Methods for establishing a skeleton transformation template database, molecular generation methods and apparatus

Abstract

This application relates to a method, a molecule generation method, and an apparatus for establishing a skeleton transformation template database. The method includes obtaining skeleton transformation molecules of a sample lead compound; obtaining reaction combinations composed of any two different molecules based on the sample lead compound and the skeleton transformation molecules, wherein the reaction combinations include reactants and products; obtaining chemical reaction templates corresponding to each of the reaction combinations, wherein the chemical reaction templates include reactant structural fragments and corresponding product structural fragments; and establishing a skeleton transformation template database using the chemical reaction templates. The solution provided in this application can obtain new molecules with greater potential for drug development and easier synthesis based on the skeleton transformation template database.

Description

Technical Field

[0001] This application relates to the field of skeleton transformation technology, and in particular to a method for establishing a skeleton transformation template database, a molecular generation method, and an apparatus. Background Technology

[0002] Skeleton transformation, a computer-aided method for finding active compounds with different skeletons, is a hot topic in medicinal chemistry research. Current skeleton transformation methods include pharmacophore-based, molecular shape-based, chemical fingerprint-based, protein structure-based molecular similarity-based, and machine learning-based methods. Skeleton transformation can help discover new molecules with ideal activities, facilitate the modification of complex natural product structures into more easily synthesized fragments, or improve pharmacological properties.

[0003] Although a large number of novel molecules can be generated through skeletal transformation, the synthesis of these new molecules is difficult and they have poor drug-like properties. This results in a large amount of computational resources being invested, but the number of molecules that ultimately meet the requirements for lead compound optimization is small. Summary of the Invention

[0004] To address or partially address the problems existing in related technologies, this application provides a method for establishing a scaffold transformation template database, a molecule generation method, and an apparatus, which can help obtain new molecules that are easier to synthesize and have potential drug-like properties.

[0005] The first aspect of this application provides a method for establishing a skeleton transformation template database, which includes:

[0006] Obtain the skeleton transformation molecule of the sample lead compound; based on the sample lead compound and the skeleton transformation molecule, obtain a reaction combination composed of any two different molecules, wherein the reaction combination includes reactants and products; obtain the chemical reaction template corresponding to each reaction combination, wherein the chemical reaction template includes reactant structural fragments and corresponding product structural fragments; and establish a skeleton transformation template database using each chemical reaction template.

[0007] In the method for establishing a skeleton transformation template database, the step of obtaining a reaction combination composed of any two different molecules based on the sample lead compound and the skeleton transformation molecule includes:

[0008] The lead compound and its corresponding skeleton transformation molecules of the same sample are represented by SMILES strings to generate corresponding string datasets; each pair of molecules in the same string dataset is associated according to a preset format to form corresponding reaction combinations, and a reaction combination dataset is obtained.

[0009] In the method for establishing a skeleton transformation template database, the step of obtaining the chemical reaction templates corresponding to each of the reaction combinations includes:

[0010] In the corresponding reaction combination, the atoms of the reactants and the atoms of the products are mapped and numbered to obtain a dataset of mapped numbers for reactants and products. Based on the dataset of mapped numbers, the atoms in the reactants whose chemical environment changes are determined. Taking the atoms in the reactants whose chemical environment changes as the reaction center, the reaction center and the structures within a preset radius from the reaction center are taken as reactant structural fragments, and the corresponding product structural fragments are obtained. Based on the reactant structural fragments and the product structural fragments, a corresponding chemical reaction template is generated.

[0011] In the method for establishing a skeleton transformation template database, the step of taking the reaction center and the structure within a preset radius from the reaction center as reactant structure fragments includes: if the reaction center and / or the structure within a preset radius from the reaction center contains atoms located in a first target group, then the reaction center, the structure within a preset radius from the reaction center, and the first target group are taken as reactant structure fragments.

[0012] In the method for establishing a skeleton transformation template database, the step of generating corresponding chemical reaction templates based on the reactant structural fragments and the product structural fragments includes:

[0013] The first target group in the reactant structural fragment is replaced with a first preset group to obtain a new reactant structural fragment; the new reactant structural fragment and the product structural fragment are used to generate a corresponding chemical reaction template.

[0014] In the method for establishing a skeleton transformation template database, the method further includes: if the product structure fragment contains a second target group, replacing the second target group with a second preset group to obtain a new product structure fragment;

[0015] The step of generating a corresponding chemical reaction template using the new reactant structural fragment and the new product structural fragment includes: generating a corresponding chemical reaction template using the new reactant structural fragment and the new product structural fragment.

[0016] In the method for establishing a skeleton transformation template database, the step of generating corresponding chemical reaction templates based on the reactant structural fragments and the product structural fragments includes:

[0017] If the product structure fragment contains a second target group, the second target group is replaced with a second preset group to obtain a new product structure fragment; using the reactant structure fragment and the new product structure fragment, a corresponding chemical reaction template is generated.

[0018] A second aspect of this application provides a skeleton transformation template database, which is constructed and obtained according to the skeleton transformation template database establishment method described in any of the above embodiments.

[0019] A third aspect of this application provides a method for generating a molecule, comprising:

[0020] Obtain the target molecule to be optimized; according to the above-mentioned skeleton transformation template database, obtain the chemical reaction template that matches the target molecule; according to each of the matched chemical reaction templates, generate the corresponding optimized molecule from the target molecule.

[0021] In the above-mentioned molecule generation method, the step of obtaining a chemical reaction template that matches the target molecule according to the skeleton transformation template database includes: matching the target molecule with each reactant structural fragment in the skeleton transformation template database to determine the corresponding substructure fragment to be replaced in the target molecule and the corresponding chemical reaction template.

[0022] In the above-described molecule generation method, generating corresponding optimized molecules from the target molecule according to each of the aforementioned chemical reaction templates includes: replacing the substructure fragments in the target molecule with corresponding product structure fragments according to the corresponding chemical reaction templates to generate corresponding optimized molecules.

[0023] A fourth aspect of this application provides an apparatus for establishing a skeleton transformation template database, comprising:

[0024] The first acquisition module is used to acquire the skeleton transformation molecule of the sample lead compound;

[0025] The combination generation module is used to obtain a reaction combination composed of any two different molecules based on the sample lead compound and skeleton transformation molecule, wherein the reaction combination includes reactants and products;

[0026] The template generation module is used to obtain the chemical reaction templates corresponding to each of the reaction combinations, wherein the chemical reaction templates include reactant structural fragments and corresponding product structural fragments;

[0027] The database construction module is used to establish a skeleton transformation template database using the aforementioned chemical reaction templates.

[0028] The fifth aspect of this application provides a molecular generation apparatus, comprising:

[0029] The second acquisition module is used to acquire the target molecule to be optimized;

[0030] The processing module is used to obtain a chemical reaction template that matches the target molecule based on the skeleton transformation template database mentioned above.

[0031] The molecule generation module is used to generate corresponding optimized molecules from the target molecule according to each of the matched chemical reaction templates.

[0032] The sixth aspect of this application provides an electronic device, comprising:

[0033] Processor; and

[0034] The memory stores executable code, which, when executed by the processor, causes the processor to perform the method for establishing a skeleton transformation template database or a molecular generation method as described above.

[0035] The seventh aspect of this application provides a computer-readable storage medium having executable code stored thereon, which, when executed by a processor of an electronic device, causes the processor to perform the method for establishing a skeleton transformation template database or a molecule generation method as described above.

[0036] The technical solution provided in this application may include the following beneficial effects:

[0037] By acquiring sample lead compounds and their skeletal transformation molecules, and by combining them in pairs to obtain multiple reaction combinations, chemical reaction templates are obtained on known molecular structures to form a template database. Based on the chemical reaction templates obtained empirically, the reliability of the optimized new molecules can be improved, which is conducive to increasing the success rate of obtaining active, easily synthesized, and highly druggable molecules.

[0038] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0039] The above and other objects, features and advantages of this application will become more apparent from the following description of exemplary embodiments of this application in conjunction with the accompanying drawings, wherein the same reference numerals generally represent the same components in the exemplary embodiments of this application.

[0040] Figure 1 This is a flowchart illustrating the method for establishing a skeleton transformation template database as shown in this application;

[0041] Figure 2 This is another flowchart illustrating the method for establishing a skeleton transformation template database shown in this application;

[0042] Figure 3 This is a molecular structure diagram of the sample lead compound and its skeleton transformation molecule as exemplified in this application;

[0043] Figure 4 It is based on Figure 3 A molecular structure diagram with a mapping number for one of the reaction combinations;

[0044] Figure 5 yes Figure 4 Chemical reaction templates extracted from reaction combinations in the process;

[0045] Figure 6 This is a schematic flowchart of the molecular generation method shown in this application;

[0046] Figure 7 This is a molecular structure diagram of the target molecule in the example of this application;

[0047] Figure 8 yes Figure 7 Multiple optimized molecules after backbone transformation of the target molecule;

[0048] Figure 9 This is a schematic diagram of the structure of the skeleton transformation template database creation device shown in this application;

[0049] Figure 10 This is another schematic diagram of the skeleton transformation template database creation device shown in this application;

[0050] Figure 11 This is a schematic diagram of the molecular generation device shown in this application;

[0051] Figure 12 This is a schematic diagram of the structure of the electronic device shown in this application. Detailed Implementation

[0052] Embodiments of this application will now be described in more detail with reference to the accompanying drawings. While embodiments of this application are shown in the drawings, it should be understood that this application may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to make this application more thorough and complete, and to fully convey the scope of this application to those skilled in the art.

[0053] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0054] It should be understood that although the terms "first," "third," "third," etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as third information, and similarly, third information may also be referred to as first information. Thus, a feature defined as "first" or "third" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0055] In related technologies, although current skeleton transformation methods can generate a large number of molecules with novel structures, the synthesis of new molecules is difficult and their drug-like properties are poor.

[0056] To address the aforementioned issues, this application provides a method for establishing a scaffold transformation template database, a molecule generation method, and an apparatus, which can generate new molecules that are easier to synthesize and have potential drug-like properties.

[0057] The technical solution of this application is described in detail below with reference to the accompanying drawings.

[0058] Figure 1 This is a flowchart illustrating the method for establishing a skeleton transformation template database as shown in this application.

[0059] See Figure 1 This application discloses a method for establishing a skeleton transformation template database, which includes:

[0060] S110, obtain the skeleton transformation molecule of the sample lead compound.

[0061] This can be achieved by collecting skeletal transformation examples of lead compounds from known sources, such as publicly available literature, and using these known lead compounds and their corresponding known active skeletal transformation molecules as sample data. In other words, these skeletal transformation molecules are optimized based on at least a portion of the structural fragments of the corresponding lead compounds, and their activity is better or similar to that of the corresponding sample lead compounds, thus providing good reference value.

[0062] S120: Based on the sample lead compound and skeleton transformation molecule, obtain a reaction combination consisting of any two different molecules, wherein the reaction combination includes reactants and products.

[0063] When there is more than one lead compound, each lead compound has its own skeleton transformation molecule. Based on this, for a single lead compound, two different molecules are arbitrarily selected from the lead compound and its corresponding skeleton transformation molecule to form a reaction combination. It can be understood that a single reaction combination contains only one reactant and one corresponding product; the reactant and product are distinct and both are selected from the lead compound and its skeleton transformation molecule.

[0064] Optionally, to obtain richer reaction rules from multiple dimensions, all sample lead compounds and their corresponding skeleton transformation molecules are sequentially paired to form corresponding reaction combinations. For example, assuming sample lead compound A has two skeleton transformation molecules B and C, pairing them according to a preset format can yield 6 reaction combinations, represented in a preset format, including A>>B, B>>A, A>>C, C>>A, B>>C, and C>>B. In the reaction combination "A>>B", A is the reactant, and B is the corresponding product, indicating that A generates (or converts) B. Optionally, the preset format can also be other formats, such as AB, A→B, A to B, A*B, etc., without limitation, by setting a preset format that is easy for computers to read, to clearly represent the reactants and products in each reaction combination.

[0065] S130, obtain the chemical reaction templates corresponding to each reaction combination, wherein the chemical reaction templates include reactant structural fragments and corresponding product structural fragments.

[0066] For each reaction combination, structural fragments in which the chemical environment changes in the reactants can be extracted as reactant structural fragments from the reaction formula in which reactants generate products. Then, product structural fragments corresponding to the reactant structural fragments can be extracted from the products, thereby obtaining a chemical reaction template composed of reactant structural fragments and corresponding product structural fragments.

[0067] Optionally, for the same reaction combination, there may be more than one structural segment in the reactants that undergoes a change in chemical environment. For structurally discontinuous segments, their corresponding product structural segments can be obtained separately, thereby generating their own independent chemical reaction templates. That is, the number of corresponding chemical reaction templates generated in the same reaction combination can be more than one.

[0068] S140, using various chemical reaction templates, establish a skeleton transformation template database.

[0069] The chemical reaction templates obtained through the above steps can be directly stored or processed, such as by converting to a unified format, data compression, and encryption. These processed templates are then centrally stored to obtain a corresponding skeleton transformation template database. This database can be updated based on newly added chemical reaction templates. Using these templates, skeleton replacement predictions for target molecules can be performed, transforming structural fragments in the target molecule into corresponding product structural fragments based on matching chemical reaction templates, thus obtaining new molecules. This design, based on chemical reaction templates generated from reliable sample data sources, makes the resulting new molecules more promising and easier to synthesize, increasing their likelihood of serving as lead compounds, thereby enhancing their drug-likeness, improving R&D efficiency, and saving R&D costs.

[0070] As can be seen from this example, the method for establishing the skeleton transformation template database in this application involves obtaining sample lead compounds and skeleton transformation molecules of sample lead compounds, and obtaining multiple reaction combinations by arranging them in pairs. Chemical reaction templates are obtained on known molecular structures to form a template database. Based on the chemical reaction templates obtained empirically, the reliability of the optimized new molecules can be improved, which is conducive to improving the success rate of obtaining active, easily synthesized, and highly druggable molecules.

[0071] Figure 2 This is a flowchart illustrating the method for establishing a skeleton transformation template database as shown in this application.

[0072] See Figure 2 This application discloses a method for establishing a skeleton transformation template database, which includes:

[0073] S210, obtain the sample lead compound and the corresponding skeleton transformation molecule.

[0074] This step can refer to the relevant description in step S110 to obtain known lead compounds and corresponding skeleton transformation molecules from various sources. Using known and reliable sample data helps improve the reliability of the template database.

[0075] S220 involves pairing all the lead compounds and skeleton transformation molecules in the sample to obtain multiple reaction combinations with a preset format, where each reaction combination includes reactants and products.

[0076] It is understood that lead compounds and skeleton transformation molecules obtained from different sources have different formats. To facilitate the extraction of chemical reaction templates in subsequent steps, in one embodiment, the lead compound and its corresponding skeleton transformation molecules from the same sample are represented using SMILES strings to generate corresponding string datasets; each pair of molecules in the same string dataset is associated according to a preset format to form corresponding reaction combinations, and a reaction combination dataset is obtained.

[0077] Specifically, after obtaining the lead compound and its corresponding skeleton substitution molecule from known literature, the lead compound and its corresponding skeleton substitution molecule can be converted into SMILES strings for representation, thus forming a unified format. Furthermore, the SMILES strings of the same lead compound and its corresponding skeleton substitution molecule can be combined into a corresponding string dataset. Optionally, a single string dataset or multiple string datasets can be stored in the same file to form a corresponding text dataset. For example, a CSV file format can be used to reduce storage space; the storage format is only illustrated here.

[0078] like Figure 3 As shown, for the lead compound A (Upadacitinib) obtained from publicly available literature and its three skeletal variants B, C, and D with good activity, they can be converted into SMILES strings, such as SMILES A, SMILES B, SMILES C, and SMILES D. These four SMILES strings are then represented using a delimiter such as ".", forming a string dataset like "smilesA.smilesB.smilesC.smilesD". Optionally, multiple string datasets can be stored in the same CSV file or separately in their own CSV files; this is only an example.

[0079] Furthermore, every two molecules in the same string dataset are paired sequentially. To improve processing efficiency, each string dataset in the CSV file can be read using Python. Specifically, for example, in the string dataset mentioned above, the corresponding string datasets are placed into a list using "." as the delimiter. A for loop is used to combine the four SMILES strings in the list in pairs, using characters such as ">>" or "<<" for association, resulting in 12 reaction combinations with a uniform preset format. Each reaction combination forms a corresponding reaction combination dataset, such as A>>B, B>>A, A>>C, C>>A, A>>D, D>>A, B>>C, C>>B, B>>D, D>>B, C>>D, D>>C. Of course, the preset format is only illustrative and not restrictive. Optionally, the reaction combinations of the same sample lead compound can be stored in a separate CSV file, distinct from the CSV file containing the string dataset.

[0080] S230, in the corresponding reaction combination, the atoms of the reactants and the atoms of the products are mapped and numbered respectively to obtain a mapping number dataset.

[0081] In this step, to facilitate subsequent identification of atoms in the reactants whose chemical environment has changed, the atoms in the reactants and products can be mapped and numbered using relevant techniques. For example, atom mapping can be performed using algorithms such as Atom Mapping or Python RXN Mapper. By matching reactants and products, heavy atoms at the same position can be mapped to each other and have the same number, and vice versa.

[0082] Specifically, for example, the CSV file of the reaction combination dataset generated in step S220 can be read using Python. For example, the atoms in the smiles can be mapped by calling the get_attention_guided_atom_maps method in rxnmapper using Python, or by calling the Java Reaction Decoder Tool API, to generate a dataset with mapping numbers.

[0083] like Figure 4 As shown, this is based on Figure 3 The reaction combination "B>>D" consisting of molecules in the image is represented by a molecular map with mapping numbers obtained using the Atom Mapping algorithm. Each heavy atom in both the reactant (i.e., the skeleton-transformed molecule B) and the product (i.e., the skeleton-transformed molecule D) has a corresponding mapping number.

[0084] S240, based on the mapping number dataset, identify the atoms in the reactants whose chemical environment has changed; take the atoms in the reactants whose chemical environment has changed as the reaction center, take the reaction center and the structures within a preset radius from the reaction center as reactant structural fragments, and obtain the corresponding product structural fragments in the products; generate the corresponding chemical reaction template based on the reactant structural fragments and product structural fragments.

[0085] Based on the mapping number dataset, atoms in the reactants whose chemical environment has changed can be identified using relevant techniques such as those in RDKit. The atoms in the reactants whose chemical environment has changed are considered the reaction center, and the reaction center and atoms within a predetermined radius from it are considered as structural fragments participating in the reaction, i.e., reactant structural fragments. Alternatively, when an atom within the predetermined radius is located within an important functional group, that functional group can also be included in the reactant structural fragment. The predetermined radius can be set according to actual needs, such as 1 Å, 2 Å, etc. Accordingly, based on the reactant structural fragments and the mapping number dataset, the corresponding product structural fragments in the product are determined. The method for obtaining product structural fragments can be found in the above-described method for obtaining reactant structural fragments, and will not be repeated here.

[0086] Specifically, based on the smarts writing rules in related technologies, chemical reaction templates can be automatically generated using a written algorithm. Specifically, smarts writing rules can include at least: chemical bond information (e.g., single bonds, double bonds, triple bonds, benzene rings, etc.) and atomic information in reactants and products; where atomic information includes the heavy atom number, the number and type of chemical bonds connected to each heavy atom, the heavy atom charge, the corresponding numbers of reactants and products, and the chiral information of reactants and products, etc. In other words, based on the content in the smarts writing rules, the corresponding structural fragments can be extracted from reactants and products by calling AllChem.MolFragmentToSmiles in RDKit to generate a chemical reaction template.

[0087] It can be understood that a reactant structural fragment and a corresponding product structural fragment can form a template for a chemical reaction. For example... Figure 5 As shown, Figure 5 yes Figure 4 Chemical reaction templates extracted from reaction combinations.

[0088] In one embodiment, if the reaction center and / or the structure within a predetermined radius of the reaction center contain atoms located in the first target group, then the reaction center, the structure within the predetermined radius of the reaction center, and the first target group are considered as reactant structural fragments. The first target group can be a predetermined important group, such as C(=O)Cl, O=C([O,N])-[*], etc. When the reaction center is located within the first target group, or when the reaction center does not belong to the first target group but its structure within the predetermined radius contains atoms from the first target group, the first target group can be considered as part of the reactant structural fragment.

[0089] To further enrich the skeleton transformation template database, existing chemical reaction templates can be updated to include more new templates. In one embodiment, the first target group in the reactant structural fragment is replaced with a first preset group to obtain a new reactant structural fragment. The new reactant structural fragment and product structural fragment are then used to generate a corresponding chemical reaction template. In other words, when a reactant structural fragment is found to have a first target group, while generating a chemical reaction template composed of the original reactant structural fragment and the original product structural fragment, the first target group in the original reactant structural fragment can be replaced with a first preset group to form a new reactant structural fragment. This new reactant structural fragment is then combined with the original product structural fragment to generate a new chemical reaction template, thereby enriching the skeleton transformation template database and increasing the diversity of newly generated molecules. For example, if the first target group is C(=O)Cl, the replaceable first preset group could be C(=O)[Cl, I, Br, F], meaning the Cl element in the first target group can be replaced with I, Br, or F. Similarly, if the first target group is O=C([O,N])-[*], the first preset group could be a carboxyl group, an amide, or an ester. This allows the replaced reaction template to cover as many atom types as possible, and makes it easier to match reaction templates when performing molecular generation based on this skeleton transformation template database.

[0090] Similarly, to enrich the skeletal transformation template database and improve the diversity of generated molecules, in one embodiment, if the product structural fragment contains a second target group, the second target group is replaced with a second preset group to obtain a new product structural fragment; using the reactant structural fragment and the new product structural fragment, a corresponding chemical reaction template is generated. That is, for an already generated chemical reaction template, while keeping the reactant structural fragment unchanged, if the product structural fragment contains a second target group, the second target group is replaced with a second preset group to obtain a new product structural fragment, thereby forming a new chemical reaction template based on the combination of the original reactant structural fragment and the new product structural fragment. Similarly, the second target group may include, for example, C(=O)Cl, and the second preset group may be, for example, C(=O)[Cl, I, Br, F], which is only illustrative and not limiting.

[0091] Furthermore, to enrich the skeletal transformation template database, in one embodiment, corresponding chemical reaction templates are generated using new reactant structural fragments and new product structural fragments. That is, for an existing chemical reaction template, if the original reactant structural fragment contains a first target group, the first target group can be replaced with a first preset group to obtain a new reactant structural fragment. Simultaneously, if the original product structural fragment contains a second target group, the second target group can be replaced with a second preset group to obtain a new product structural fragment. Thus, new chemical reaction templates are formed based on the new reactant structural fragments and the new product structural fragments. This design allows the template database to contain more chemical reaction templates, thereby improving the matching rate of reaction templates and the diversity of new molecule generation.

[0092] The reactant structural fragment may contain more than one target group. These target groups may be identical, partially different, or completely different. Each target group can have at least one corresponding preset group for replacement. Identical target groups can have corresponding preset groups of the same or different nature. When there are multiple target groups in the reactant structural fragment, each replacement can target only one target group, or multiple or all target groups simultaneously; this is not limited here. The principle for replacing target groups in the product structural fragment is the same as above and will not be repeated here.

[0093] S250 processes the chemical reaction templates to establish a skeleton transformation template database.

[0094] By collecting chemical reaction templates, a skeleton transformation template database can be established. Furthermore, the chemical reaction templates can be converted according to different storage formats, thus enabling the skeleton transformation template database to have more computer-readable format types.

[0095] Based on the skeleton transformation template database, the target molecule to be optimized can be optimized by referring to the chemical reaction templates therein, and new molecules that can be used for reference can be obtained.

[0096] As can be seen from this example, the method for establishing the skeleton transformation template database in this application can create a template database with reference value, which can be used as a reference for skeleton transformation when optimizing various target molecules, and is conducive to obtaining new molecules that are available for reference and easier to synthesize more efficiently.

[0097] An embodiment of this application also provides a skeleton transformation template database, which can be established and obtained according to the establishment method in the above embodiments.

[0098] Figure 6 This is a schematic flowchart of the molecular generation method shown in this application.

[0099] See Figure 6 This application discloses a method for generating a molecule, comprising:

[0100] S310, obtain the target molecule to be optimized.

[0101] In this step, any molecule to be optimized can be used as the target molecule. For ease of understanding, as follows: Figure 7 As shown, Figure 7 The target molecule shown can be used to test this molecule generation method.

[0102] S320: Obtain a chemical reaction template that matches the target molecule based on the skeleton transformation template database.

[0103] In this step, in one specific implementation, the target molecule is matched with the reactant structure fragments of each chemical reaction template in the skeleton transformation template database to determine the corresponding substructure fragment to be replaced in the target molecule.

[0104] It is understood that the target molecule has its own molecular structure. The molecular structure of the target molecule can be matched with reactant structural fragments in chemical reaction templates in a database. When at least a portion of the target molecule's structure is identical or similar to a reactant structural fragment in the chemical reaction template (e.g., similarity greater than 80%), the match is considered successful. This identifies the corresponding chemical reaction template as the template for the target molecule's skeleton transformation, and simultaneously identifies the substructure fragments in the target molecule that require skeleton transformation. It should be noted that, depending on the inherent structure of each target molecule, a single target molecule may have one or more substructure fragments to be replaced. That is, a single target molecule may contain multiple substructure fragments that match their respective reactant structural fragments, thus allowing for the matching of multiple chemical reaction templates. Furthermore, a single substructure fragment in the target molecule may match multiple identical or similar reactant structural fragments, thus identifying multiple corresponding chemical reaction templates.

[0105] The number of chemical reaction templates that can be matched in the skeleton transformation template database is uncertain due to the different structures of the target molecules themselves. The richer the chemical reaction templates in the skeleton transformation template database, the more reactant structural fragments can be matched, and the more chemical reaction templates can be obtained for the target molecule.

[0106] Specifically, to facilitate matching, the target molecule can be converted into a smileys string as input data. Furthermore, the AllChem.ReactionFromSmarts method in rdkit can be called in Python to convert the chemical reaction template into a Reaction object, and the smileys string of the input target molecule can be converted into a Mol object using Chem.MolFromSmiles() for matching.

[0107] S330 generates corresponding optimized molecules from the target molecules based on the matched chemical reaction templates.

[0108] In one specific implementation, based on a corresponding chemical reaction template, the substructure fragments in the target molecule are replaced with the corresponding product structure fragments to generate the corresponding optimized molecule. In other words, based on the matched chemical reaction template, the matched substructure fragments in the target molecule undergo skeletal transformation, replacing them with the corresponding product structure fragments from the chemical reaction template, thereby obtaining the corresponding optimized molecule.

[0109] like Figure 8 As shown, Figure 8 The 8 molecules in it are Figure 7 The optimized molecule is the target molecule after skeleton replacement. That is, it represents... Figure 7The target molecule was successfully matched with eight chemical reaction templates. Based on the corresponding chemical reaction templates, the matching substructure fragments in the target molecule were subjected to skeletal transformation and replaced with the corresponding product structure fragments in the chemical reaction templates, thereby obtaining the corresponding optimized molecules.

[0110] Specifically, the RunReactants() method of the Reaction object can be called to generate a new optimized molecule with a transformed skeleton from the target molecule of the previous step.

[0111] As can be seen from this example, the molecular generation method of this application, based on a reliable database of skeletal transformation templates, can more efficiently and reliably transform the skeletal structure of a target molecule according to a matching chemical reaction template, thereby obtaining a new optimized molecule.

[0112] Corresponding to the aforementioned application function implementation method embodiments, this application also provides an apparatus for establishing a skeleton transformation template database, a molecular generation apparatus, and corresponding embodiments.

[0113] Figure 9 This is a schematic diagram of the structure of the skeleton transformation template database creation device shown in this application.

[0114] See Figure 9 The skeleton transformation template database establishment apparatus 40 shown in this application includes a first acquisition module 410, a combination generation module 420, a template generation module 430, and a database construction module 440, wherein:

[0115] The first acquisition module 410 is used to acquire the skeleton transformation molecule of the sample lead compound.

[0116] The combination generation module 420 is used to obtain a reaction combination composed of any two different molecules based on the sample lead compound and skeleton transformation molecule, wherein the reaction combination includes reactants and products.

[0117] The template generation module 430 is used to obtain the chemical reaction templates corresponding to each reaction combination, wherein the chemical reaction templates include reactant structural fragments and corresponding product structural fragments.

[0118] The database construction module 440 is used to establish a skeleton transformation template database using various chemical reaction templates.

[0119] See Figure 10In one specific embodiment, the combination generation module 420 includes a format conversion module 421 and a combination module 422. The format conversion module 421 is used to represent the same sample lead compound and its corresponding skeleton transformation molecules using SMILES strings, generating corresponding string datasets. The combination module 422 is used to associate every two molecules in the same string dataset according to a preset format to form corresponding reaction combinations and obtain a reaction combination dataset.

[0120] In one specific embodiment, the template generation module 430 includes a mapping and numbering module 431 and a template extraction module 432. The mapping and numbering module 431 maps and numbers the atoms of reactants and the atoms of products in the corresponding reaction combinations, obtaining a dataset of mapping numbers for reactants and products. The template extraction module 432, based on the mapping number dataset, identifies atoms in the reactants whose chemical environment changes; uses these atoms as reaction centers, and identifies the reaction centers and structures within a preset radius as reactant structural fragments, obtaining corresponding product structural fragments in the products; and generates corresponding chemical reaction templates based on the reactant and product structural fragments.

[0121] Furthermore, the template extraction module 432 is used to, if the reaction center and / or the structure within a preset radius from the reaction center contain atoms located in the first target group, then use the reaction center, the structure within the preset radius from the reaction center, and the first target group as reactant structural fragments. The template extraction module 432 is used to replace the first target group in the reactant structural fragment with a first preset group to obtain a new reactant structural fragment; using the new reactant structural fragment and the product structural fragment, a corresponding chemical reaction template is generated.

[0122] The template extraction module 432 is used to replace the second target group with a second preset group if the product structure fragment contains a second target group, thereby obtaining a new product structure fragment; the template extraction module 432 is used to generate a corresponding chemical reaction template using the new reactant structure fragment and the new product structure fragment. And / or, the template extraction module 432 is used to generate a corresponding chemical reaction template using the reactant structure fragment and the new product structure fragment.

[0123] As can be seen from this example, the apparatus for establishing a skeletal transformation template database of this application can establish a database for molecular skeletal transformation based on reliable data sources, which helps to generate new molecules that are easier to synthesize and have more reliable drug-like properties.

[0124] Figure 11 This is a schematic diagram of the molecular generation device shown in this application.

[0125] See Figure 11 The molecular generation apparatus 50 shown in this application includes a second acquisition module 510, a processing module 520, and a molecular generation module 530, wherein:

[0126] The second acquisition module 510 is used to acquire the target molecule to be optimized.

[0127] The processing module 520 is used to obtain a chemical reaction template that matches the target molecule based on the skeleton transformation template database.

[0128] The molecule generation module 530 is used to generate corresponding optimized molecules from the target molecules according to the matched chemical reaction templates.

[0129] Specifically, the processing module 520 includes a function to match the target molecule with each reactant structural fragment in the skeleton transformation template database to determine the corresponding substructure fragment to be replaced in the target molecule and the corresponding chemical reaction template. The molecule generation module 530 is used to replace the substructure fragment in the target molecule with the corresponding product structural fragment according to the corresponding chemical reaction template, thereby generating the corresponding optimized molecule.

[0130] As can be seen from this example, the molecular generation device of this application can perform skeletal transformation on any target molecule based on a reliable skeletal transformation template database, and generate new molecules that are easier to synthesize and have more reliable drug-like properties based on matching chemical reaction templates. This helps to reduce unnecessary trial and error, save R&D costs, and improve R&D efficiency.

[0131] The specific manner in which each module performs its operation in the above embodiments has been described in detail in the embodiments related to the method, and will not be elaborated further here.

[0132] Figure 12 This is a schematic diagram of the structure of the electronic device shown in this application.

[0133] See Figure 12 The electronic device 1000 includes a memory 1010 and a processor 1020.

[0134] The processor 1020 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0135] Memory 1010 may include various types of storage units, such as system memory, read-only memory (ROM), and permanent storage devices. ROM may store static data or instructions required by processor 1020 or other modules of the computer. Permanent storage devices may be read-write storage devices. Permanent storage devices may be non-volatile storage devices that retain stored instructions and data even when the computer is powered off. In some embodiments, permanent storage devices use mass storage devices (e.g., magnetic or optical disks, flash memory) as permanent storage devices. In other embodiments, permanent storage devices may be removable storage devices (e.g., floppy disks, optical drives). System memory may be a read-write storage device or a volatile read-write storage device, such as dynamic random access memory. System memory may store some or all of the instructions and data required by the processor during operation. Furthermore, memory 1010 may include any combination of computer-readable storage media, including various types of semiconductor memory chips (e.g., DRAM, SRAM, SDRAM, flash memory, programmable read-only memory), and disks and / or optical disks may also be used. In some embodiments, the memory 1010 may include a removable storage device that is readable and / or writable, such as a laser disc (CD), a read-only digital multifunction optical disc (e.g., DVD-ROM, dual-layer DVD-ROM), a read-only Blu-ray disc, a high-density optical disc, a flash memory card (e.g., SD card, mini SD card, Micro-SD card, etc.), a magnetic floppy disk, etc. Computer-readable storage media do not contain carrier waves or transient electronic signals transmitted wirelessly or via wired connections.

[0136] The memory 1010 stores executable code, which, when processed by the processor 1020, can cause the processor 1020 to execute part or all of the methods described above.

[0137] Furthermore, the method according to this application can also be implemented as a computer program or computer program product, which includes computer program code instructions for performing some or all of the steps in the method described above.

[0138] Alternatively, this application may be implemented as a computer-readable storage medium (or a non-transitory machine-readable storage medium or a machine-readable storage medium) storing executable code (or computer program or computer instruction code) thereon, which, when executed by a processor of an electronic device (or an electronic device, etc.), causes the processor to perform part or all of the steps of the above-described method according to this application.

[0139] The various embodiments of this application have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A method for establishing a skeleton transformation template database, characterized by, The method comprises the following steps: obtain skeleton transformation molecules of sample lead compounds; obtain reaction combinations composed of any two different molecules according to the sample lead compounds and the skeleton transformation molecules, wherein the reaction combinations comprise reactants and products; obtain chemical reaction templates corresponding to each of the reaction combinations, wherein in each reaction combination, a structure fragment in which the chemical environment of the reactant is changed is extracted as a reactant structure fragment, and a product structure fragment corresponding to the reactant structure fragment is extracted from the corresponding product, to obtain a chemical reaction template composed of the reactant structure fragment and the corresponding product structure fragment; establish a skeleton transformation template database using the chemical reaction templates.

2. The method of claim 1, wherein, The method of obtaining reaction combinations composed of any two different molecules according to the sample lead compounds and the skeleton transformation molecules comprises: express the same sample lead compound and each corresponding skeleton transformation molecule in the same sample lead compound using SMILES strings to generate corresponding string data sets; associate each two molecules in the same string data set according to a preset format to form corresponding reaction combinations, and obtain a reaction combination data set.

3. The method of claim 1, wherein, The method of obtaining chemical reaction templates corresponding to each of the reaction combinations comprises: map and number the atoms of the reactant and the atoms of the product in the corresponding reaction combination to obtain a mapping and numbering data set of the reactant and the product; determine the atoms in the reactant in which the chemical environment is changed according to the mapping and numbering data set; take the atoms in the reactant in which the chemical environment is changed as a reaction center, and take the structure within a preset radius from the reaction center as a reactant structure fragment, and obtain a corresponding product structure fragment in the product; generate a corresponding chemical reaction template according to the reactant structure fragment and the product structure fragment.

4. The method of claim 3, wherein, The method of taking the structure within a preset radius from the reaction center as a reactant structure fragment comprises: if the reaction center and / or the structure within a preset radius from the reaction center contains an atom located at a first target group, then the reaction center, the structure within a preset radius from the reaction center, and the first target group are taken as a reactant structure fragment.

5. The method of claim 4, wherein, The method of generating a corresponding chemical reaction template according to the reactant structure fragment and the product structure fragment comprises: replace the first target group in the reactant structure fragment with a first preset group to obtain a new reactant structure fragment; generate a corresponding chemical reaction template using the new reactant structure fragment and the product structure fragment.

6. The method of claim 5, wherein, The method further comprises: if the product structure fragment contains a second target group, replace the second target group with a second preset group to obtain a new product structure fragment; The method of generating a corresponding chemical reaction template using the new reactant structure fragment and the product structure fragment comprises: generate a corresponding chemical reaction template using the new reactant structure fragment and the new product structure fragment.

7. The method of claim 4, wherein, The method of generating a corresponding chemical reaction template according to the reactant structure fragment and the product structure fragment comprises: If the second target group is included in the product structure fragment, the second target group is replaced by a second preset group to obtain a new product structure fragment; A corresponding chemical reaction template is generated by using the reactant structure fragment and the new product structure fragment.

8. A skeleton transformation template database, characterized by, The method is constructed according to the establishing method of the skeleton transformation template database in any one of claims 1 to 7.

9. A method of molecular generation, characterized by, Comprise: Obtaining a target molecule to be optimized; According to the skeleton transformation template database in claim 8, a chemical reaction template matched with the target molecule is obtained; According to each matched chemical reaction template, the target molecule is generated into a corresponding optimized molecule.

10. The method of claim 9, wherein, According to the skeleton transformation template database, a chemical reaction template matched with the target molecule is obtained, comprising: Matching the target molecule with each reactant structure fragment in the skeleton transformation template database to determine a corresponding substructure fragment to be replaced in the target molecule and a corresponding chemical reaction template; According to each chemical reaction template, the target molecule is generated into a corresponding optimized molecule, comprising: According to the corresponding chemical reaction template, the substructure fragment in the target molecule is replaced by a corresponding product structure fragment to generate a corresponding optimized molecule.

11. An apparatus for creating a skeleton transformation template database, characterized by comprising: Comprise: A first obtaining module is configured to obtain a skeleton transformation molecule of a sample lead compound; A combination generating module is configured to obtain a reaction combination composed of any two different molecules according to the sample lead compound and the skeleton transformation molecule, wherein the reaction combination comprises a reactant and a product; A template generating module is configured to obtain a chemical reaction template corresponding to each reaction combination, wherein in each reaction combination, a structure fragment in which a chemical environment is transformed is extracted from the reactant as a reactant structure fragment, and a product structure fragment corresponding to the reactant structure fragment is extracted from the corresponding product to obtain a chemical reaction template composed of the reactant structure fragment and the corresponding product structure fragment; A database constructing module is configured to establish a skeleton transformation template database by using each chemical reaction template.

12. A molecular production device, characterized by, Comprise: A second obtaining module is configured to obtain a target molecule to be optimized; A processing module is configured to obtain a chemical reaction template matched with the target molecule according to the skeleton transformation template database in claim 8; A molecule generating module is configured to generate a corresponding optimized molecule from the target molecule according to each matched chemical reaction template.

13. An electronic device, comprising: Comprise: A processor; And A memory having executable code stored thereon, when the executable code is executed by the processor, causes the processor to execute the establishing method of the skeleton transformation template database in any one of claims 1 to 7 or the molecule generating method in any one of claims 9 to 10.

14. A computer readable storage medium having executable code stored thereon, when the executable code is executed by a processor of an electronic device, causes the processor to execute the establishing method of the skeleton transformation template database in any one of claims 1 to 7 or the molecule generating method in any one of claims 9 to 10.

Images