined whether the surrounding structures are exactly the same (i.e., r=∞ is equivalent to deleting parts with symmetry). When adopting the above method, the radius r for defining the reference range may be set by receiving a specification of radius r from the user in the input information receiving unit 101. The preprocessing unit 102 may also suggest (recommend) the value of radius r to the user using a machine learning model and / or user log data (history data, etc.), or it may be determined automatically.

[0050] Figure 5 illustrates a pretreatment that "masks" part of the molecular structure. Figure 5(a) shows the molecular structure of N-phenyl-2,7-dibutyl-10-(piperidin-1-yl)pyrene-4-carboxamide, which is exemplified as Example 2. This compound of Example 2 has butyl groups as indicated by substituents IIa and IIb. These butyl groups are substituents with limited reactivity and may not be able to be added to the precursor at the end of the sequential synthesis (in other words, at the beginning of the retrosynthesis). If the butyl group bond is cleaved at the beginning of the retrosynthesis analysis of Example 2, the resulting synthetic route, even if theoretically feasible, will be irreproducible (unrealistic). To design an appropriate synthetic route, the reactivity of substituents must be considered.

[0051] Substituents with restricted reactivity, such as the butyl group in Example 2, include substituents with constraints due to bonding properties, such as substituents with strong covalent bonds or substituents with resonance-stabilized electronic structures (e.g., phenyl group, cyclohexyl group, trifluoromethyl group, etc.); substituents with constraints due to steric factors that cause access hindrance (steric hindrance), such as isopropyl group, tert-butyl group, mesityl group, etc.; substituents with constraints due to electronic factors, such as substituents that are inert to electrophilic attack or substituents that are inert to nucleophilic attack (e.g., nitro group, sulfonyl group, trimethylsilyl group, etc.); and substituents with constraints due to physicochemical properties, such as substituents that are difficult to solvate or substituents with limited intermolecular interactions (e.g., perfluoroalkyl groups, crown ether rings, biphenyl group, etc.).

[0052] In addition to substituents whose reactivity is restricted as described above, there are also cases where a specific substituent (functional group) is temporarily protected during the synthesis process to restrict its reaction, for purposes such as protecting it from a predetermined chemical reaction or achieving a regioselective reaction (so-called protecting group). In this embodiment, substituents with restricted reactivity, substituents to be protected, etc., are collectively referred to as "reaction-restricting substituents." In the retrosynthetic analysis of this embodiment, reaction-restricting substituents are defined as "substituents that temporarily restrict bond cleavage."

[0053] The preprocessing unit 102 may simplify part of the molecular structure and set up a provisional structure compound by masking reaction-constraining substituents that temporarily restrict bond cleavage among the substituents included in the molecular structure of the base compound. "Masking reaction-constraining substituents" means virtually covering a predetermined substituent with a mask, but it also means temporarily excluding the masked area from the chemical reaction search in retrosynthesis analysis so that it is not subject to bond cleavage.

[0054] The parts to be masked, i.e., reaction-constraining substituents, may be specified by the user. In this case, the input information receiving unit 101 accepts the user's specification of substituents to be set as reaction-constraining substituents. Such specification of reaction-constraining substituents by the user may be performed when the target compound is entered, or it may be performed in the middle of the synthesis route search. In the latter case, for example, if the candidate synthesis route branches into multiple parts, or if no candidate chemical reaction for the next synthesis step is found, the user may be notified of an alert, and the specification of reaction-constraining substituents may be accepted triggered by the alert. The preprocessing unit 102 identifies the reaction-constraining substituents specified by the user from the molecular structure of the base compound and sets a provisional structure compound with the reaction-constraining substituents masked.

[0055] The preprocessing unit 102 may automatically identify reaction-constraining substituents or their candidates, rather than being limited to a method specified by the user. In this case, the preprocessing unit 102 may refer to the aforementioned specific structure DB124 and, based on a database of candidate substructures for reaction-constraining substituents, automatically identify reaction-constraining substituents from among the molecular structures of the base compound and set up a provisional structure compound with the reaction-constraining substituents masked. Alternatively, the preprocessing unit 102 may identify candidate substituents to be used as reaction-constraining substituents from among the molecular structures of the base compound based on a database of candidate substructures for reaction-constraining substituents and present (suggest) them to the user. The preprocessing unit 102 may then set up a provisional structure compound by having the user, upon receiving the suggestion, select a substituent to be masked as a reaction-constraining substituent from among the presented candidates.

[0056] Furthermore, the database of candidate substructures for reaction-constraining substituents may, as mentioned above, be constructed based on past retrosynthetic analysis results (i.e., historical data showing substituents specified by the user as reaction-constraining substituents in past analyses), or it may be constructed based on expert knowledge, or it may be a database that reflects data from both.

[0057] In Example 2 of Figure 5, the butyl groups of substituents IIa and IIb shown in (a) are identified as reaction-constraining substituents, and a provisional structure compound is set up with these butyl groups masked (Figure 5(b)). In addition, in the compound of Example 2, substituents IIa and IIb also fall under the category of "substituents with the same structure (substituents in symmetrical regions)." In such cases, the pretreatment unit 102 may set up a provisional structure compound by applying both a pretreatment to the base compound, which involves deleting some of the substituents with the same structure, and a pretreatment that masks the reaction-constraining substituents. For example, at least one substituent with the same structure may be deleted (either substituent IIa or substituent IIb may be deleted), and then the remaining substituent may be masked. Alternatively, both substituents IIa and IIb may be masked, and the synthetic pathway search by retrosynthesis analysis may be performed. At the timing when the mask is removed at a predetermined step, a pretreatment to delete at least one substituent with the same structure may be performed.

[0058] Figure 6, like Figure 5, illustrates a pretreatment that "masks" a part of the molecular structure. Figure 6(a) shows the molecular structure of 2-propyl-6-(piperidin-1-yl)-4-phenylbenzoate-naphthalene, which is exemplified as Example 3. In the compound of Example 3, the butyl group indicated by substituent III is identified as a reaction-constraining substituent, and a provisional structural compound (Figure 6(b)) is established by masking this butyl group.

[0059] The analysis unit 103 is a functional unit that performs retrosynthesis analysis to determine the synthesis route of the base compound. The analysis unit 103 may be implemented using a machine learning model for reaction prediction and a search algorithm for reaction route exploration. The architecture of the retrosynthesis analysis performed by the analysis unit 103 may employ known algorithms and is not necessarily limited.

[0060] If preprocessing has been performed, the analysis unit 103 takes a provisional structure compound, whose molecular structure has been partially simplified, as input information and performs a retrosynthesis analysis on the provisional structure compound. Specifically, the analysis unit 103 targets the molecular structure of the provisional structure compound, excluding structures simplified by "deletion," "masking," "conversion," etc., and searches for a chemical reaction to synthesize the molecular structure of the provisional structure compound by referring to the compound DB 121 and / or chemical reaction DB 122. Then, based on the various DBs and the retrosynthesis analysis architecture, it determines the chemical reaction to generate the molecular structure of the provisional structure compound and the intermediate compounds required for that chemical reaction.

[0061] When searching for chemical reactions, search conditions may be set to prioritize reactions with low synthesis costs, or reactions in which readily available starting compounds are output as intermediate compounds (other conditions such as prioritizing well-known chemical reactions or frequently performed chemical reactions may also be adopted). Furthermore, if the analysis unit 103 finds multiple candidate chemical reactions through the search, it may branch the synthesis route in parallel and continue retrosynthesis analysis for each branched route. Alternatively, if multiple candidate chemical reactions are found, the analysis unit 103 may present the candidates to the user and allow them to select the chemical reaction to adopt. In addition, if multiple candidate chemical reactions exist, the analysis unit 103 may adopt the reaction with the lowest synthesis cost, or it may refer to information such as inventory status recorded in the compound DB 121 and prioritize adopting a chemical reaction that can be handled with a compound that is in stock. If no pretreatment of the base compound has been performed before the retrosynthesis analysis in a predetermined step, the analysis unit 103 can simply accept the molecular structure of the base compound as input information and perform a normal retrosynthesis analysis.

[0062] After determining the chemical reaction in the first step as described above, the analysis unit 103 continues the retrosynthesis analysis using the intermediate compound required for the determined chemical reaction as the base compound. At this time, the analysis unit 103 may perform a verification of whether to restore the simplified substructure that was performed by the pretreatment. For example, the analysis unit 103 may perform a process to temporarily remove the mask from the intermediate compound determined by the retrosynthesis analysis of the masked provisional structure compound, and verify whether the next synthesis route can be output from the intermediate compound with the molecular structure without the mask (whether the appropriate chemical reaction for the next step can be found).

[0063] If the analysis unit 103 determines that it can output the next synthesis route, it removes the mask and performs a retrosynthesis analysis on the intermediate compound after mask removal to determine the synthesis reaction (chemical reaction) of the intermediate compound. On the other hand, if it determines that it cannot output the next synthesis route (or that removing the mask is not appropriate), the analysis unit 103 reapplies the mask to the intermediate compound, performs a retrosynthesis analysis, and determines the synthesis reaction of the intermediate compound. The area to which the mask is reapplied is the same area to which the mask was applied in the previous step.

[0064] Furthermore, after determining the chemical reactions in each step, the analysis unit 103 may determine whether the ongoing retrosynthetic analysis satisfies the limiting conditions (i.e., whether the upper limit of the search time, the upper limit of the search steps, or the upper limit of the synthesis cost has been reached) and determine whether to continue the retrosynthetic analysis. If the ongoing analysis is within the limits, the retrosynthetic analysis will continue. If the ongoing analysis falls outside the limits (i.e., the upper limit of the search time, the upper limit of the search steps, or the upper limit of the synthesis cost has been reached), the process may be stopped as a search failure. If, within the limits, all intermediate compounds of the determined chemical reactions become registered compounds, the analysis unit 103 determines that the analysis was a success and terminates the analysis for the target compound. The analysis unit 103 may set registered compounds, such as commercially available drugs or user-registered compounds, as the goal for the synthesis route search, or it may refer to the inventory status of the compound DB 121 and set only registered compounds with available inventory as the goal for terminating the synthesis route search.

[0065] For example, Figures 8, 10, and 12 are examples of the retrosynthesis analysis results of Examples 1-3 performed by the analysis unit 103, and Figures 7, 9, and 11 are the corresponding comparative example data. The analysis results for each comparative example and example will be described below.

[0066] Comparative Example 1 Figure 7 shows the results of retrosynthesis analysis performed on the compound shown in Figure 4(a) without any pretreatment (Comparative Example 1). As shown in Figure 7, when the compound in Figure 4(a) with substituents (symmetrical sites) of the same structure was not pretreated, the bonds were finely cleaved at the ends of the molecular structure, and the synthetic route did not converge easily, exceeding the limiting number of steps and resulting in the failure of the synthetic route search. Note that Figure 7 only shows the output results up to step 2, but in reality, the search for chemical reactions continued beyond step 2, and the search results from step 2 onwards are omitted from the illustration.

[0067] Example 1 On the other hand, Figure 8 shows the results of retrosynthesis analysis performed on the compound shown in Figure 4(a) after pretreatment by removing some of the substituents (symmetrical sites) of the same structure (Example 1). Specifically, in Example 1, one of the two hydroxyphenyl groups included in the molecular structure was removed to set up a provisional structure compound. When this provisional structure compound was input into the retrosynthesis analysis algorithm and the analysis was performed, as shown in Figure 8, both intermediate compounds converged to the registered compound in step 2, and a more efficient synthesis route than Comparative Example 1 was output.

[0068] Comparative Example 2 Figure 9 shows the results of retrosynthesis analysis performed on the compound shown in Figure 5(a) without any pretreatment (Comparative Example 2). As shown in Figure 9, in Comparative Example 2, the synthetic pathway was searched for in which bonds were finely cleaved from the ends of the molecular structure to avoid the butyl group. As a result, the synthetic pathway did not converge easily, and the number of steps exceeded the limit condition, resulting in the failure of the synthetic pathway search. Note that the search results from step 3 onwards are omitted from Figure 9.

[0069] Example 2 On the other hand, Figure 10 shows the results of retrosynthesis analysis performed on the compound shown in Figure 5(a) after performing a pretreatment to simplify the molecular structure by masking part of it (Example 2). As shown in Figure 10, in Example 2, two butyl groups included in the molecular structure were designated as reaction-constraining substituents, and a provisional structure compound was set up with these butyl groups masked. This masked provisional structure compound was input into a machine learning model and retrosynthesis analysis was performed. In the case of Example 2, after determining the chemical reaction in Step 1, the analysis unit 103 temporarily removed the mask and verified whether the next synthesis reaction could be output from the intermediate compound in Step 1 with the mask removed. As a result, the chemical reaction in Step 2 was determined after reapplying the mask. After Step 2, a similar verification was performed, and the chemical reaction in Step 3 was determined after reapplying the mask.

[0070] After step 3, the mask is removed, and the pretreatment unit 102 pretreatments the intermediate compound from step 3 by removing one of the two substituents (butyl groups) of the same structure. The analysis unit 103 then determines the chemical reaction in step 4 by retrosynthesis analysis of the provisional structure compound after pretreatment. Note that the removal of substituents of the same structure may be performed at the pretreatment stage of the initial target compound, rather than between steps 3 and 4. In Example 2, ultimately, in step 5, a synthesis reaction using two registered compounds was output, and an efficient synthesis route was output with fewer steps than in Comparative Example 2.

[0071] Comparative Example 3 Figure 11 shows the results of retrosynthesis analysis performed on the compound shown in Figure 6(a) without any pretreatment (Comparative Example 3). In Comparative Example 3, as in Comparative Example 2, a synthetic route that finely cleaves bonds was selected, and the synthetic route did not converge, exceeding the limit condition in terms of the number of steps, resulting in the failure of the synthetic route search. In Figure 11, the chemical reaction search from step 5 onwards is omitted from the illustration.

[0072] Example 3 On the other hand, Figure 12 shows the results of retrosynthesis analysis performed on the compound shown in Figure 6(a) after performing a pretreatment to simplify the molecular structure by masking a part of it (Example 3). As shown in Figure 12, in Example 3, the butyl group included in the molecular structure was designated as a reaction-constraining substituent, and a provisional structure compound was set up with this butyl group masked. In Example 3, the mask was removed after step 5, and in step 6, a synthesis reaction converging to two registered compounds was output. In Example 3, as with the other examples, a synthesis route with fewer steps was output more efficiently than in Comparative Example 3.

[0073] As shown in Examples 1-3, the retrosynthesis analysis system of this embodiment allowed for the establishment of a more efficient synthesis route than those shown in the comparative examples by simplifying (deleting, masking, transforming, etc.) a part of the molecular structure of the base compound as a pretreatment for retrosynthesis analysis.

[0074] The display control unit 104 executes a process to output the synthesis route of the target compound determined by the analysis unit 103 to the display unit 22 of the user terminal 2. The synthesis route output by the analysis unit 103 may be output in the form of a phylogenetic tree (e.g., Figures 8, 10, and 12) showing the chemical reactions of the series of steps determined by the retrosynthetic analysis. The display control unit 104 may output the synthesis route determined by the retrosynthetic analysis to the display unit 22 after the retrosynthetic analysis is completed up to the final step (after the search has converged), or it may sequentially update the phylogenetic tree after the chemical reactions at each step are determined and output the analysis results.

[0075] Although Figures 8, 10, and 12 illustrate analysis results without branching of the synthesis pathway, if multiple suitable synthesis reaction candidates are found during the steps, the synthesis pathway may be branched and the analysis results output accordingly. In such cases, the display control unit 104 may notify the user of the multiple branched synthesis reaction candidates and accept a selection operation from the user on the UI (User Interface) presented on the display unit 22 regarding which pathway to continue the chemical reaction search for.

[0076] If pretreatment is performed in the initial stage of retrosynthesis and / or during the synthesis pathway, the display control unit 104 may display the simplified parts due to pretreatment on the synthesis pathway output to the display unit 22 in a visually identifiable manner. The method of displaying the pretreated parts is not particularly limited; for example, methods such as changing the color of the target part, highlighting the target part, or showing the steps of the pretreatment (transition from base compound to provisional structure compound) may be employed. Furthermore, since pretreatment itself is a step that is not performed in the actual synthesis reaction, whether or not to display the target part (step) of pretreatment on the synthesis pathway may be switchable according to the user's selection.

[0077] When the user specifies reaction-constraining substituents during input of the target compound or during a step in retrosynthesis analysis, the display control unit 104 may display the molecular structure (e.g., structural formula) of the base compound on the UI displayed on the display unit 22 and allow the user to select the substituent to be set as the reaction-constraining substituent via the UI screen displaying the molecular structure. Alternatively, when the preprocessing unit 102 automatically sets the reaction-constraining substituents, the display control unit 104 may display the reaction-constraining substituents set by the preprocessing unit 102 along with the molecular structure of the base compound on the UI displayed on the display unit 22 and allow the user to choose whether or not to accept the setting of the reaction-constraining substituents identified by the preprocessing unit 102 via the UI screen. Even when the preprocessing unit 102 automatically suggests (suggests) candidate substituents to be used as reaction-constraining substituents, the display control unit 104 may display the candidate substituents on the UI screen of the display unit 22 and allow the user to choose whether or not to set the reaction-constraining substituents (whether or not to apply a mask), and which of the multiple candidate substituents to use as the reaction-constraining substituents.

[0078] The management server 1 may have functions not illustrated in Figure 3. For example, the management server 1 may have a function unit (learning unit) that executes prediction and / or search algorithms using a machine learning model for retrosynthesis analysis that realizes the functions of the analysis unit 103. This learning unit may have a function to additionally learn the contents of various databases that have been updated in accordance with updates to various databases such as the compound DB 121 and the chemical reaction DB 122. The learning unit may also use the retrosynthesis analysis result data (data including information indicating the success / failure of the search) stored in the analysis information storage unit 123 to perform additional learning of the machine learning model and / or modification of the search algorithm. In addition to the function unit that executes machine learning, the management server 1 may also have a function unit that constructs learning data used for machine learning.

[0079] Some or all of the functions of the management server 1 described above may be implemented by the processor of the user terminal 2.

[0080] <Functions of User Terminal 2> As shown in Figure 3, the user terminal 2 may include, for example, an input unit 21, a display unit 22, a transmitting / receiving unit 23, etc. The input unit 21 is composed of, for example, information input devices such as a keyboard, mouse, or touch panel, and directly accepts various input operations by the user, such as specifying the target compound, specifying reaction constraint substituents, and specifying limiting conditions during the search. The display unit 22 is composed of information display devices such as a display or touch panel, and displays a UI that accepts input operations, a UI that presents analysis results, and other data transmitted from the management server 1 on the screen. The input unit 21 and the display unit 22 may be configured as an integral unit (e.g., a touch panel). The transmitting / receiving unit 23 is a communication interface that communicates between the user terminal 2 and the management server 1. The transmitting / receiving unit 23 has the function of transmitting user operation information input via the user terminal 2 to the management server 1, and the function of receiving various processing result information, such as analysis results output by the management server 1, from the management server 1.

[0081] <An example of a retrosynthesis analysis method> Next, the retrosynthesis analysis method according to this embodiment will be described with reference to the flowcharts illustrated in Figures 13 and 14.

[0082] When performing retrosynthesis analysis, the input information receiving unit 101 first receives input from the user via the user terminal 2 regarding the target compound to be analyzed (step SQ101 in Figure 13). At this time, the input information receiving unit 101 may also receive from the user the specification of limitations information during the search, such as an upper limit on the search time, an upper limit on the number of search steps, and an upper limit on the synthesis cost.

[0083] Next, the preprocessing unit 102 determines whether or not a substructure to be preprocessed (for example, multiple substituents of the same structure, a symmetrical site, or a reaction-constraining substituent, etc.) exists within the molecular structure of the target compound input by the user (step SQ102). If it is determined that a substructure to be preprocessed exists within the molecular structure of the target compound ("Yes" in step SQ102), the preprocessing unit 102 performs preprocessing to simplify a part of the molecular structure of the target compound (base compound) and sets up a provisional structure compound with a simplified part of the structure (step SQ103). "Simplified" in this preprocessing includes, for example, "deleting" a part of the molecular structure, "masking" a part of the molecular structure, or "converting" the molecular structure to a state in which a part of the molecular structure has been deleted or to a simpler structure. After setting up the provisional structure compound, the analysis unit 103 uses the provisional structure compound as input information to perform retrosynthesis analysis on the provisional structure compound and determines the chemical reaction that produces the target compound (step SQ104).

[0084] If, in step SQ102, the preprocessing unit 102 determines that there is no substructure to be preprocessed within the molecular structure of the target compound ("No" in step SQ102), the process proceeds to step SQ104 without performing preprocessing, and the retrosynthesis analysis is performed using the molecular structure of the target compound as input information.

[0085] In the retrosynthetic analysis of step SQ104, the analysis unit 103 refers to various databases such as compound DB121 and chemical reaction DB122 to determine the locations where bonds are cleaved within the molecular structure of the input information, the intermediate compounds produced by the bond cleavage, and the chemical reactions. The function of the analysis unit 103 in step SQ104 may be performed as a machine learning model or search algorithm for retrosynthetic analysis, and the architecture of the retrosynthetic analysis is not particularly limited.

[0086] If any of the intermediate compounds determined by the retrosynthetic analysis in step SQ104 are not registered compounds ("Yes" in step SQ105), the analysis unit 103 continues the retrosynthetic analysis step using the intermediate compound as a base compound to be input into the machine learning model and search algorithm. At this time, the analysis unit 103 determines whether the ongoing analysis is within the range of predetermined limiting conditions (whether the upper limit of search time has been reached, whether the upper limit of the number of search steps has been reached, whether the upper limit of synthesis cost has been reached, etc.) (step SQ106). If it is determined that the analysis can be continued within the range of limiting conditions, the analysis unit 103 determines whether or not to perform pretreatment on the intermediate compound determined in step SQ104 (step SQ107).

[0087] If the molecular structure of the intermediate compound contains a substructure to be pretreated, and the analysis unit 103 determines that pretreatment should be performed to determine the next synthetic route ("Yes" in step SQ107), the process proceeds to step SQ103. The pretreatment unit 102 then sets a provisional structure compound with a simplified part of the molecular structure of the intermediate compound and performs a retrosynthesis analysis on this provisional structure compound (step SQ104). On the other hand, if the analysis unit 103 determines that pretreatment should not be performed on the intermediate compound ("No" in step SQ107), the process proceeds to step SQ104, and the analysis unit 103 performs a retrosynthesis analysis on the molecular structure of the intermediate compound that has not undergone pretreatment.

[0088] The cycle of steps SQ103 to SQ107 described above is repeated, and the one-step retrosynthesis analysis continues until all intermediate compounds in the analysis converge to registered compounds that are available as starting materials. In step SQ105, if all intermediate compounds determined by the retrosynthesis analysis are registered compounds, the process moves to step SQ108, where the analysis unit 103 completes the chemical reaction search and outputs the series of synthesis routes, which are the analysis results, to the user terminal 2. On the other hand, if the analysis reaches a predetermined upper limit set in the limiting conditions (upper limit of search time, upper limit of number of search steps, upper limit of synthesis cost, etc.) during the repetition of the cycle of steps SQ103 to SQ107 ("No" in step SQ106), the analysis unit 103 determines that the synthesis route search has failed and terminates the analysis.

[0089] Figure 14 shows an example flowchart for performing retrosynthesis analysis after receiving the user's specification of reaction-constraining substituents. The flowchart in Figure 14 concretizes steps SQ101, SQ102, SQ103, and SQ107 in Figure 13, and the other steps are the same as in the flowchart in Figure 13 and are denoted by the same reference numerals.

[0090] As shown in step SQ1011 of Figure 14, the input information receiving unit 101 accepts input of the target compound from the user and may also accept the user's specification of reaction-constraining substituents that temporarily restrict bond cleavage, such as substituents with reactivity constraints (substituents with low reactivity). Next, the preprocessing unit 102 determines whether or not there are multiple substituents with the same structure in the molecular structure of the target compound input by the user (in other words, whether or not there are symmetrical sites) (step SQ1021 of Figure 14). If there are multiple substituents with the same structure ("Yes" in step SQ1021), the preprocessing unit 102 simplifies part of the molecular structure of the target compound by deleting at least one of the substituents with the same structure (step SQ1031). On the other hand, if there are no substituents with the same structure in the molecular structure of the target compound ("No" in step SQ1021), the process proceeds to step SQ1032.

[0091] In step SQ1032, the preprocessing unit 102 sets up a provisional structure compound by masking the reaction-constraining substituents specified by the user. Then, the analysis unit 103 performs a retrosynthesis analysis on the provisional structure compound set up in steps SQ1031 to SQ1032 to determine the synthesis reaction of the base compound (step SQ104). After one chemical reaction is determined in a single step, if the analysis is to continue in the same flow as in Figure 13, in step SQ1071, the analysis unit 103 temporarily removes the mask from the intermediate compound determined in step SQ104 and verifies whether the next synthesis route can be output (step SQ1072).

[0092] If the next synthesis route can be output ("Yes" in step SQ1072), the analysis unit 103 removes the mask set in the pretreatment and proceeds to step SQ104, where it performs retrosynthesis analysis on the intermediate compound after the mask has been removed. On the other hand, if it is determined that the next synthesis route cannot be output, the process proceeds to step SQ1032, where the pretreatment unit 102 reapplies the mask to the reaction-constraining substituents in the molecular structure of the intermediate compound. Then, retrosynthesis analysis is performed on the intermediate compound with the mask applied.

[0093] In the flow shown in Figure 14, the cycle from step SQ1031 to step SQ1072 is repeated, and when the synthesis route converges to the registered compound (step SQ105 "No"), the search for chemical reactions is terminated and the series of synthesis routes are output.

[0094] The retrosynthetic analysis method flow shown in Figures 13 and 14 is merely illustrative, and the retrosynthetic analysis method of this embodiment is not limited to the examples shown in Figures 13 and 14. For example, the preprocessing unit 102 may automatically set reaction-constraining substituents from among the molecular structures of the base compound based on a database (a database of candidate substructures for reaction-constraining substituents) constructed based on past retrosynthetic analysis results and / or expert knowledge, or it may suggest candidate reaction-constraining substituents. Also, some of the steps shown in Figures 13 and 14 may be in a different order. For example, the user's specification of reaction-constraining substituents may be performed in an intermediate step of the retrosynthetic analysis (SQ107) rather than in the first step (step SQ1011). Furthermore, in Figure 14, a preprocessing step is performed to delete some substituents of the same structure before masking the reaction-constraining substituents, but the deletion of some substituents of the same structure may be performed at the stage when the mask is removed from the intermediate compound (SQ107).

[0095] In the retrosynthesis analysis system according to this embodiment, as described above, the target compound or intermediate compound is used as a base compound, and after performing a pretreatment to simplify a part of the molecular structure of the base compound, retrosynthesis analysis is performed, thereby enabling efficient exploration of the synthesis route of the target compound. As a result, feasible and efficient synthesis routes can be devised for compounds having predetermined structures, such as molecular structures with multiple substituents of the same structure, molecular structures with symmetry, and molecular structures containing reaction-constraining substituents.

[0096] The embodiments described above are merely illustrative to facilitate understanding of this disclosure and are not intended to limit it. This disclosure may be modified and improved without departing from its intent, and its equivalents are included.

[0097] <Variation> For example, pretreatment for retrosynthesis analysis is not limited to simplifying a part of the molecular structure. The pretreatment unit 102 may perform pretreatment to add a predetermined substructure (substituent such as a protecting group) to a part of the molecular structure of the base compound. In the synthesis pathway of organic compounds, specific substituents are sometimes temporarily modified (so-called protecting groups) to control the reactivity of those substituents and protect them from chemical reactions during synthesis. Such protecting groups are not limited to protecting from chemical reactions; they may also be formed to achieve regioselective reactions or to control solubility in a given solvent. After the protecting group has served its purpose in the synthesis pathway, it is removed by a deprotection reaction or the like.

[0098] While the formation of a protecting group is an addition reaction, retrosynthetic analysis is a method that sequentially determines the chemical reactions at the reaction site and the reacting compounds, such as bond cleavage, to explore the synthesis route. Therefore, in conventional retrosynthetic analysis, it can be difficult to form and select an appropriate protecting group. In the retrosynthetic analysis system of this disclosure, the system may accept a designation from the user of a protecting group to be added to a part of the molecular structure of a base compound, and then perform retrosynthetic analysis by masking the substituent on which the protecting group is formed. In this case, the preprocessing unit 102 may accept the designation of the protecting group during the retrosynthetic analysis, or it may accept an instruction to remove the protecting group at a predetermined timing in the retrosynthetic analysis. When accepting the designation of a protecting group from the user, it is not limited to accepting the designation of a single structure to be used as a protecting group, but may also accept the designation of multiple candidate structures and / or selection from multiple candidates. For example, in addition to accepting the addition of an F group, the system may accept the designation of a group of candidates including other halogen atoms such as Cl, Br, and I, and the preprocessing unit 102 may appropriately determine the structure to be added to the base compound from among the accepted candidate group. Furthermore, not limited to cases where the user specifies a protecting group, the preprocessing unit 102 may automatically set or suggest the formation and removal of protecting groups by referring to a database based on past retrosynthesis analysis results and / or expert knowledge.

[0099] In this disclosure, the pretreatment unit 102 may perform a combination of multiple pretreatments on the base compound. For example, when combining the removal of substituents of the same structure with a masking treatment, various combinations are possible, such as (1) performing the masking treatment after removing substituents of the same structure, (2) performing the masking treatment and then removing substituents of the same structure, or (3) performing the masking treatment and the removal of substituents of the same structure in parallel. The order of the pretreatment combinations may be set as appropriate depending on the structure of the target compound, user specifications, etc.

[0100] The retrosynthesis analysis system described herein may be implemented as a single device, or it may be implemented by multiple devices (e.g., cloud servers) that are partially or entirely connected by a network. For example, the functions of the control unit (processor 10) and storage 12 of the management server 1 may be implemented by different servers that are connected to each other by a network.

[0101] Furthermore, the series of processes performed by the retrosynthesis analysis system described herein may be implemented using software, hardware, or a combination of software and hardware. It is also possible to create a computer program to implement each function of the management server 1 according to this embodiment and implement it on a PC or the like. A computer-readable recording medium on which such a computer program is stored can also be provided. Examples of recording media include magnetic disks, optical disks, magneto-optical disks, flash memory, etc. In addition, the above-mentioned computer program may be distributed, for example, via a network, without using a recording medium.

[0102] Furthermore, the effects described herein are merely descriptive or illustrative and not limiting. In other words, the technology relating to this disclosure may produce other effects that will be apparent to those skilled in the art from the description herein, in addition to or in lieu of the effects described herein.

[0103] The retrosynthesis analysis system, retrosynthesis analysis apparatus, retrosynthesis analysis method, and retrosynthesis analysis program of this disclosure may have the following configurations. [Item 1] An input information receiving unit that receives the target compound to be analyzed, A preprocessing unit that uses the target compound and / or an intermediate compound of the target compound as a base compound, performs a preprocessing to simplify a part of the molecular structure of the base compound, and sets up a provisional structure compound, A retrosynthesis analysis system comprising: an analysis unit that performs retrosynthesis analysis on the provisional structure compound and determines the synthesis pathway of the base compound. [Item 2] The aforementioned pre-processing unit, Determine whether or not multiple substituents with the same structure exist within the molecular structure of the base compound. The retrosynthesis analysis system described in item 1, which, when it is determined that multiple substituents with the same structure exist, sets the provisional structure compound by deleting at least one of the multiple substituents with the same structure. [Item 3] The aforementioned pre-processing unit, From the molecular structure of the aforementioned base compound, a symmetrical region is identified, The retrosynthesis analysis system according to item 1, wherein the provisional structure compound is set by removing at least one substituent included in the symmetrical region. [Item 4] The aforementioned pre-processing unit, A retrosynthesis analysis system according to any one of items 1 to 3, which simplifies a part of the molecular structure by masking reaction-constraining substituents that temporarily restrict bond cleavage among the substituents included in the molecular structure of the base compound, thereby setting up the provisional structure compound. [Item 5] The input information receiving unit receives from the user the specification of the substituent to be set as the reaction constraint substituent. The retrosynthesis analysis system according to item 4, wherein the preprocessing unit identifies the reaction-constraining substituent specified by the user from among the molecular structures of the base compound and sets up the provisional structure compound in which the reaction-constraining substituent is masked. [Item 6] The retrosynthesis analysis system according to item 4, wherein the preprocessing unit automatically identifies the reaction-constraining substituent from among the molecular structures of the base compound based on a database of candidate substructures of the reaction-constraining substituent, and sets up the provisional structure compound in which the reaction-constraining substituent is masked. [Item 7] The aforementioned analysis unit, From the intermediate compound determined by retrosynthesis analysis of the masked provisional structure compound, a process is performed to temporarily remove the mask. A retrosynthesis analysis system as described in item 4, which verifies whether the next synthesis route can be output from the intermediate compound with the aforementioned mask removed. [Item 8] The retrosynthesis analysis system described in item 7, wherein the analysis unit, when it determines that it can output the next synthesis route, performs a retrosynthesis analysis on the intermediate compound with the mask removed and determines the synthesis reaction of the intermediate compound. [Item 9] The retrosynthesis analysis system according to item 7, wherein the analysis unit determines that it cannot output the next synthesis route, reapplies the mask to the intermediate compound to perform retrosynthesis analysis and determines the synthesis reaction of the intermediate compound. [Item 10] The system further comprises a display control unit that outputs the synthesis path determined by the analysis unit to a display unit, The display control unit presents the parts simplified by the preprocessing on the synthesis path output to the display unit, in the retrosynthesis analysis system according to any one of items 1 to 3. [Item 11] An input information receiving unit that receives the target compound to be analyzed, A preprocessing unit that uses the target compound and / or an intermediate compound of the target compound as a base compound, performs a preprocessing to simplify a part of the molecular structure of the base compound, and sets up a provisional structure compound, A retrosynthesis analysis apparatus comprising: an analysis unit that performs retrosynthesis analysis on the provisional structure compound and determines the synthesis pathway of the base compound. [Item 12] The system accepts the target compound to be analyzed, Using the aforementioned target compound and / or an intermediate compound of the aforementioned target compound as a base compound, a pretreatment is performed to simplify a part of the molecular structure of the base compound, and a pretreatment is performed to establish a provisional structure compound. A retrosynthetic analysis method in which a computer performs retrosynthetic analysis on the provisional structure compound to determine the synthesis pathway of the base compound. [Item 13] The system accepts the target compound to be analyzed, Using the aforementioned target compound and / or an intermediate compound of the aforementioned target compound as a base compound, a pretreatment is performed to simplify a part of the molecular structure of the base compound, and a pretreatment is performed to establish a provisional structure compound. A retrosynthetic analysis program for causing a computer to perform retrosynthetic analysis on the aforementioned hypothetical structure compound and determine the synthesis route of the base compound. [Explanation of Symbols]

[0104] 1. Management Server 2 User terminals 101 Input Information Reception Section 102 Pre-processing section 103 Analysis Department 104 Display Control Unit

Claims

1. An input information receiving unit that receives the target compound to be analyzed, A preprocessing unit that uses the target compound and / or an intermediate compound of the target compound as a base compound, performs a preprocessing to simplify a part of the molecular structure of the base compound, and sets up a provisional structure compound, A retrosynthesis analysis system comprising: an analysis unit that performs retrosynthesis analysis on the provisional structure compound and determines the synthesis pathway of the base compound.

2. The aforementioned pre-processing unit, Determine whether or not multiple substituents with the same structure exist within the molecular structure of the base compound. The retrosynthesis analysis system according to claim 1, wherein, when it is determined that there are multiple substituents with the same structure, the provisional structure compound is set by deleting at least one of the multiple substituents with the same structure.

3. The aforementioned pre-processing unit, From the molecular structure of the aforementioned base compound, a symmetrical region is identified, The retrosynthesis analysis system according to claim 1, wherein the provisional structure compound is set by removing at least one substituent included in the symmetrical region.

4. The aforementioned pre-processing unit, A retrosynthesis analysis system according to any one of claims 1 to 3, wherein a part of the molecular structure is simplified by masking reaction-constraining substituents that temporarily restrict bond cleavage among the substituents included in the molecular structure of the base compound, thereby setting up the provisional structure compound.

5. The input information receiving unit receives from the user the specification of the substituent to be set as the reaction constraint substituent. The retrosynthesis analysis system according to claim 4, wherein the preprocessing unit identifies the reaction-constraining substituent specified by the user from among the molecular structures of the base compound and sets up the provisional structure compound in which the reaction-constraining substituent is masked.

6. The retrosynthesis analysis system according to claim 4, wherein the preprocessing unit automatically identifies the reaction-constraining substituent from among the molecular structures of the base compound based on a database of candidate substructures of the reaction-constraining substituent, and sets up the provisional structure compound in which the reaction-constraining substituent is masked.

7. The aforementioned analysis unit, From the intermediate compound determined by retrosynthesis analysis of the masked provisional structure compound, a process is performed to temporarily remove the mask. The retrosynthesis analysis system according to claim 4, which verifies whether the next synthesis route can be output from the intermediate compound with the mask removed.

8. The retrosynthesis analysis system according to claim 7, wherein the analysis unit, when it determines that it can output the next synthesis route, performs a retrosynthesis analysis on the intermediate compound with the mask removed and determines the synthesis reaction of the intermediate compound.

9. The retrosynthesis analysis system according to claim 7, wherein the analysis unit determines that it cannot output the next synthesis route, it reapplies the mask to the intermediate compound to perform retrosynthesis analysis and determines the synthesis reaction of the intermediate compound.

10. The system further comprises a display control unit that outputs the synthesis path determined by the analysis unit to a display unit, The inverse synthesis analysis system according to any one of claims 1 to 3, wherein the display control unit presents the parts simplified by the preprocessing on the synthesis path output to the display unit.

11. An input information receiving unit that receives the target compound to be analyzed, A preprocessing unit that uses the target compound and / or an intermediate compound of the target compound as a base compound, performs a preprocessing to simplify a part of the molecular structure of the base compound, and sets up a provisional structure compound, A retrosynthesis analysis apparatus comprising: an analysis unit that performs retrosynthesis analysis on the provisional structure compound and determines the synthesis pathway of the base compound.

12. The system accepts the target compound to be analyzed, Using the aforementioned target compound and / or an intermediate compound of the aforementioned target compound as a base compound, a pretreatment is performed to simplify a part of the molecular structure of the base compound, and a pretreatment is performed to establish a provisional structure compound. A retrosynthetic analysis method in which a computer performs retrosynthetic analysis on the provisional structure compound to determine the synthesis pathway of the base compound.

13. The system accepts the target compound to be analyzed, Using the aforementioned target compound and / or an intermediate compound of the aforementioned target compound as a base compound, a pretreatment is performed to simplify a part of the molecular structure of the base compound, and a pretreatment is performed to establish a provisional structure compound. A retrosynthetic analysis program for causing a computer to perform retrosynthetic analysis on the aforementioned hypothetical structure compound and determine the synthesis route of the base compound.

Images