A drug screening method, device, and computer-readable storage medium

Abstract

Embodiments of the present application disclose a drug screening method, device and computer readable storage medium; after obtaining protein information of a target protein and molecular information of a candidate drug molecule, the target protein is a target macromolecule which can be acted on by the drug molecule, then, feature extraction is performed on the protein information and the molecular information respectively to obtain protein features of the target protein and molecular features of the candidate drug molecule, according to the protein information and the molecular information, interaction information of the target protein and the candidate drug molecule under a molecular force field is determined, then, based on the protein features, the molecular features and the interaction information, a binding strength between the target protein and the candidate drug molecule is calculated, and according to the binding strength, a target drug molecule is screened from the candidate drug molecule; the scheme can improve the accuracy of drug screening.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of information processing, in particular to a drug screening method, device and computer readable storage medium. BACKGROUND

[0002] In recent years, with the popularity of neural network technology in the field of artificial intelligence, the application of neural network in drug design has also made great progress. Especially the application of neural network technology in drug screening for specific target proteins. The existing drug screening technology is often that the neural network predicts the binding strength between the protein and the small molecule based on the chemical information of the drug small molecule and the atoms around it or constructs the graph data of the drug small molecule and the atoms around it.

[0003] In the research and practice of the prior art, the inventors of the present application found that the chemical information and the graph data are only a simple description of the chemical relationship and position connection relationship between the drug small molecule and the protein atoms, and cannot accurately express the binding information between the protein and the drug small molecule, so that the accuracy of the binding strength between the protein and the drug small molecule is low, thus greatly reducing the accuracy of drug screening. SUMMARY

[0004] The embodiments of the present application provide a drug screening method, device and computer readable storage medium, which can improve the accuracy of drug screening.

[0005] A drug screening method, comprising:

[0006] obtaining protein information of a target protein and molecular information of a candidate drug molecule, the target protein being a target macromolecule that can be acted on by a drug molecule;

[0007] respectively extracting features from the protein information and the molecular information to obtain protein features of the target protein and molecular features of the candidate drug molecule;

[0008] determining interaction information between the target protein and the candidate drug molecule under a molecular force field according to the protein information and the molecular information;

[0009] calculating a binding strength between the target protein and the candidate drug molecule based on the protein features, the molecular features and the interaction information;

[0010] screening a target drug molecule from the candidate drug molecule according to the binding strength.

[0011] Correspondingly, the embodiments of the present application provide a drug screening device, comprising:

[0012] An acquisition unit is configured to acquire protein information of a target protein and molecule information of a candidate drug molecule, the target protein being a target macromolecule capable of being acted on by a drug molecule;

[0013] An extraction unit is configured to respectively perform feature extraction on the protein information and the molecule information to obtain protein features of the target protein and molecule features of the candidate drug molecule.

[0014] A determination unit is configured to determine interaction information of the target protein and the candidate drug molecule under a molecular force field according to the protein information and the molecule information.

[0015] A calculation unit is configured to calculate a binding strength between the target protein and the candidate drug molecule based on the protein features, the molecule features and the interaction information.

[0016] A screening unit is configured to screen a target drug molecule from the candidate drug molecule according to the binding strength.

[0017] Optionally, in some embodiments, the extraction unit can be specifically configured to identify protein atom information of protein atoms in the protein information and drug atom information of drug atoms in the molecule information; perform feature extraction on the drug atom information to obtain the molecule features of the candidate drug molecule; and perform feature extraction on the protein atom information to obtain the protein features of the target protein.

[0018] Optionally, in some embodiments, the extraction unit can be specifically configured to identify, according to an atom type, at least one target drug atom of a target atom type for interaction in the drug atoms; screen target drug atom information corresponding to the target drug atom from the drug atom information; and perform feature extraction on the target drug atom information to obtain the molecule features of the candidate drug molecule.

[0019] Optionally, in some embodiments, the extraction unit can be specifically configured to identify first position information of the target drug atom in the target drug atom information; perform feature extraction on the target drug atom information to obtain initial molecule features of the candidate drug molecule; construct molecule point cloud features of the candidate drug molecule according to the first position information and the initial molecule features, and take the molecule point cloud features as the molecule features of the candidate drug molecule.

[0020] Optionally, in some embodiments, the determining unit can be specifically configured to identify, from the protein information and the first position information, a target protein atom in the protein atoms that interacts with the target drug atom under the molecular force field; calculate an interaction force between the target protein atom and the target drug atom under the molecular force field; and determine, based on the molecular information and the interaction force, the interaction information between the target protein and the candidate drug molecule under the molecular force field.

[0021] Optionally, in some embodiments, the determining unit can be specifically configured to identify, from the protein information, second position information of the protein atoms; calculate, according to the first position information and the second position information, an atomic distance between the target drug atom and each protein atom; and screen, based on the atomic distance, a target protein atom in the protein atoms that interacts with the target drug atom under the molecular force field.

[0022] Optionally, in some embodiments, the determining unit can be specifically configured to determine, according to a type of the molecular force field, a distance threshold for interaction between the protein atom and the target drug atom; and screen, from the protein atoms, at least one protein atom with an atomic distance not exceeding the distance threshold, to obtain the target protein atom.

[0023] Optionally, in some embodiments, the determining unit can be specifically configured to determine, according to a type of the molecular force field, an action force type of the interaction force under the molecular force field; screen, from a preset mapping relationship set, a mapping relationship corresponding to the action force type, the mapping relationship being a mapping relationship between the interaction force and the atomic distance; and map, according to the mapping relationship, the interaction force between the target drug atom and the target protein atom under the molecular force field.

[0024] Optionally, in some embodiments, the determining unit can be specifically configured to map, according to the mapping relationship, an initial interaction force corresponding to the atomic distance between the target drug atom and the target protein atom; classify the initial interaction force based on the action force type; and fuse, according to a classification result, the initial interaction force corresponding to each action force type, to obtain the interaction force.

[0025] Optionally, in some embodiments, the determining unit can be specifically configured to identify, from the molecular information, target attribute information of the target drug atom for interaction; and determine, according to the target attribute information and the interaction force, the interaction information between the target protein and the candidate drug molecule under the molecular force field.

[0026] Optionally, in some embodiments, the determining unit can be specifically configured to extract at least one attribute information of the target drug atom from the molecular information; identify the attribute information to obtain an attribute type corresponding to the attribute information; and filter out target attribute information for interaction from the attribute information according to the attribute type.

[0027] Optionally, in some embodiments, the calculating unit can be specifically configured to perform feature extraction on the interaction information to obtain interaction features; filter out protein features corresponding to the target protein atom from the protein features to obtain target protein features; and perform intensity calculation on the target protein features, the molecular features and the interaction features by using a preset intensity calculation model to obtain a binding strength between the target protein and the candidate drug molecule.

[0028] Optionally, in some embodiments, the calculating unit can be specifically configured to perform feature transformation on the target protein features, the molecular features and the interaction features to obtain local binding strength features of multiple dimensions; perform fusion on the local binding strength features to obtain global binding strength features; and determine the binding strength between the target protein and the candidate drug molecule according to the global binding strength features.

[0029] In addition, an electronic device is also provided in the embodiments of the present application, which comprises a processor and a memory, the memory stores an application program, and the processor is configured to run the application program in the memory to implement the drug screening method provided in the embodiments of the present application.

[0030] In addition, a computer readable storage medium is also provided in the embodiments of the present application, which stores a plurality of instructions, and the instructions are suitable for being loaded by a processor to execute the steps in any one of the drug screening methods provided in the embodiments of the present application.

[0031] After obtaining protein information of a target protein and molecular information of a candidate drug molecule, the target protein is a target macromolecule that can be acted on by the drug molecule, then, feature extraction is respectively performed on the protein information and the molecular information to obtain protein features of the target protein and molecular features of the candidate drug molecule, according to the protein information and the molecular information, interaction information of the target protein and the candidate drug molecule under a molecular force field is determined, then, based on the protein features, the molecular features and the interaction information, a binding strength between the target protein and the candidate drug molecule is calculated, and according to the binding strength, a target drug molecule is screened out from the candidate drug molecule; since the scheme determines the interaction information of the target protein and the candidate drug molecule under the molecular force field on the basis of the extracted protein features and molecular features, the binding information between the protein and the drug small molecule can be more accurately expressed by introducing the molecular force field and the interaction information under the molecular force field, so that the accuracy of the binding strength between the protein and the drug small molecule is improved, and therefore, the accuracy of drug screening can be greatly improved. BRIEF DESCRIPTION OF DRAWINGS

[0032] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative effort on the basis of these drawings.

[0033] Figure 1 is a scene schematic diagram of the drug screening method provided by the embodiment of the present application;

[0034] Figure 2 is a flow schematic diagram of the drug screening method provided by the embodiment of the present application;

[0035] Figure 3 is a schematic diagram of calculating the binding strength between the target protein and the candidate drug molecule provided by the embodiment of the present application;

[0036] Figure 4 is another flow schematic diagram of the drug screening method provided by the embodiment of the present application;

[0037] Figure 5 is a structural schematic diagram of the drug screening device provided by the embodiment of the present application;

[0038] Figure 6 is a structural schematic diagram of the electronic device provided by the embodiment of the present application. DETAILED DESCRIPTION

[0039] With reference to the accompanying drawings, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative work fall within the scope of the present application.

[0040] The embodiments of the present application provide a drug screening method, a device and a computer readable storage medium. The drug screening device can be integrated in an electronic device, which can be a server or a terminal.

[0041] The server can be a physical server, a server cluster or a distributed system composed of multiple physical servers, a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, network acceleration services (CDN), and basic cloud computing services such as big data and artificial intelligence platforms. The terminal can be a smart phone, a tablet computer, a notebook computer, a desktop computer, a smart speaker, a smart watch, etc., but is not limited thereto. The terminal and the server can be directly or indirectly connected through wired or wireless communication, which is not limited in the present application.

[0042] For example, referring to Figure 1 For example, the electronic device integrates the drug screening device. After obtaining the protein information of the target protein and the molecular information of the candidate drug molecule, the target protein is a target macromolecule that can be acted on by the drug molecule. Then, the protein information and the molecular information are respectively extracted to obtain the protein features of the target protein and the molecular features of the candidate drug molecule. According to the protein information and the molecular information, the interaction information of the target protein and the candidate drug molecule under the molecular force field is determined. Then, based on the protein features, the molecular features and the interaction information, the binding strength between the target protein and the candidate drug molecule is calculated. According to the binding strength, the target drug molecule is screened from the candidate drug molecule.

[0043] The drug screening in the scheme can be understood as virtual drug screening. The so-called virtual drug screening is to simulate the process of drug screening on a computer, predict the possible activity of a compound, and then perform targeted physical screening on the compound that is more likely to be a drug, thereby greatly reducing the cost of drug development. The virtual drug screening in the scheme mainly applies molecular docking technology. To implement such screening, the molecular structure of the drug action target needs to be known. The ability of small molecules in the compound library to bind to the target is calculated through molecular simulation means to predict the physiological activity of the candidate compound. Thus, the required drug molecules are screened from the candidate drug molecules.

[0044] The drug screening method provided in the embodiments of the present application relates to the field of artificial intelligence. The calculation of the binding strength between the target protein and the candidate drug molecule in the embodiments of the present application can be calculated by a preset strength calculation model.

[0045] Artificial intelligence (AI) is the use of digital computers or digital computer-controlled machines to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technology of computer science, which tries to understand the essence of intelligence and produce a new intelligent machine that can react in a similar way to human intelligence. Artificial intelligence is the design principle and implementation method of various intelligent machines, so that machines have the functions of perception, reasoning and decision-making. Artificial intelligence technology is a comprehensive discipline, which involves a wide range of fields, including hardware and software technologies. The artificial intelligence software technology mainly includes computer vision technology, machine learning / deep learning and other directions.

[0046] The following will be described in detail. It should be noted that the order of the following embodiments is not limited as the preferred order of the embodiments.

[0047] The present embodiment will be described from the perspective of a drug screening device, which can be integrated in an electronic device. The electronic device can be a server, a terminal or other device. The terminal can include a tablet computer, a notebook computer, a personal computer (PC), a wearable device, a virtual reality device or other intelligent devices that can screen drugs.

[0048] A drug screening method comprises:

[0049] After obtaining the protein information of the target protein and the molecular information of the candidate drug molecule, the target protein is a target macromolecule that can be acted on by the drug molecule, the protein information and the molecular information are respectively subjected to feature extraction to obtain protein features of the target protein and molecular features of the candidate drug molecule, the interaction information of the target protein and the candidate drug molecule under a molecular force field is determined according to the protein information and the molecular information, the binding strength between the target protein and the candidate drug molecule is calculated based on the protein features, the molecular features and the interaction information, and the target drug molecule is screened from the candidate drug molecule according to the binding strength.

[0050] As shown in Figure 2 , the specific process of the drug screening method is as follows:

[0051] 101. Obtain the protein information of the target protein and the molecular information of the candidate drug molecule.

[0052] The target protein is a target macromolecule that can be acted on by the drug molecule. The protein information can be information describing the protein, which can include, for example, the composition elements of the protein, the related information of atoms in the protein, and some attribute information of the protein. The molecular information can be information describing the drug molecule, which can be a small molecule.

[0053] For example, the protein information of the target protein and the molecular information of the candidate drug molecule sent by the terminal can be directly obtained, for example, a user triggers the terminal to generate a drug screening request, adds the protein information of the target protein and the molecular information of at least one candidate drug molecule to the drug screening request, and sends the drug screening request with the added protein information and molecular information to the drug screening device, so that the drug screening device obtains the protein information of the target protein and the molecular information of the candidate drug molecule, or the terminal directly sends the protein information of the target protein and the molecular information of the candidate drug molecule to the drug screening device. The protein information of the target protein and the molecular information of the candidate drug molecule can also be indirectly obtained, for example, a user triggers the terminal to generate a drug screening request, adds the storage address of the protein information and the molecular information of the candidate drug molecule to the drug screening request, and then sends the drug screening request with the added storage address to the drug screening device, the drug screening device obtains the storage address in the drug screening request, and obtains the protein information of the target protein and the molecular information of the candidate drug molecule according to the storage address. The protein information of the target protein and the molecular information of the candidate drug molecule corresponding to the target protein can also be directly obtained from the protein database and the drug molecule database.

[0054] 102. The protein information and the molecular information are respectively subjected to feature extraction to obtain protein features of the target protein and molecular features of the candidate drug molecule.

[0055] The target protein can include a plurality of protein atoms, such as hydrogen atoms, oxygen atoms, carbon atoms, nitrogen atoms, and sulfur atoms, and the like. The candidate drug molecule can also include a plurality of drug atoms, such as hydrogen atoms, oxygen atoms, carbon atoms, nitrogen atoms, and sulfur atoms, and the like.

[0056] For example, protein atom information of protein atoms identified in the protein information and drug atom information of drug atoms identified in the molecule information are extracted as features, to obtain the molecular features of the candidate drug molecule, and the protein atom information is extracted as features, to obtain the protein features of the target protein atoms. Specifically, the following can be performed:

[0057] (1) Protein atom information of protein atoms identified in the protein information and drug atom information of drug atoms identified in the molecule information are extracted as features.

[0058] For example, information related to protein atoms identified in the protein information can be obtained as protein atom information, such as identification of the number information, atom type information, and position information of each protein atom in the target protein, which can be three-dimensional coordinate information, and the like. Information related to the candidate drug molecule identified in the molecule information can be obtained as drug atom information, such as identification of the number information, atom type, and position information of each drug atom, which can be obtained as drug atom information.

[0059] (2) The drug atom information is extracted as features, to obtain the molecular features of the candidate drug molecule.

[0060] For example, target drug atoms for interaction of at least one target atom type can be identified in drug atoms according to atom types, for example, drug atoms for interaction of at least one non-hydrogen atom can be identified in drug atoms according to atom types, to obtain target drug atoms. The target drug atom information corresponding to the target drug atoms is screened out in the drug atom information, for example, the target drug atom information corresponding to the non-hydrogen target drug atoms is screened out in the drug atom information. Feature extraction is performed on the target drug atom information to obtain the molecular features of the candidate drug molecules, for example, the first position information of the target drug atoms is identified in the target drug atom information, feature extraction is performed on the target drug atom information to obtain the initial molecular features of the target candidate drug molecules, for example, one-hot vectorization is performed on the target drug atoms in the target drug atom information, the initial drug atom features corresponding to each type of target drug atom can be obtained, and a vector of one dimension is added to the initial drug atom features to represent that the target drug atom comes from a drug molecule, so that the drug atom features of the target drug atoms can be obtained, and the drug atom features are fused to obtain the initial molecular features of the candidate drug molecules. According to the first position information and the initial molecular features, the molecular point cloud features of the candidate drug molecules are constructed, for example, according to the three-dimensional coordinates of the target drug atoms in the candidate drug molecules and the initial molecular features of the candidate drug molecules, the molecular point cloud features of the candidate drug molecules can be constructed, and the molecular point cloud features are taken as the molecular features of the candidate drug molecules.

[0061] (3) Feature extraction is performed on the protein atom information to obtain the protein features of the target protein.

[0062] For example, candidate protein atoms for interaction in protein atoms according to atom types, such as, candidate protein atoms for at least one non-hydrogen atom for interaction in protein atoms according to atom types, so as to obtain candidate protein atoms. Screening candidate protein atom information corresponding to candidate protein atoms in protein information, such as, screening candidate protein atom information corresponding to non-hydrogen atoms in candidate protein atom information. Feature extraction is performed on the candidate protein atom information to obtain the protein feature of the target protein, such as, identifying the protein position information of the candidate protein atom in the candidate protein atom information, performing feature extraction on the candidate protein atom information to obtain the initial protein feature of the target protein, for example, one-hot vectorization of the candidate protein atom in the candidate protein atom information can obtain the initial protein atom feature corresponding to each type of candidate protein atom, and adding a dimension of vector in the initial protein atom feature to represent that the candidate protein atom comes from the target protein can obtain the protein atom feature of the candidate protein atom. The protein point cloud feature of the target protein is constructed according to the protein position information and the initial protein feature, and the protein point cloud feature is taken as the protein feature of the target protein.

[0063] 103. According to the protein information and the molecular information, the interaction information of the target protein and the candidate drug molecule in the molecular force field is determined.

[0064] Wherein, the molecular force field can be understood as a force field for accurately calculating the interaction between atoms, in which the bonding interaction between atoms constituting the same molecule and the van der Waals interaction between different molecules can be calculated by a molecular force field function, and some molecules have hydrogen bond interaction.

[0065] Wherein, the interaction information can be understood as the information of the target protein atom and the target drug atom interacting in the molecular force field, such as, it can include the information of the interaction force between the target protein atom and the target drug atom in the molecular force field, and it can also include the attribute information of the target drug atom itself interacting with the target protein atom in the molecular force field.

[0066] For example, according to the protein information and the first position information, the target protein atom interacting with the target drug atom in the molecular force field is identified in the protein atom, the interaction force between the target protein atom and the target drug atom in the molecular force field is calculated, and the interaction information of the target protein and the candidate drug molecule in the molecular force field is determined based on the molecular information and the interaction force. Specifically, it can be as follows:

[0067] S1, identifying target protein atoms in the protein atoms that interact with the target drug atom under the molecular force field according to the protein information and the first position information.

[0068] For example, the second position information of the protein atoms is identified in the protein information, such as the three-dimensional coordinates of each protein atom in the target protein are identified in the protein information, so as to obtain the second position information of the protein atoms. According to the first position information and the second position information, the atomic distance between the target drug atom and each protein atom is calculated, such as the combination of the protein atom and the target drug atom under the molecular force field is simulated, and the atomic distance between the target drug atom and each protein atom is calculated according to the three-dimensional coordinates of the protein atom and the target drug atom. Based on the atomic distance, the target protein atom in the protein atoms that interacts with the target drug atom under the molecular force field is screened, such as the distance threshold value of the interaction between the protein atom and the target drug atom is determined according to the type of the molecular force field, for example, when the type of the molecular force field is X-score, the distance threshold value of this type of molecular force field can be 8 angstroms, and the distance threshold values under different molecular force fields can be the same or different. At least one protein atom whose atomic distance does not exceed the distance threshold value is screened from the protein atoms, and the target protein atom is obtained, such as the adjacent protein atom whose atomic distance with the target drug atom does not exceed 8 angstroms is screened from the protein atoms as the target protein atom when the distance threshold value is 8 angstroms.

[0069] S2, calculating the interaction force between the target protein atom and the target drug atom under the molecular force field.

[0070] Among them, the interaction force can be the cumulative van der Waals force, non-polar interaction force, hydrogen bond interaction force and polar interaction force of the target drug atom on each target protein atom.

[0071] For example, according to the type of the molecular force field, the type of the force of the interaction under the molecular force field is determined, such as when the type of the molecular force field is X-score, the type of the force under the molecular force field of this type can include cumulative van der Waals force, non-polar interaction force, hydrogen bond interaction force and polar interaction force, etc. The type of the force under different molecular force fields can be the same or different. The mapping relationship corresponding to the type of the force is screened out from the preset mapping relationship set, which is the mapping relationship between the interaction force and the atomic distance, such as when the type of the force is van der Waals force, the mapping relationship between the van der Waals force and the atomic distance can be screened out from the preset mapping relationship set, when the type of the force is non-polar interaction force, the mapping relationship between the non-polar interaction force and the atomic distance can be screened out from the preset mapping relationship set, and so on. According to the mapping relationship, the interaction force between the target drug atom and the target protein atom under the molecular force field is mapped, such as according to the mapping relationship, the initial interaction force corresponding to the atomic distance between the target drug atom and the target protein atom is mapped, for example, when the mapping relationship is the mapping relationship corresponding to the van der Waals force, according to the atomic distance between the target drug atom and the target protein atom, the initial van der Waals force corresponding to the atomic distance can be mapped, when the mapping relationship is the mapping relationship corresponding to the non-polar interaction force, the initial non-polar interaction force corresponding to the atomic distance can be mapped, and so on. According to the type of the force, the initial interaction force is classified, such as the initial van der Waals force is classified into a category, the initial non-polar interaction force is classified into a category, and so on. According to the classification result, the initial interaction force corresponding to each type of force is fused to obtain the interaction force, such as the initial interaction forces of the same type can be accumulated, such as all the initial van der Waals forces are accumulated to obtain the van der Waals force between the target drug atom and the target protein atom corresponding to the target drug atom, and so on. The interaction force between the target drug atom and the target protein atom can include one or more interaction forces.

[0072] S3、based on the molecular information and the interaction force, determining the interaction information between the target protein and the candidate drug molecule under the molecular force field.

[0073] For example, the target drug atom in the molecular information is identified as target attribute information for interaction, such as extracting at least one attribute information of the target drug atom in the molecular information, identifying the attribute information, obtaining the attribute type corresponding to the attribute information, and screening the target attribute information for interaction in the attribute information according to the attribute type. For example, the attribute type corresponding information such as rotatable bond information, atomic mass and oil-water separation coefficient (logP) can be screened in the attribute information to obtain the target attribute information of the target drug atom for interaction. According to the target attribute information and the interaction force, the interaction information of the target protein and the candidate drug molecule in the molecular force field is determined, such as the numerical value of the interaction force of the target protein atom and the target drug atom in the molecular force field and the target attribute information can be used as the interaction information of the target protein and the candidate drug molecule in the molecular force field.

[0074] 104. Based on the protein characteristics, molecular characteristics and interaction information, the binding strength between the target protein and the candidate drug molecule is calculated.

[0075] The binding strength refers to the strength of the interaction between the protein and other molecules (such as proteins or small organic molecules).

[0076] For example, the interaction information is extracted to obtain interaction features, such as the interaction information can be converted into a 9-dimensional feature vector, which is used to describe the cumulative van der Waals force, non-polar interaction force, hydrogen bond interaction force and polar interaction force of each target drug atom on the adjacent target protein atom, and the rotatable bond information, atomic mass and oil-water separation coefficient of the target drug atom itself. In the protein characteristics, the protein characteristics corresponding to the target protein atom are screened to obtain the target protein characteristics, such as the protein characteristics corresponding to the target protein atom interacting with the target drug atom in the protein characteristics, and the target protein characteristics can be obtained. The target protein characteristics, molecular characteristics and interaction characteristics are calculated by using a preset strength calculation model to obtain the binding strength between the target protein and the candidate drug molecule, such as the target protein characteristics, molecular characteristics and interaction characteristics are transformed to obtain local binding strength features of multiple dimensions, and the local binding strength features are fused to obtain global binding strength features, such as the feature mean of the local binding strength features can be calculated, and the local binding strength features are fused based on the feature mean, so that the global binding strength features can be obtained. According to the global binding strength features, the binding strength between the target protein and the candidate drug molecule is determined, such as the global binding strength features can be processed by using a full connection layer, and finally the binding strength between the target protein and the candidate drug molecule is output.

[0077] It should be noted that the preset intensity calculation model can be a point cloud depth neural network, which is used to process the input target protein features and molecular features such as point cloud features, and the interaction features of each target drug atom are also fused in the molecular features such as point cloud features, which can be specifically as shown in Figure 3 Thus, the preset intensity calculation model is more universal, and overfitting of the preset intensity calculation model can be effectively prevented. The training of the preset intensity calculation model can be performed by using a high-precision time-delay structure database of protein and drug molecule complexes, and by sample collection, the contributions of strong and weak interaction samples to the intensity calculation model are well balanced.

[0078] 105. According to the binding strength, the target drug molecule is screened from the candidate drug molecules.

[0079] For example, the binding strength threshold corresponding to the target protein can be obtained, the binding strength threshold is compared with the binding strength, and the drug molecule with the binding strength exceeding the binding strength threshold is screened from the candidate drug molecules as the target drug molecule. According to the binding strength, the interaction level of the candidate drug molecule and the target protein can be determined, and the drug molecule with the interaction level exceeding the preset level threshold is screened from the candidate drug molecules as the target drug molecule.

[0080] As can be seen from the above, after obtaining the protein information of the target protein and the molecular information of the candidate drug molecule, the target protein is a target macromolecule that can be acted on by the drug molecule. Then, the protein information and the molecular information are respectively extracted to obtain the protein features of the target protein and the molecular features of the candidate drug molecule. According to the protein information and the molecular information, the interaction information of the target protein and the candidate drug molecule under the molecular force field is determined. Then, based on the protein features, the molecular features and the interaction information, the binding strength between the target protein and the candidate drug molecule is calculated. According to the binding strength, the target drug molecule is screened from the candidate drug molecules. Since the scheme determines the interaction information of the target protein and the candidate drug molecule under the molecular force field on the basis of the extracted protein features and molecular features, the binding information between the protein and the drug small molecule can be more accurately expressed by introducing the molecular force field and the interaction information under the molecular force field, thereby improving the accuracy of the binding strength between the protein and the drug small molecule. Therefore, the accuracy of drug screening can be greatly improved.

[0081] According to the method described in the above embodiment, the following examples will be further described in detail.

[0082] In this embodiment, the drug screening device is specifically integrated in an electronic device, which is a server, and the type of the molecular force field is X-score, which will be described as an example.

[0083] As shown in the following, a drug screening method comprises the following steps: Figure 4

[0084] 201. The server acquires protein information of a target protein and molecular information of a candidate drug molecule.

[0085] For example, a user triggers a terminal to generate a drug screening request, adds protein information of a target protein and molecular information of at least one candidate drug molecule into the drug screening request, and sends the drug screening request with the added protein information and molecular information to the server, so that the server acquires the protein information of the target protein and the molecular information of the candidate drug molecule. Alternatively, the terminal directly sends the protein information of the target protein and the molecular information of the candidate drug molecule to the server. The user can also trigger the terminal to generate a drug screening request, add storage addresses of the protein information and the molecular information of the candidate drug molecule into the drug screening request, and then send the drug screening request with the added storage addresses to the server. The server acquires the storage addresses in the drug screening request, and acquires the protein information of the target protein and the molecular information of the candidate drug molecule according to the storage addresses. The protein information of the target protein and the molecular information of the candidate drug molecule corresponding to the target protein can also be directly acquired from a protein database and a drug molecule database.

[0086] 202. The server identifies protein atom information of protein atoms in the protein information and drug atom information of drug atoms in the molecular information.

[0087] For example, the server identifies number information, atom type information and position information of each protein atom of the protein atoms in the protein information, and the position information can be three-dimensional coordinate information in the target protein. These information can be used as the protein atom information. The server also identifies number information, atom type and position information of each drug atom of the drug atoms in the molecular information. These information can be used as the drug atom information.

[0088] 203. The server extracts features from the drug atom information to obtain molecular features of the candidate drug molecule.

[0089] ​For example, the server can identify, according to the atom type, a drug atom with at least one non-hydrogen atom for interaction in the drug atom, to obtain a target drug atom. The target drug atom information corresponding to the target drug atom with the non-hydrogen atom is screened out in the drug atom information. The first position information of the target drug atom is identified in the target drug atom information. The one-hot vectorization of the target drug atom in the target drug atom information can obtain the initial drug atom feature corresponding to each type of target drug atom, and a dimension vector is added to the initial drug atom feature to represent that the target drug atom comes from the drug molecule, so that the drug atom feature of the target drug atom can be obtained. The initial molecular feature of the candidate drug molecule can be obtained by fusing the drug atom features. According to the three-dimensional coordinates of the target drug atom in the candidate drug molecule and the initial molecular feature of the candidate drug molecule, the molecular point cloud feature of the candidate drug molecule can be constructed, and the molecular point cloud feature is taken as the molecular feature of the candidate drug molecule.

[0090] 204, the server extracts features from the protein atom information to obtain the protein feature of the target protein.

[0091] For example, according to the atom type, a protein atom with at least one non-hydrogen atom for interaction can be identified in the protein atom, so as to obtain a candidate protein atom. The candidate protein atom information corresponding to the candidate protein atom corresponding to the non-hydrogen atom is screened out in the candidate protein atom information. The protein position information of the candidate protein atom is identified in the candidate protein atom information. The one-hot vectorization of the candidate protein atom in the candidate protein atom information can obtain the initial protein atom feature corresponding to each type of candidate protein atom, and a dimension vector is added to the initial protein atom feature to represent that the candidate protein atom comes from the target protein, so that the protein atom feature of the candidate protein atom can be obtained. The initial protein feature of the target protein can be obtained by fusing the protein atom features. According to the protein position information and the initial protein feature, the protein point cloud feature of the target protein is constructed, and the protein point cloud feature is taken as the protein feature of the target protein.

[0092] 205, the server identifies, according to the protein information and the first position information, a target protein atom that interacts with the target drug atom under the molecular force field in the protein atom.

[0093] For example, the server identifies the three-dimensional coordinates of each protein atom in the target protein in the protein information, thereby obtaining second position information of the protein atom. The server simulates the binding of the target drug atom and the protein atom under the molecular force field, calculates the atomic distance between the target drug atom and each protein atom according to the three-dimensional coordinates of the target drug atom and the protein atom. When the type of the molecular force field is X-score, the distance threshold of the type of the molecular force field can be 8 angstroms, and the target protein atom is screened from the protein atom whose atomic distance with the target drug atom is not more than 8 angstroms as the target protein atom.

[0094] 206. The server calculates the interaction force between the target protein atom and the target drug atom under the molecular force field.

[0095] For example, when the type of the molecular force field is X-score, the type of the interaction force under the type of the molecular force field can include cumulative van der Waals force, non-polar interaction force, hydrogen bond interaction force, and polar interaction force. When the type of the interaction force is van der Waals force, the server can screen the mapping relationship between the van der Waals force and the atomic distance in the preset mapping relationship set, when the type of the interaction force is non-polar interaction force, the server can screen the mapping relationship between the non-polar interaction force and the atomic distance in the preset mapping relationship set, and so on. According to the mapping relationship, the server maps the initial interaction force corresponding to the atomic distance between the target drug atom and the target protein atom. For example, when the mapping relationship is the mapping relationship corresponding to the van der Waals force, the initial van der Waals force corresponding to the atomic distance between the target drug atom and the target protein atom can be mapped according to the atomic distance between the target drug atom and the target protein atom, when the mapping relationship is the mapping relationship corresponding to the non-polar interaction force, the initial non-polar interaction force corresponding to the atomic distance between the target drug atom and the target protein atom can be mapped, and so on. The initial van der Waals force is classified into a category, the initial non-polar interaction force is classified into a category, and so on. The cumulative van der Waals force between the target drug atom and the target protein atom corresponding to the target drug atom can be obtained by accumulating all the initial van der Waals forces, and so on. The interaction force between the target drug atom and the target protein atom can be obtained.

[0096] 207. The server determines the interaction information between the target protein and the candidate drug molecule under the molecular force field based on the molecular information and the interaction force.

[0097] For example, the server extracts at least one attribute information of the target drug atom in the molecular information, identifies the attribute information, obtains an attribute type corresponding to the attribute information, and can filter out information corresponding to the attribute type of rotatable bond information, atomic mass, and oil-water partition coefficient (logP) in the attribute information to obtain target attribute information of the target drug atom for interaction. The numerical value of the interaction force of the target protein atom and the target drug atom under the molecular force field and the target attribute information can be used as the interaction information of the target protein and the candidate drug molecule under the molecular force field.

[0098] 208. The server calculates the binding strength between the target protein and the candidate drug molecule based on the protein features, the molecular features, and the interaction information.

[0099] For example, the server can convert the interaction information into a 9-dimensional feature vector, which is used to describe the cumulative van der Waals force, non-polar interaction force, hydrogen bond interaction force, and polar interaction force of each target drug atom on the adjacent target protein atom, as well as the rotatable bond information, atomic mass, and oil-water partition coefficient of the target drug atom itself. In the protein features, the protein features corresponding to the target protein atom interacting with the target drug atom are filtered out to obtain the target protein features. The target protein features, the molecular features, and the interaction features are transformed by using a preset strength calculation model to obtain a plurality of dimensions of local binding strength features. The feature mean value of the local binding strength features is calculated, and the local binding strength features are fused based on the feature mean value, so that the global binding strength features can be obtained. The global binding strength features are processed by using a fully connected layer, and finally the binding strength between the target protein and the candidate drug molecule is output.

[0100] 209. The server screens the target drug molecule from the candidate drug molecules according to the binding strength.

[0101] For example, the server can obtain a binding strength threshold value corresponding to the target protein, compare the binding strength threshold value with the binding strength, and screen the drug molecule with the binding strength exceeding the binding strength threshold value from the candidate drug molecules as the target drug molecule. The server can also determine the interaction level of the candidate drug molecule and the target protein according to the binding strength, and screen the drug molecule with the interaction level exceeding a preset level threshold from the candidate drug molecules as the target drug molecule.

[0102] From the above, the server in this embodiment obtains the protein information of the target protein and the molecular information of the candidate drug molecule, the target protein is a target macromolecule that can be acted on by the drug molecule, then the protein information and the molecular information are extracted respectively to obtain the protein features of the target protein and the molecular features of the candidate drug molecule, the interaction information of the target protein and the candidate drug molecule under the molecular force field is determined according to the protein information and the molecular information, then the binding strength between the target protein and the candidate drug molecule is calculated based on the protein features, the molecular features and the interaction information, and the target drug molecule is screened from the candidate drug molecule according to the binding strength. Since the scheme determines the interaction information of the target protein and the candidate drug molecule under the molecular force field on the basis of the extracted protein features and molecular features, the binding information between the protein and the drug small molecule can be more accurately expressed by introducing the molecular force field and the interaction information under the molecular force field, thereby improving the accuracy of the binding strength between the protein and the drug small molecule, and therefore the accuracy of drug screening can be greatly improved.

[0103] In order to better implement the above method, the embodiment of the present application further provides a drug screening device, which can be integrated in an electronic device such as a server or a terminal device. The terminal device can include a tablet computer, a notebook computer and / or a personal computer, etc.

[0104] For example, as shown in Figure 5 The drug screening device can include an acquisition unit 301, an extraction unit 302, a determination unit 303, a calculation unit 304 and a screening unit 305, as follows:

[0105] (1) The acquisition unit 301;

[0106] The acquisition unit 301 is configured to acquire the protein information of the target protein and the molecular information of the candidate drug molecule, the target protein being a target macromolecule that can be acted on by the drug molecule.

[0107] For example, the acquisition unit 301 can be specifically configured to enable a user to generate a drug screening request by triggering a terminal, add the protein information of the target protein and the molecular information of at least one candidate drug molecule to the drug screening request, and send the drug screening request with the added target protein information and molecular information to the drug screening device, so that the drug screening device acquires the protein information of the target protein and the molecular information of the candidate drug molecule, or the terminal directly sends the protein information of the target protein and the molecular information of the candidate drug molecule to the drug screening device.

[0108] (2) The extraction unit 302;

[0109] The extraction unit 302 is configured to extract features of the protein information and the molecule information respectively, and obtain the protein features of the target protein and the molecule features of the candidate drug molecule.

[0110] For example, the extraction unit 302 can be specifically configured to identify protein atom information of protein atoms in the protein information, and identify drug atom information of drug atoms in the molecule information, extract features of the drug atom information to obtain the molecule features of the candidate drug molecule, and extract features of the protein atom information to obtain the protein features of the target protein.

[0111] (3) The determination unit 303;

[0112] The determination unit 303 is configured to determine interaction information of the target protein and the candidate drug molecule in a molecular force field according to the protein information and the molecule information.

[0113] For example, the determination unit 303 can be specifically configured to identify target protein atoms in the protein atoms that interact with the target drug atoms in the molecular force field according to the protein information and the first position information, calculate the interaction force between the target protein atoms and the target drug atoms in the molecular force field, and determine the interaction information of the target protein and the candidate drug molecule in the molecular force field based on the molecule information and the interaction force.

[0114] (4) The calculation unit 304;

[0115] The calculation unit 304 is configured to calculate the binding strength between the target protein and the candidate drug molecule based on the protein features, the molecule features and the interaction information.

[0116] For example, the calculation unit 304 can be specifically configured to extract features of the interaction information to obtain interaction features, filter the protein features corresponding to the target protein atoms in the protein features to obtain target protein features, and calculate the strength of the target protein features, the molecule features and the interaction features by using a preset strength calculation model to obtain the binding strength between the target protein and the candidate drug molecule.

[0117] (5) The screening unit 305;

[0118] The screening unit 305 is configured to screen the target drug molecule from the candidate drug molecule according to the binding strength.

[0119] For example, the screening unit 305 can be specifically used to obtain a binding strength threshold corresponding to the target protein, compare the binding strength threshold with the binding strength, screen out a drug molecule with a binding strength exceeding the binding strength threshold from the candidate drug molecules as the target drug molecule, and determine an interaction level of the candidate drug molecule and the target protein according to the binding strength, and screen out a drug molecule with an interaction level exceeding a preset level threshold from the candidate drug molecules as the target drug molecule.

[0120] In a specific implementation, the above units can be implemented as independent entities, or combined as the same or several entities, and the specific implementation of the above units can be referred to the method embodiments above, which will not be described here.

[0121] As can be seen from the above, after the obtaining unit 301 obtains the protein information of the target protein and the molecular information of the candidate drug molecule, the target protein is a target macromolecule that can be acted on by the drug molecule, then the extracting unit 302 respectively extracts the protein information and the molecular information to obtain the protein features of the target protein and the molecular features of the candidate drug molecule, the determining unit 303 determines the interaction information of the target protein and the candidate drug molecule under the molecular force field according to the protein information and the molecular information, then the calculating unit 304 calculates the binding strength between the target protein and the candidate drug molecule based on the protein features, the molecular features and the interaction information, and the screening unit 305 screens out the target drug molecule from the candidate drug molecules according to the binding strength; since the scheme further determines the interaction information of the target protein and the candidate drug molecule under the molecular force field on the basis of the extracted protein features and molecular features, the binding information between the protein and the drug small molecule can be more accurately expressed by introducing the molecular force field and the interaction information under the molecular force field, thereby improving the accuracy of the binding strength between the protein and the drug small molecule, and thus the accuracy of the drug screening can be greatly improved.

[0122] The embodiment of the present application further provides an electronic device, as shown in the figure, which shows a structural schematic diagram of the electronic device related to the embodiment of the present application, and specifically: Figure 6

[0123] The electronic device can include a processor 401 with one or more processing cores, a memory 402 with one or more computer readable storage media, a power supply 403 and an input unit 404, and the like. Those skilled in the art can understand that the structure of the electronic device shown in the figure does not constitute a limitation on the electronic device, and can include more or fewer components than the figure, or combine certain components, or different component arrangements. Among them: Figure 6

[0124] ​​The processor 401 is the control center of the electronic device, connects each part of the entire electronic device through various interfaces and lines, executes various functions of the electronic device and processes data by running or executing software programs and / or modules stored in the memory 402 and calling data stored in the memory 402, thereby overall detecting the electronic device. Optionally, the processor 401 can include one or more processing cores; preferably, the processor 401 can integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface and application program, etc., and the modem processor mainly processes wireless communication. It can be understood that the above-mentioned modem processor can also not be integrated into the processor 401.

[0125] The memory 402 can be used to store software programs and modules, and the processor 401 executes various functions and data processing by running the software programs and modules stored in the memory 402. The memory 402 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application program required by a function (such as a sound playing function, an image playing function, etc.), etc.; the data storage area can store data created according to the use of the electronic device, etc. In addition, the memory 402 can include a high-speed random access memory, and can also include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other volatile solid-state memory device. Accordingly, the memory 402 can also include a memory controller to provide access for the processor 401 to the memory 402.

[0126] The electronic device also includes a power supply 403 for powering each component, and preferably the power supply 403 can be logically connected to the processor 401 through a power management system, thereby realizing functions such as management of charging, discharging and power consumption management through the power management system. The power supply 403 can also include one or more than one direct current or alternating current power supply, a recharging system, a power failure detection circuit, a power supply converter or inverter, a power supply state indicator, etc. Any component.

[0127] The electronic device can also include an input unit 404, which can be used to receive input digital or character information, and generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function control.

[0128] Although not shown, the electronic device can also include a display unit, etc., which will not be described here. Specifically in the present embodiment, the processor 401 in the electronic device will load the executable file corresponding to the process of one or more than one application program into the memory 402 according to the following instructions, and run the application program stored in the memory 402 by the processor 401, thereby realizing various functions, as follows:

[0129] The protein information of the target protein and the molecular information of the candidate drug molecule are obtained, the target protein being a target macromolecule that can be acted on by the drug molecule, the protein information and the molecular information are respectively subjected to feature extraction to obtain protein features of the target protein and molecular features of the candidate drug molecule, the interaction information of the target protein and the candidate drug molecule under a molecular force field is determined according to the protein information and the molecular information, the binding strength between the target protein and the candidate drug molecule is calculated based on the protein features, the molecular features and the interaction information, and the target drug molecule is screened from the candidate drug molecule according to the binding strength.

[0130] For example, the user triggers the terminal to generate a drug screening request, adds the protein information of the target protein and the molecular information of at least one candidate drug molecule to the drug screening request, and sends the drug screening request with the added target protein information and molecular information to the electronic device, so that the electronic device obtains the protein information of the target protein and the molecular information of the candidate drug molecule, or the terminal directly sends the protein information of the target protein and the molecular information of the candidate drug molecule to the electronic device. The protein atom information of the protein atom is identified in the protein information, and the drug atom information of the drug atom is identified in the molecular information. The drug atom information is subjected to feature extraction to obtain the molecular features of the candidate drug molecule, and the protein atom information is subjected to feature extraction to obtain the protein features of the target protein. The target protein atom that interacts with the target drug atom under a molecular force field is identified from the protein atoms according to the protein information and the first position information, the interaction force between the target protein atom and the target drug atom under the molecular force field is calculated, the interaction information of the target protein and the candidate drug molecule under the molecular force field is determined based on the molecular information and the interaction force. The interaction features are obtained by performing feature extraction on the interaction information, the protein features corresponding to the target protein atom are screened from the protein features to obtain the target protein features, and the binding strength between the target protein and the candidate drug molecule is obtained by performing strength calculation on the target protein features, the molecular features and the interaction features by using a preset strength calculation model. The binding strength threshold value corresponding to the target protein is obtained, the binding strength threshold value is compared with the binding strength, and the drug molecule with the binding strength exceeding the binding strength threshold value is screened from the candidate drug molecule as the target drug molecule. The interaction level of the candidate drug molecule and the target protein can also be determined according to the binding strength, and the drug molecule with the interaction level exceeding a preset level threshold value is screened from the candidate drug molecule as the target drug molecule.

[0131] The specific implementation of each operation can be referred to the foregoing embodiments, which will not be repeated here.

[0132] From the above, the embodiment of the present application obtains the protein information of the target protein and the molecular information of the candidate drug molecule after the target protein is a target macromolecule that can be acted on by the drug molecule, then the protein information and the molecular information are respectively extracted to obtain the protein characteristics of the target protein and the molecular characteristics of the candidate drug molecule, the interaction information of the target protein and the candidate drug molecule under the molecular force field is determined according to the protein information and the molecular information, then the binding strength between the target protein and the candidate drug molecule is calculated based on the protein characteristics, the molecular characteristics and the interaction information, and the target drug molecule is screened from the candidate drug molecule according to the binding strength. Since the scheme determines the interaction information of the target protein and the candidate drug molecule under the molecular force field on the basis of the extracted protein characteristics and molecular characteristics, the binding information between the protein and the drug small molecule can be more accurately expressed by introducing the molecular force field and the interaction information under the molecular force field, thereby improving the accuracy of the binding strength between the protein and the drug small molecule, and therefore the accuracy of drug screening can be greatly improved.

[0133] Those skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructions or controlled by instructions related to hardware, which can be stored in a computer readable storage medium and loaded and executed by a processor.

[0134] Therefore, the embodiment of the present application provides a computer readable storage medium, which stores a plurality of instructions capable of being loaded by a processor to execute the steps in any drug screening method provided by the embodiment of the present application. For example, the instructions can execute the following steps:

[0135] The protein information of the target protein and the molecular information of the candidate drug molecule are obtained, the target protein is a target macromolecule that can be acted on by the drug molecule, the protein information and the molecular information are respectively extracted to obtain the protein characteristics of the target protein and the molecular characteristics of the candidate drug molecule, the interaction information of the target protein and the candidate drug molecule under the molecular force field is determined according to the protein information and the molecular information, the binding strength between the target protein and the candidate drug molecule is calculated based on the protein characteristics, the molecular characteristics and the interaction information, and the target drug molecule is screened from the candidate drug molecule according to the binding strength.

[0136] For example, the user triggers the terminal to generate a drug screening request, adds protein information of a target protein and molecular information of at least one candidate drug molecule to the drug screening request, and sends the drug screening request with the added protein information and molecular information to the electronic device, so that the electronic device acquires the protein information of the target protein and the molecular information of the candidate drug molecule, or the terminal directly sends the protein information of the target protein and the molecular information of the candidate drug molecule to the electronic device. Protein atom information of a protein atom is identified in the protein information, and drug atom information of a drug atom is identified in the molecular information. The drug atom information is feature-extracted to obtain molecular features of the candidate drug molecule, and the protein atom information is feature-extracted to obtain protein features of the target protein. According to the protein information and the first position information, a target protein atom that interacts with the target drug atom under a molecular force field is identified in the protein atom, interaction force between the target protein atom and the target drug atom under the molecular force field is calculated, and based on the molecular information and the interaction force, interaction information between the target protein and the candidate drug molecule under the molecular force field is determined. The interaction information is feature-extracted to obtain interaction features, target protein features corresponding to the target protein atom are screened out from the protein features to obtain target protein features, a preset strength calculation model is used to calculate the strength of the target protein features, the molecular features and the interaction features, and the binding strength between the target protein and the candidate drug molecule is obtained. A binding strength threshold value corresponding to the target protein is acquired, the binding strength threshold value is compared with the binding strength, and a drug molecule with a binding strength exceeding the binding strength threshold value is screened out from the candidate drug molecule as a target drug molecule. In addition, according to the binding strength, an interaction level of the candidate drug molecule and the target protein can be determined, and a drug molecule with an interaction level exceeding a preset level threshold value is screened out from the candidate drug molecule as a target drug molecule.

[0137] The specific implementation of each operation can refer to the foregoing embodiments, which will not be described here again.

[0138] The computer readable storage medium can include a read only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0139] Due to the instructions stored in the computer readable storage medium, the steps in any drug screening method provided by the embodiments of the present application can be executed, and thus the beneficial effects of any drug screening method provided by the embodiments of the present application can be achieved. Details can refer to the foregoing embodiments, which will not be described here again.

[0140] According to an aspect of the present application, a computer program product or computer program is provided, which includes computer instructions stored in a computer readable storage medium. A processor of a computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions, so that the computer device executes the method provided in various optional implementation manners of the above-mentioned drug screening aspect.

[0141] The above describes in detail a drug screening method, device and computer readable storage medium provided by an embodiment of the present application. The principles and implementation manners of the present application are described by applying specific examples. The above description of the embodiments is only used to help understand the method of the present application and its core idea. Meanwhile, for those skilled in the art, the specific implementation manners and application ranges will be changed according to the idea of the present application. In summary, the content of the specification should not be understood as a limitation of the present application.

Claims

1. A method of drug screening, characterized by, The method comprises the following steps: obtaining protein information of a target protein and molecular information of a candidate drug molecule, the target protein being a target macromolecule capable of being acted on by a drug molecule; identifying protein atom information of a protein atom in the protein information; identifying candidate protein atoms of at least one target atom type for interaction from the protein atoms according to atom types; screening candidate protein atom information corresponding to the candidate protein atoms from the protein atom information; and performing feature extraction on the candidate protein atom information to obtain protein features of the target protein; identifying drug atom information of a drug atom in the molecular information; identifying target drug atoms of at least one target atom type for interaction from the drug atoms according to atom types; screening target drug atom information corresponding to the target drug atoms from the drug atom information; and performing feature extraction on the target drug atom information to obtain molecular features of the candidate drug molecule; determining interaction information of the target protein and the candidate drug molecule under a molecular force field according to the protein information and the molecular information; calculating a binding strength between the target protein and the candidate drug molecule based on the protein features, the molecular features and the interaction information; screening a target drug molecule from the candidate drug molecule according to the binding strength.

2. The drug screening method according to claim 1, characterized in that, The target protein comprises at least one protein atom, and the candidate drug molecule comprises at least one drug atom.

3. The drug screening method according to claim 1, characterized in that, The feature extraction on the target drug atom information to obtain the molecular features of the candidate drug molecule comprises the following steps: identifying first position information of the target drug atom in the target drug atom information; performing feature extraction on the target drug atom information to obtain initial molecular features of the candidate drug molecule; constructing molecular point cloud features of the candidate drug molecule according to the first position information and the initial molecular features, and taking the molecular point cloud features as the molecular features of the candidate drug molecule.

4. The drug screening method according to claim 3, characterized in that, The determination of the interaction information of the target protein and the candidate drug molecule under the molecular force field according to the protein information and the molecular information comprises the following steps: identifying target protein atoms that interact with the target drug atom under the molecular force field from the protein atoms according to the protein information and the first position information; calculating interaction forces between the target protein atoms and the target drug atom under the molecular force field; determining the interaction information of the target protein and the candidate drug molecule under the molecular force field based on the molecular information and the interaction forces.

5. The drug screening method according to claim 4, characterized in that, The identification of the target protein atoms that interact with the target drug atom under the molecular force field from the protein atoms according to the protein information and the first position information comprises the following steps: identifying second position information of the protein atom in the protein information; calculating atom distances between the target drug atom and each protein atom according to the first position information and the second position information; and Screening, based on the atomic distance, target protein atoms in the protein atoms that interact with the target drug atoms under the molecular force field.

6. The drug screening method according to claim 5, characterized in that, The screening, based on the atomic distance, target protein atoms in the protein atoms that interact with the target drug atoms under the molecular force field comprises: Determining, according to the type of the molecular force field, a distance threshold for the interaction of the protein atoms with the target drug atoms; Screening, in the protein atoms, at least one protein atom with the atomic distance not exceeding the distance threshold, to obtain the target protein atoms.

7. The drug screening method according to claim 4, characterized in that, The calculation of the interaction force between the target protein atoms and the target drug atoms under the molecular force field comprises: Determining, according to the type of the molecular force field, the type of the interaction force under the molecular force field; Screening, in a preset mapping relationship set, a mapping relationship corresponding to the type of the interaction force, the mapping relationship being a mapping relationship between the interaction force and the atomic distance; Mapping, according to the mapping relationship, the interaction force between the target drug atoms and the target protein atoms under the molecular force field.

8. The drug screening method according to claim 7, characterized in that, The mapping, according to the mapping relationship, of the interaction force between the target drug atoms and the target protein atoms under the molecular force field comprises: Mapping, according to the mapping relationship, an initial interaction force corresponding to the atomic distance between the target drug atoms and the target protein atoms; Classifying the initial interaction force based on the type of the interaction force; Fusing, according to the classification result, the initial interaction force corresponding to each type of the interaction force to obtain the interaction force.

9. The drug screening method according to claim 4, characterized in that, The determination, based on the molecular information and the interaction force, of the interaction information between the target protein and the candidate drug molecule under the molecular force field comprises: Identifying, in the molecular information, target attribute information used for interaction of the target drug atoms; Determining, according to the target attribute information and the interaction force, the interaction information between the target protein and the candidate drug molecule under the molecular force field.

10. The drug screening method according to claim 9, characterized in that, The identification, in the molecular information, of the target attribute information used for interaction of the target drug atoms comprises: Extracting, in the molecular information, at least one attribute information of the target drug atoms; Identifying the attribute information to obtain an attribute type corresponding to the attribute information; Screening, according to the attribute type, target attribute information used for interaction in the attribute information.

11. The drug screening method according to claim 4, characterized in that, The calculation, based on the protein characteristics, molecular characteristics and interaction information, of the binding strength between the target protein and the candidate drug molecule comprises: Feature extraction of the interaction information to obtain interaction features; Screening, in the protein characteristics, protein characteristics corresponding to the target protein atoms to obtain target protein characteristics; Strength calculation of the target protein characteristics, molecular characteristics and interaction features by using a preset strength calculation model to obtain the binding strength between the target protein and the candidate drug molecule.

12. The drug screening method according to claim 11, characterized in that, The strength calculation model is used to calculate the target protein characteristics, molecular characteristics and interaction characteristics, and the binding strength between the target protein and the candidate drug molecule is obtained, including: The feature transformation is performed on the target protein characteristics, molecular characteristics and interaction characteristics to obtain local binding strength characteristics in multiple dimensions; The local binding strength characteristics are fused to obtain global binding strength characteristics; According to the global binding strength characteristics, the binding strength between the target protein and the candidate drug molecule is determined.

13. A drug screening device characterized by, It includes: An acquisition unit is configured to acquire protein information of a target protein and molecular information of a candidate drug molecule, wherein the target protein is a target macromolecule that can be acted on by a drug molecule; An extraction unit is configured to identify protein atom information of protein atoms in the protein information; According to the atom type, at least one target atom type of candidate protein atoms for interaction is identified in the protein atoms; The candidate protein atom information corresponding to the candidate protein atoms is screened out in the protein atom information; and the protein characteristics of the target protein are obtained by performing feature extraction on the candidate protein atom information; Drug atom information of drug atoms is identified in the molecular information; According to the atom type, at least one target atom type of target drug atoms for interaction is identified in the drug atoms; target drug atom information corresponding to the target drug atoms is screened out in the drug atom information; and the molecular characteristics of the candidate drug molecule are obtained by performing feature extraction on the target drug atom information; A determination unit is configured to determine interaction information between the target protein and the candidate drug molecule under a molecular force field according to the protein information and the molecular information; A calculation unit is configured to calculate the binding strength between the target protein and the candidate drug molecule based on the protein characteristics, the molecular characteristics and the interaction information; A screening unit is configured to screen a target drug molecule from the candidate drug molecules according to the binding strength.

14. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a plurality of instructions, which are suitable for being loaded by the processor to perform the steps in the drug screening method of any one of claims 1 to 12.

15. An electronic device, comprising: The computer readable storage medium stores a plurality of instructions, which are suitable for being loaded by the processor to perform the steps in the drug screening method of any one of claims 1 to 12.

16. A computer program product, characterised in that, The computer program product includes computer instructions stored in a computer readable storage medium; the processor of the computer device reads the computer instructions from the computer readable storage medium; the processor executes the computer instructions, so that the computer device performs the steps in the drug screening method of any one of claims 1 to 12.

Images