Molecular screening method, device, electronic device and storage medium

Abstract

The present disclosure provides a method, device, electronic device and storage medium for molecular screening, and relates to the fields of deep learning technology, graph theory technology and biological and information technology in the field of artificial intelligence technology. The method is as follows: obtain a first label map of the molecule to be screened and a second label map of the reference molecule, each molecule to be screened and the reference molecule form a molecule pair, for each molecule pair, obtain the mapping between the vertices of the first label map and the second label map and the conflict information between the mappings to generate a mapping map of the molecule pair, sample the mapping map, obtain the maximum weight fully connected subgraph of the molecule pair, and according to the maximum weight fully connected subgraph of each molecule pair, screen out the molecule to be screened with the greatest similarity to the reference molecule as the target molecule. A framework for molecular screening based on Gaussian boson sampling is proposed, so that Gaussian boson sampling can be used to efficiently implement the problem of ligand-based drug virtual screening and improve the efficiency of molecular screening.

Description

Technical Field

[0001] The present disclosure relates to the fields of deep learning technology, graph theory technology, and biology and information technology in the field of artificial intelligence technology, and in particular to a method, device, electronic device, and storage medium for molecular screening. Background Art

[0002] Bose sampling is a modern quantum optical algorithm with strong computing power. Its principle is to send the same photons into a quantum optical system and use a photon counter to measure the output state of the quantum optical system. Gaussian boson sampling, as an extension of Bose sampling, can be used to solve graph combination problems. However, how to better apply Gaussian boson sampling to solve specific problems and increase the application scenarios of Gaussian boson sampling is still a huge challenge. Among them, how to apply Gaussian boson sampling to efficiently realize ligand-based drug virtual screening and improve the effect of molecular screening has become a problem that needs to be solved urgently. Summary of the invention

[0003] Provided are a method, device, electronic equipment and storage medium for molecular screening.

[0004] According to a first aspect, a method for molecular screening is provided, comprising: obtaining a first label graph of a molecule to be screened and a second label graph of a reference molecule in a molecule set to be screened; each molecule to be screened and the reference molecule form a molecule pair, and for each molecule pair, obtaining a mapping between vertices of the first label graph and the second label graph and conflict information between the mappings, and generating a mapping graph of the molecule pair based on the conflict information between the mappings; sampling the mapping graph to obtain a maximum weight fully connected subgraph of the molecule pair; and screening out a molecule to be screened with the greatest similarity to the reference molecule from the molecule set to be screened as a target molecule according to the maximum weight fully connected subgraph of each molecule pair.

[0005] According to a second aspect, a device for molecular screening is provided, comprising: an acquisition module, used to acquire a first label map of a molecule to be screened and a second label map of a reference molecule in a molecule set to be screened; a mapping module, used to form a molecule pair with each molecule to be screened and the reference molecule, and for each molecule pair, to acquire a mapping between vertices of the first label map and the second label map and conflict information between the mappings, and to generate a mapping map of the molecule pair based on the conflict information between the mappings; a sampling module, used to sample the mapping map to acquire a maximum weight fully connected subgraph of the molecule pair; and a screening module, used to screen out a molecule to be screened with the greatest similarity to the reference molecule from the molecule set to be screened as a target molecule according to the maximum weight fully connected subgraph of each molecule pair.

[0006] According to a third aspect, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can perform the molecular screening method described in the first aspect of the present disclosure.

[0007] According to a fourth aspect, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to cause the computer to execute the molecular screening method according to the first aspect of the present disclosure.

[0008] According to a fifth aspect, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the steps of the molecular screening method according to the first aspect of the present disclosure.

[0009] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0011] Figure 1 is a schematic flow chart of a molecular screening method according to the first embodiment of the present disclosure;

[0012] Figure 2 is a schematic flow chart of a molecular screening method according to a second embodiment of the present disclosure;

[0013] Figure 3 is a schematic flow chart of a molecular screening method according to the third embodiment of the present disclosure;

[0014] Figure 4 is a schematic flow chart of a molecular screening method according to a fourth embodiment of the present disclosure;

[0015] Figure 5 is a schematic flow chart of a molecular screening method according to a fifth embodiment of the present disclosure;

[0016] Figure 6 is a schematic diagram of generating a label map of molecules in a method for molecular screening according to an embodiment of the present disclosure;

[0017] Figure 7 is a schematic flow chart of a molecular screening method according to a sixth embodiment of the present disclosure;

[0018] Figure 8is a schematic flow chart of a molecular screening method according to the seventh embodiment of the present disclosure;

[0019] Fig. 9 is an overall schematic diagram of a method for molecular screening according to an embodiment of the present disclosure;

[0020] Fig.10 is a block diagram of a molecular screening apparatus according to a first embodiment of the present disclosure;

[0021] Fig.11 is a block diagram of a molecular screening apparatus according to a second embodiment of the present disclosure;

[0022] Fig.12 The block diagram is a block diagram of an electronic device for implementing the method of the embodiment of the present disclosure. DETAILED DESCRIPTION

[0023] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0024] Artificial Intelligence (AI) is a technical science that studies and develops theories, methods, technologies and application systems for simulating, extending and expanding human intelligence. At present, AI technology has the advantages of high automation, high precision and low cost, and has been widely used.

[0025] Deep Learning (DL) is a new research direction in the field of Machine Learning (ML). It learns the inherent laws and representation levels of sample data. The information obtained in the learning process is of great help in the interpretation of data such as text, images and sounds. Its ultimate goal is to enable machines to have analytical learning capabilities like humans and to recognize data such as text, images and sounds. In terms of specific research content, it mainly includes neural network systems based on convolution operations, namely convolutional neural networks; autoencoding neural networks based on multi-layer neurons; deep belief networks that are pre-trained in the form of multi-layer autoencoding neural networks and then further optimize the weights of neural networks in combination with identification information. Deep learning has achieved many results in search technology, data mining, machine learning, machine translation, natural language processing, multimedia learning, speech, recommendation and personalization technology, and other related fields. Deep learning enables machines to imitate human activities such as audio-visual and thinking, solves many complex pattern recognition problems, and makes great progress in artificial intelligence related technologies.

[0026] Graph Theory is a branch of mathematics. It studies graphs. A graph in graph theory is a figure composed of a number of given points and lines connecting two points. This kind of graph is usually used to describe a certain relationship between certain things. Points represent things, and lines connecting two points represent the relationship between the corresponding two things.

[0027] Biotechnology is a comprehensive technology based on life sciences, which uses the characteristics and functions of organisms (or biological tissues, cells and other components) to design and construct new substances or new strains with expected performance, and combines them with engineering principles to process and produce products or provide services. Information Technology is a technology that studies the acquisition, transmission and processing of information. It is a combination of computer technology, communication technology and microelectronics technology. It is a new discipline that uses computers to process information and uses modern electronic communication technology to engage in information collection, storage, processing, utilization, and related product manufacturing, technology development, and information services. Both information technology and biotechnology are high-tech. In the new economy, the two are not in a relationship of one rising and the other falling, but complement each other and jointly promote the rapid development of the economy in the 21st century.

[0028] The molecular screening method, device, electronic device and storage medium of the embodiments of the present disclosure are described below with reference to the accompanying drawings.

[0029] Figure 1 It is a schematic flow chart of the molecular screening method according to the first embodiment of the present disclosure.

[0030] like Figure 1As shown, the molecular screening method of the embodiment of the present disclosure may specifically include the following steps:

[0031] S101, obtaining a first label map of a molecule to be screened in a molecule set to be screened and a second label map of a reference molecule.

[0032] Specifically, the execution subject of the molecular screening method of the embodiment of the present disclosure may be the molecular screening device provided by the embodiment of the present disclosure, and the molecular screening device may be a hardware device with data information processing capabilities and / or the necessary software required to drive the hardware device to work. Optionally, the execution subject may include a workstation, a server, a computer, a user terminal and other devices. Among them, the user terminal includes but is not limited to a mobile phone, a computer, an intelligent voice interaction device, a smart home appliance, a car terminal, etc.

[0033] The molecular screening method of the disclosed embodiment can be applied to realize ligand-based virtual screening (LVS), which relies on the principle that "molecules with similar structures tend to have similar properties" to screen out potential drug molecules from the molecular database.

[0034] In the disclosed embodiments, multiple molecules in a set of molecules to be screened (such as a molecular database) are used as molecules to be screened, and each molecule to be screened is compared with a reference molecule to obtain a target molecule with a high similarity to the reference molecule, wherein the molecule may be a drug molecule. In some embodiments, for each molecule to be screened in the set of molecules to be screened, a label map of the molecule to be screened (molecule A) is obtained as a first label map (G A ), and obtain the label map of the reference molecule (molecule B) as the second label map (G B ), where the relationship between the vertices and edges of the label graph corresponds to the structure of the molecule and can be viewed as an undirected graph containing the characteristics of the molecule.

[0035] S102, each molecule to be screened and the reference molecule form a molecule pair, for each molecule pair, obtain the mapping between the vertices of the first label graph and the second label graph and the conflict information between the mappings, and generate a mapping graph of the molecule pair based on the conflict information between the mappings.

[0036] In the disclosed embodiment, each molecule to be screened forms a molecule pair with a reference molecule. For each molecule pair, a mapping between the vertices of the first label graph of the molecule to be screened and the second label graph of the reference molecule in the molecule pair is obtained. For example, the first label graph G A Vertex A1 in the second label graph G BThe mapping between the vertex B1 in the first label graph and the vertex B1 in the second label graph can be understood as building a corresponding relationship between the vertex A1 and the vertex B1. It should be noted that the obtained mapping should be all possible mappings between the vertices of the first label graph and the second label graph. In addition, the conflict information between any two of all possible mappings is obtained. The conflict information can be understood as if there is a conflict between the two mappings, then the two mappings cannot exist at the same time. The embodiment of the present disclosure generates a mapping diagram (G) of the molecule pair based on the obtained mapping and the conflict information between the mappings. AB ).

[0037] S103, sampling the mapping graph to obtain the maximum weighted fully connected subgraph of the molecule pair.

[0038] In the embodiment of the present disclosure, the mapping diagram G AB It is an undirected graph, including a vertex set, an edge set and a weight set. Based on Gaussian Boson Sampling (GBS), the mapping graph of the molecular pair is sampled to obtain the maximum weighted connected subgraph of the molecular pair. The sampling post-processing can be performed through greedy contraction, local expansion and using historical sampling information to obtain a better maximum weighted fully connected subgraph.

[0039] S104, according to the maximum weighted fully connected subgraph of each molecule pair, the molecule to be screened having the greatest similarity to the reference molecule is screened from the set of molecules to be screened as the target molecule.

[0040] In the disclosed embodiment, the maximum weighted fully connected subgraph of each molecule pair is obtained, and the similarity between the molecule to be screened and the reference molecule is obtained based on these maximum weighted fully connected subgraphs. The molecule to be screened with the greatest similarity to the reference molecule is screened out from the set of molecules to be screened, and the molecule to be screened with the greatest similarity is used as the target molecule.

[0041] In summary, the method for molecular screening of the embodiment of the present disclosure obtains the first label map of the to-be-screened molecule in the to-be-screened molecule set and the second label map of the reference molecule, each to-be-screened molecule and the reference molecule form a molecule pair, for each molecule pair, obtains the mapping between the vertices of the first label map and the second label map and the conflict information between the mappings, and generates a mapping map of the molecule pair based on the conflict information between the mappings, samples the mapping map, obtains the maximum weight fully connected subgraph of the molecule pair, and screens out the molecule to be screened with the greatest similarity to the reference molecule from the to-be-screened molecule set according to the maximum weight fully connected subgraph of each molecule pair as the target molecule. The present disclosure models the molecule into an undirected graph with a label (i.e., feature), generates a mapping map of the molecule pair based on the label map, obtains the maximum weight fully connected subgraph of the mapping map using Gaussian boson sampling, obtains the similarity score of the molecule pair based on the maximum weight fully connected subgraph, and proposes a framework for molecular screening based on Gaussian boson sampling, so that Gaussian boson sampling can be applied to efficiently implement the problem of ligand-based drug virtual screening, thereby improving the efficiency of molecular screening.

[0042] Figure 2 It is a schematic flow chart of the molecular screening method according to the second embodiment of the present disclosure.

[0043] like Figure 2 As shown, in Figure 1 Based on the illustrated embodiment, the molecular screening method of the embodiment of the present disclosure may specifically include the following steps:

[0044] S201, obtaining a first label map of a molecule to be screened in a molecule set to be screened and a second label map of a reference molecule.

[0045] The above step S102 may specifically include the following steps:

[0046] S202 , performing key feature matching on each first vertex in the first label graph and each second vertex in the second label graph, respectively, to obtain a mapping between the first vertex and the second vertex.

[0047] In the embodiment of the present disclosure, the first label map G A Each vertex in is regarded as the first vertex Ai, and the second label graph G B Each vertex in is regarded as the second vertex Bj, and the key features of one of the first vertices Ai are obtained as the first key features, and the key features of one of the second vertices Bj are obtained as the second key features, and it is determined whether the attributes of the first key feature and the second key feature are the same. If there is at least one key feature whose attributes are the same, a mapping between the first vertex Ai and the second vertex Bj is generated.

[0048] It should be noted that a molecule includes a single atom or a ring composed of multiple atoms. The above-mentioned key features are the features of the atoms or rings corresponding to the vertices. There are multiple such features, which can be divided into key features and non-key features.

[0049] S203: Acquire two non-conflicting mappings from all mappings to form a mapping pair, and form an edge between the two mappings included in the mapping pair.

[0050] In the disclosed embodiment, according to the conflict information of the acquired mappings, two non-conflicting mappings are acquired from all the mappings, and the two mappings are determined as a non-conflicting mapping pair. When generating a mapping graph of a molecule pair, an edge is formed between the two mappings corresponding to the mapping pair. It can be understood that no edge is formed between conflicting mappings.

[0051] S204, obtaining a weight corresponding to the mapping.

[0052] In the embodiment of the present disclosure, for each mapping, a weight corresponding to the mapping is obtained according to the number and characteristics of atoms of the first vertex and the second vertex corresponding to the mapping.

[0053] S205 , generating a mapping graph by mapping the vertices of the mapping graph, the edges between the non-conflicting mapping pairs, and the weights of the mappings.

[0054] In the disclosed embodiment, each possible mapping is used as a vertex of a mapping graph, edges are formed between conflict-free mapping pairs, and a weight set is formed according to the weight of each mapping, thereby forming a mapping graph including a vertex set, an edge set and a weight set.

[0055] S206, sampling the mapping graph to obtain the maximum weighted fully connected subgraph of the molecule pair.

[0056] S207, according to the maximum weighted fully connected subgraph of each molecule pair, the molecule to be screened having the greatest similarity to the reference molecule is screened from the set of molecules to be screened as the target molecule.

[0057] In the embodiment of the present disclosure, step S201 is the same as step S101 in the above embodiment, and steps S206-S207 are the same as steps S103-S104, which will not be described again here.

[0058] Based on the above embodiments, Figure 3 As shown, in step S203, "obtaining two non-conflicting mappings from all mappings to form a mapping pair" may include the following steps:

[0059] S301, selecting two mappings from all mappings, and determining a first vertex and a second vertex corresponding to the two mappings respectively.

[0060] In the embodiment of the present disclosure, two mappings are randomly selected from all the mappings of the first label graph and the second label graph to determine the first vertex and the second vertex corresponding to each of the two mappings. For example, one mapping corresponds to the first vertex A1 and the second vertex B1, and the other mapping corresponds to the first vertex A2 and the second vertex B2.

[0061] S302: Obtain the first atom or the first ring represented by the first vertices corresponding to the two mappings.

[0062] In the embodiment of the present disclosure, the atom represented by the first vertex A1 (ie, the first atom and the first ring) and the atom or ring represented by A2 (ie, the first atom or the first ring) are obtained.

[0063] S303: Obtain the second atom or the second ring represented by the second vertices corresponding to the two mappings.

[0064] In the embodiment of the present disclosure, the atom represented by the second vertex B1 (ie, the second atom and the second ring) and the atom or ring represented by B2 (ie, the second atom or the second ring) are obtained.

[0065] S304: In response to no atom conflict and no distance conflict between the two first atoms or first rings, and no atom conflict and no distance conflict between the two second atoms or second rings, determine that the two mappings are two non-conflicting mappings, and form a mapping pair.

[0066] In the embodiment of the present disclosure, it is determined whether there is an atomic conflict and a distance conflict between two first atoms or first rings, and whether there is an atomic conflict and a distance conflict between two second atoms or second rings:

[0067] The atomic conflict can be determined by: performing an atomic comparison on two first atoms or first rings, and performing an atomic comparison on two second atoms or second rings; in response to the fact that the same atom does not exist in the two first atoms or first rings, and the same atom does not exist in the two second atoms or second rings, it is determined that there is no atomic conflict between the two mappings, which can be expressed by the following formula 1:

[0068]

[0069] In Formula 1, A1 and A2 respectively represent a single atom represented by two first vertices or a set of multiple atoms in a ring; B1 and B2 respectively represent a single atom represented by two second vertices or a set of multiple atoms in a ring.

[0070] The distance conflict can be determined by: obtaining a first distance between two first atoms or first rings, obtaining a second distance between two second atoms or second rings, and in response to the difference between the first distance and the second distance being less than or equal to a preset distance, determining that there is no distance conflict between the two mappings.

[0071] It is not difficult to understand that the similarity of a pair of molecules depends on the number of labels of the common substructure between the molecules, that is, the similarity of the pair of molecules is judged based on the maximum labeled common substructure (Labelled Maximum Common Substructure, LMCS for short) between the pairs of molecules. The disclosed embodiment uses GBS to obtain the maximum weighted fully connected subgraph of the mapping graph of the pair of molecules, so as to be equivalent to the maximum labeled common substructure between the pairs of molecules. To ensure that the obtained maximum weighted fully connected subgraph can effectively correspond to the common substructure, the distances between the two vertices in the same label graph corresponding to the two mappings are roughly the same as the condition of no distance conflict. For example, for two mappings, the difference between the distance between the two atoms corresponding to the first vertices of the two mappings (first distance) and the distance between the two atoms corresponding to the second vertices of the two mappings (second distance) does not exceed the preset distance. It can be expressed by the following formula 2:

[0072] |dist(A1, A2)-dist(B1, B2)|>τ

[0073] Wherein, in formula 1, A1 and A2 represent single atoms or rings represented by two first vertices, respectively; B1 and B2 represent single atoms or rings represented by two second vertices, respectively. The first distance and the second distance can be calculated based on the Euclidean distance, wherein the 3D coordinates of the atom are used as the atomic coordinates, and the 3D coordinates of the geometric center of the ring are used as the coordinates of the ring. In some embodiments, τ 设置为

[0074] Based on the above embodiment, "obtaining the weight corresponding to the mapping" in step S204 may include the following steps: for each mapping, determining the first vertex and the second vertex corresponding to the mapping, and generating the weight of the mapping according to the number and characteristics of atoms in the first vertex and the second vertex respectively, where the characteristics are the characteristics of the atoms or rings corresponding to the vertex.

[0075] Based on the above embodiments, Figure 4 As shown, the molecular screening method of the embodiment of the present disclosure also includes a process of generating a mapping weight according to the number and characteristics of atoms of the first vertex and the second vertex, which can be achieved by the following steps:

[0076] S401, obtaining an average value of the number of atoms at the first vertex and the number of atoms at the second vertex.

[0077] In the embodiment of the present disclosure, for each mapping, the average value of the number of atoms at the first vertex and the number of atoms at the second vertex corresponding to the mapping is obtained.

[0078] S402: Obtain the number of features having the same attributes or values ​​between the first vertex and the second vertex.

[0079] In the embodiment of the present disclosure, feature matching is performed on the feature corresponding to the first vertex (i.e., the feature of the atom or ring represented by the first vertex) and the feature corresponding to the second vertex (i.e., the feature of the atom or ring represented by the second vertex), wherein feature matching can be performed by comparing whether the attributes of the features or the values ​​of the features are the same to obtain the number of features with the same attributes or values ​​between the first vertex and the second vertex.

[0080] S403: Determine a weight of a mapping between the first vertex and the second vertex according to the average value and the number of features having the same attribute or value.

[0081] In the embodiment of the present disclosure, for each mapping, the weight between the first vertex and the second vertex is determined according to the obtained average value and the number of features with the same attribute or value, which can be expressed by Formula 3:

[0082]

[0083] Among them, V in formula 3 AB is the mapping graph G AB =(V AB , E AB , W AB ), m i represents the i-th mapping in the mapping graph, a represents the first vertex corresponding to the i-th mapping, A i represents a single atom or a collection of multiple atoms in a ring represented by the first vertex, b represents the second vertex corresponding to the i-th mapping, and B i represents a single atom or a collection of multiple atoms in a ring represented by the second vertex, |L a ∩L b | represents the number of features with the same attributes or values ​​between the first vertex and the second vertex.

[0084] Based on the above embodiments, Figure 5 As shown, the method for molecular screening of the embodiment of the present disclosure also includes a process of generating a label map corresponding to the molecule, wherein the molecule can be a molecule to be screened or a reference molecule, which can be achieved by the following steps:

[0085] S501, generating an undirected graph of any molecule according to the molecular structure of any molecule, wherein the vertices in the undirected graph correspond to single atoms of the molecule or a ring containing multiple atoms, and the edges in the undirected graph correspond to chemical bonds in the molecule.

[0086] In the disclosed embodiment, an undirected graph can be generated with atoms as vertices and bonds as edges. For a ring containing multiple atoms, the ring is shrunk into a single vertex. For two rings with a common bond, an auxiliary edge is added between the corresponding vertices in the graph, such as Figure 6 The label graph G is constructed by molecules A and B. A and G B .

[0087] S502, extracting features of each atom in any molecule, and generating a feature label set of any molecule based on the extracted features, wherein the feature label set includes features of a single atom and features of a ring, and the features of the ring are obtained by aggregating features of multiple atoms contained in the ring.

[0088] In the disclosed embodiment, as shown in Table 1, the features of an atom may include atomic number, implicit H-bond, formal charge, degree, etc. The features of each atom included in the molecule are extracted, the feature type of the extracted feature is identified, and the key feature label of the extracted feature is determined based on the feature type, wherein the key feature label is used to indicate whether the feature is a key feature, and a feature label set is generated based on the extracted features, the feature type of the feature, and the key feature label of the feature.

[0089] Among them, in the disclosed embodiment, the chemical characteristics and pharmaceutical characteristics of atoms are considered, and the characteristics are divided into critical (C) and non-critical (NC) based on the "critical" mechanism for easy mapping. The feature types include additive chemical characteristics, non-additive chemical characteristics and pharmaceutical characteristics, non-additive chemical characteristics and pharmaceutical characteristics are critical characteristics, and additive chemical characteristics are non-critical characteristics.

[0090] Table 1. Correspondence between atomic features and key feature labels

[0091]

[0092]

[0093] Among them, "(medicine)" is used to mark pharmaceutical characteristics.

[0094] It should be noted that the characteristics of the ring can be obtained by aggregating the characteristics of the atoms contained in the ring. For example, for additive characteristics (i.e., implicit H-bonds, formal charge, and degree), we add the corresponding values ​​of each atom in each ring. For non-additive characteristics (i.e., atomic number and attached chemical bonds), we keep a sorted list. Finally, for the other seven pharmacological characteristics, if any atom in the ring has the characteristic, we set its characteristic label to "true", if no atom has the characteristic, then set the characteristic label to "false".

[0095] S503, generating a label graph corresponding to any molecule according to the undirected graph and feature label set of any molecule, wherein any molecule is a molecule to be screened, the corresponding label graph is a first label graph, and any molecule is a reference molecule, the corresponding label graph is a second label graph.

[0096] In the embodiment of the present disclosure, two molecules are represented as A and B, and their corresponding label graphs are represented as G and B, respectively. A =(V A , E A , L A ) and G B =(V B , E B , L B ), the label graph includes the vertex set V, the edge set E and the feature label set L.

[0097] Based on the above embodiments, Figure 7 As shown, the above step S207 of "screening out the molecule to be screened having the greatest similarity to the reference molecule from the molecule set to be screened according to the maximum weighted fully connected subgraph of each molecule pair as the target molecule" may include the following steps:

[0098] S701, based on the maximum fully connected graph of the molecule pair, obtaining the similarity between the label graphs of the molecule to be screened and the reference molecule in the molecule pair.

[0099] S702, selecting the molecule to be screened with the greatest similarity to the reference molecule from the set of molecules to be screened as the target molecule.

[0100] Based on the above embodiments, Figure 8 As shown, in the above step S701, "obtaining the similarity between the label graphs of the molecule to be screened and the reference molecule in the molecule pair based on the maximum fully connected graph of the molecule pair" may include the following steps:

[0101] S801: Obtain a first total weight of a first label graph.

[0102] In the disclosed embodiment, the sum of the weights of all vertices in the first label graph is taken as the first total weight, wherein the weight of each vertex can be determined by the number of features corresponding to the vertex and the number of atoms corresponding to the vertex.

[0103] S802: Obtain a second total weight of the second label graph.

[0104] In the disclosed embodiment, the sum of the weights of all vertices in the second label graph is taken as the second total weight, wherein the weight of each vertex can be determined by the number of features corresponding to the vertex and the number of atoms corresponding to the vertex.

[0105] S803: Determine a third total weight of the maximum weight fully connected subgraph based on the weights of the vertices included in the maximum weight fully connected subgraph.

[0106] S804: Determine the similarity between the first label and the second label graph according to the first total weight, the second total weight, and the third total weight.

[0107] In the embodiment of the present disclosure, the average value of the Bunke and Shearer similarity metrics may be used as the similarity score, which may be expressed by Formula 4:

[0108]

[0109] Among them, w v Represented as the weight w of the vertex in the label graph v =|L v |+|v|, where |v| is the number of atoms contained in vertex v, M represents the vertex set of the maximum weight fully connected subgraph, ω mi and ω in Formula 3 mi The same represents the weight of the vertex in the maximum weight fully connected subgraph. When A and B are the same molecule, Sim(A, B) = 1. When , Sim(A, B) = 0. Otherwise, Sim(A, B)∈(0, 1). Therefore, it can correctly characterize the similarity of molecules.

[0110] In summary, the method for molecular screening of the embodiment of the present disclosure obtains the first label map of the to-be-screened molecule in the to-be-screened molecule set and the second label map of the reference molecule, each to-be-screened molecule and the reference molecule form a molecule pair, for each molecule pair, obtains the mapping between the vertices of the first label map and the second label map and the conflict information between the mappings, and generates a mapping map of the molecule pair based on the conflict information between the mappings, samples the mapping map, obtains the maximum weight fully connected subgraph of the molecule pair, and screens out the molecule to be screened with the greatest similarity to the reference molecule from the to-be-screened molecule set according to the maximum weight fully connected subgraph of each molecule pair as the target molecule. The present disclosure models the molecule into an undirected graph with a label (i.e., feature), generates a mapping map of the molecule pair based on the label map, obtains the maximum weight fully connected subgraph of the mapping map using Gaussian boson sampling, obtains the similarity score of the molecule pair based on the maximum weight fully connected subgraph, and proposes a framework for molecular screening based on Gaussian boson sampling, so that Gaussian boson sampling can be applied to efficiently implement the problem of ligand-based drug virtual screening, thereby improving the efficiency of molecular screening.

[0111] In order to clearly describe the molecular screening method of the present disclosure, Fig. 9 To elaborate, Fig. 9 is an overall schematic diagram of the molecular screening method according to an embodiment of the present disclosure, such as Fig. 9 As shown, molecule A is used as the molecule to be screened, and molecule B is used as the reference molecule. The first label graph G corresponding to the molecule to be screened is generated according to the structure of the molecule. A The second label graph G corresponding to the reference molecule B B , according to the first label graph G A and the second label graph G B Establish a mapping graph G corresponding to the molecule to be screened A and the reference molecule B AB , based on the GBS algorithm to generate the mapping graph G AB The maximum weight clique (MWC) is based on the maximum weight fully connected subgraph of the mapping graph, and the target atoms (such as Fig. 9 The gray solid circle shown in FIG. 4 is mapped back to the first label map and the second label map to determine the common substructure of the two molecules, thereby generating a similarity score (ie, similarity) between the first label map and the second label map.

[0112] Fig.10 is a block diagram of a molecular screening apparatus according to the first embodiment of the present disclosure.

[0113] like Fig.10 As shown, the molecular screening device 1000 of the embodiment of the present disclosure includes: an acquisition module 1001, a mapping module 1002, a sampling module 1003 and a screening module 1004.

[0114] The acquisition module 1001 is used to acquire a first label map of a molecule to be screened and a second label map of a reference molecule in a molecule set to be screened.

[0115] The mapping module 1002 is used to form a molecule pair with each molecule to be screened and a reference molecule, obtain the mapping between the vertices of the first label graph and the second label graph and the conflict information between the mappings for each molecule pair, and generate a mapping graph of the molecule pair based on the conflict information between the mappings.

[0116] The sampling module 1003 is used to sample the mapping graph to obtain the maximum weighted fully connected subgraph of the molecule pair.

[0117] The screening module 1004 is used to screen out the molecule to be screened with the greatest similarity to the reference molecule from the molecule set to be screened according to the maximum weighted fully connected subgraph of each molecule pair as the target molecule.

[0118] It should be noted that the above explanation of the embodiment of the method for molecular screening is also applicable to the device for molecular screening of the embodiment of the present disclosure, and the specific process will not be repeated here.

[0119] In summary, the molecular screening device of the embodiment of the present disclosure obtains the first label map of the molecules to be screened in the set of molecules to be screened and the second label map of the reference molecules, each molecule to be screened and the reference molecule form a molecule pair, for each molecule pair, obtains the mapping between the vertices of the first label map and the second label map and the conflict information between the mappings, and generates a mapping map of the molecule pair based on the conflict information between the mappings, samples the mapping map, obtains the maximum weight fully connected subgraph of the molecule pair, and screens out the molecule to be screened with the greatest similarity to the reference molecule from the set of molecules to be screened according to the maximum weight fully connected subgraph of each molecule pair as the target molecule. The present disclosure models the molecule into an undirected graph with a label (i.e., feature), generates a mapping map of the molecule pair based on the label map, obtains the maximum weight fully connected subgraph of the mapping map using Gaussian boson sampling, obtains the similarity score of the molecule pair based on the maximum weight fully connected subgraph, and proposes a framework for molecular screening based on Gaussian boson sampling, so that Gaussian boson sampling can be applied to efficiently implement the problem of ligand-based drug virtual screening, thereby improving the efficiency of molecular screening.

[0120] Fig.11 is a block diagram of a molecular screening apparatus according to a second embodiment of the present disclosure.

[0121] like Fig.11 As shown, the molecular screening device 1100 of the embodiment of the present disclosure includes: an acquisition module 1101, a mapping module 1102, a sampling module 1103 and a screening module 1104.

[0122] Among them, the acquisition module 1101 has the same structure and function as the acquisition module 1001 in the previous embodiment, the mapping module 1102 has the same structure and function as the mapping module 1002 in the previous embodiment, the sampling module 1103 has the same structure and function as the sampling module 1003 in the previous embodiment, and the screening module 1104 has the same structure and function as the screening module 1004 in the previous embodiment.

[0123] Furthermore, the acquisition module 1101 includes: a first generation submodule 11011, which is used to generate an undirected graph of any molecule according to the molecular structure of any molecule, wherein the vertices in the undirected graph correspond to single atoms of the molecule or a ring containing multiple atoms, and the edges in the undirected graph correspond to chemical bonds in the molecule; a second generation submodule 11012, which extracts the features of each atom in any molecule, and generates a feature label set of any molecule based on the extracted features, wherein the feature label set includes features of single atoms and features of the ring, and the features of the ring are obtained by aggregating features of multiple atoms contained in the ring; a third generation submodule 11013, which is used to generate a label graph corresponding to any molecule according to the undirected graph and feature label set of any molecule, wherein if any molecule is a molecule to be screened, the corresponding label graph is a first label graph, and if any molecule is a reference molecule, the corresponding label graph is a second label graph.

[0124] Furthermore, the second generation submodule 11012 includes: a first determination unit, used to identify the feature type of the extracted feature, and determine the key feature tag of the extracted feature based on the feature type, and the key feature tag indicates whether the feature is a key feature; a second generation unit, used to generate a feature tag set based on the extracted feature, the feature type of the feature and the key feature tag of the feature.

[0125] Furthermore, the molecule is a drug molecule, and the feature types include additive chemical features, non-additive chemical features, and pharmaceutical features, the non-additive chemical features and pharmaceutical features are key features, and the additive chemical features are non-key features.

[0126] Furthermore, the mapping module 1102 includes: a feature matching submodule, which is used to match key features of each first vertex in the first label graph with each second vertex in the second label graph, respectively, to obtain a mapping between the first vertex and the second vertex; a fourth generation submodule, which is used to obtain two non-conflicting mappings from all mappings to form a mapping pair, and form an edge between the two mappings included in the mapping pair; a first acquisition submodule, which is used to obtain weights corresponding to the mappings; and a fifth generation submodule, which is used to generate a mapping graph with the mappings as the vertices of the mapping graph, the edges between the non-conflicting mapping pairs, and the weights of the mappings.

[0127] Furthermore, the fourth generation submodule includes: a second determination unit, used to select two mappings from all the mappings, and determine the first vertex and the second vertex corresponding to each of the two mappings; a first acquisition unit, used to acquire the first atom or the first ring represented by the first vertex corresponding to each of the two mappings; a second acquisition unit, used to acquire the second atom or the second ring represented by the second vertex corresponding to each of the two mappings; and a third determination unit, used to determine that the two mappings are two non-conflicting mappings in response to the fact that there is no atomic conflict and distance conflict between the two first atoms or the first rings, and there is no atomic conflict and distance conflict between the two second atoms or the second rings, and to form a mapping pair.

[0128] Furthermore, the third determination unit includes: an atomic comparison subunit, used to atomically compare the two first atoms or first rings, and to atomically compare the two second atoms or second rings; a second determination subunit, used to determine that there is no atomic conflict between the two mappings in response to the fact that the same atom does not exist in the two first atoms or first rings, and the same atom does not exist in the two second atoms or second rings.

[0129] Furthermore, the third determination unit includes: a first acquisition subunit, used to acquire a first distance between two first atoms or first rings; a second acquisition subunit, used to acquire a second distance between two second atoms or second rings; and a second determination subunit, used to determine that there is no distance conflict between the two mappings in response to the difference between the first distance and the second distance being less than or equal to a preset distance.

[0130] Furthermore, the feature matching submodule includes: a third acquisition unit, used to acquire a first key feature of one of the first vertices, and a second key feature of one of the second vertices; a judgment unit, used to judge whether the attributes of the first key feature and the second key feature are the same; and a second generation unit, used to generate a mapping between one of the first vertices and one of the second vertices if there is at least one key feature with the same attribute.

[0131] Furthermore, the acquisition submodule includes: a third generation unit, which is used to determine the first vertex and the second vertex corresponding to each mapping, and generate the weight of the mapping according to the number and characteristics of atoms in the first vertex and the second vertex respectively, where the characteristics are the characteristics of the atoms or rings corresponding to the vertices.

[0132] Furthermore, the third generation unit includes: a third acquisition subunit, used to obtain the average number of atoms of the first vertex and the average number of atoms of the second vertex; a fourth acquisition subunit, used to obtain the number of features with the same attributes or values ​​between the first vertex and the second vertex; and a third determination unit, used to determine the weight of the mapping between the first vertex and the second vertex based on the average value and the number of features with the same attributes or values.

[0133] Furthermore, the sampling module 1103 includes: a sampling submodule, which is used to sample the mapping graph based on a Gaussian boson sampling algorithm to obtain the maximum weight fully connected subgraph.

[0134] Furthermore, the screening module 1104 includes: a second acquisition submodule, used to obtain the similarity between the label graphs of the to-be-screened molecule and the reference molecule in the molecule pair based on the maximum fully connected graph of the molecule pair; and a screening submodule, used to screen out the to-be-screened molecule with the greatest similarity to the reference molecule from the to-be-screened molecule set as the target molecule.

[0135] Furthermore, the second acquisition submodule includes: a fourth acquisition unit, used to obtain a first total weight of the first label graph; a fifth acquisition unit, used to obtain a second total weight of the second label graph; a third determination unit, used to determine the third total weight of the maximum weight fully connected subgraph based on the weights of the vertices included in the maximum weight fully connected subgraph; and a fourth determination unit, used to determine the similarity between the first label and the second label graph based on the first total weight, the second total weight and the third total weight.

[0136] In summary, the molecular screening device of the embodiment of the present disclosure obtains the first label map of the molecules to be screened in the set of molecules to be screened and the second label map of the reference molecules, each molecule to be screened and the reference molecule form a molecule pair, for each molecule pair, obtains the mapping between the vertices of the first label map and the second label map and the conflict information between the mappings, and generates a mapping map of the molecule pair based on the conflict information between the mappings, samples the mapping map, obtains the maximum weight fully connected subgraph of the molecule pair, and screens out the molecule to be screened with the greatest similarity to the reference molecule from the set of molecules to be screened according to the maximum weight fully connected subgraph of each molecule pair as the target molecule. The present disclosure models the molecule into an undirected graph with a label (i.e., feature), generates a mapping map of the molecule pair based on the label map, obtains the maximum weight fully connected subgraph of the mapping map using Gaussian boson sampling, obtains the similarity score of the molecule pair based on the maximum weight fully connected subgraph, and proposes a framework for molecular screening based on Gaussian boson sampling, so that Gaussian boson sampling can be applied to efficiently implement the problem of ligand-based drug virtual screening, thereby improving the efficiency of molecular screening.

[0137] In the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0138] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

[0139] Fig.12A schematic block diagram of an example electronic device 1200 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0140] like Fig.12 As shown, the electronic device 1200 includes a computing unit 1201, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 1202 or a computer program loaded from a storage unit 1208 into a random access memory (RAM) 1203. In the RAM 1203, various programs and data required for the operation of the electronic device 1200 can also be stored. The computing unit 1201, the ROM 1202, and the RAM 1203 are connected to each other via a bus 1204. An input / output (I / O) interface 1205 is also connected to the bus 1204.

[0141] Multiple components in the electronic device 1200 are connected to the I / O interface 1205, including: an input unit 1206, such as a keyboard, a mouse, etc.; an output unit 1207, such as various types of displays, speakers, etc.; a storage unit 1208, such as a disk, an optical disk, etc.; and a communication unit 1209, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 1209 allows the electronic device 1200 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0142] The computing unit 1201 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 1201 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The computing unit 1201 performs the various methods and processes described above, such as Figures 1 to 9The method of molecular screening shown. For example, in some embodiments, the method of molecular screening can be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 1208. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 1200 via ROM 1202 and / or communication unit 1209. When the computer program is loaded into RAM 1203 and executed by the computing unit 1201, one or more steps of the semantic parsing method described above can be performed. Alternatively, in other embodiments, the computing unit 1201 can be configured to perform the method of molecular screening by any other appropriate means (e.g., by means of firmware).

[0143] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0144] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0145] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0146] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0147] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), the Internet, and a blockchain network.

[0148] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship between the client and the server is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services ("Virtual Private Server", or "VPS" for short). The server may also be a server of a distributed system, or a server combined with a blockchain.

[0149] According to an embodiment of the present disclosure, the present disclosure further provides a computer program product, including a computer program, wherein the computer program implements the steps of the molecular screening method shown in the above embodiment of the present disclosure when executed by a processor.

[0150] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0151] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A method for molecular screening, comprising: Obtaining a first label graph of the to-be-screened molecule in the to-be-screened molecule set and a second label graph of the reference molecule, wherein each label graph is an undirected graph including features of the molecule corresponding to the label graph, a vertex in the undirected graph corresponds to a single atom of the molecule or a ring containing multiple atoms, and an edge in the undirected graph corresponds to a chemical bond in the molecule; Each of the molecules to be screened and the reference molecule form a molecule pair, and for each of the molecule pairs, the mapping between the vertices of the first label graph and the second label graph and the conflict information between the mappings are obtained, and a mapping graph of the molecule pair is generated based on the mapping and the conflict information between the mappings; wherein the mapping between the vertices is obtained by matching key features between the vertices; the conflict information between the mappings is determined by the atomic conflict and distance conflict between the first atoms or the first rings represented by the first vertices corresponding to the two mappings, and the atomic conflict and distance conflict between the second atoms or the second rings represented by the second vertices corresponding to the two mappings, the first vertex being the vertex in the first label graph, and the second vertex being the vertex in the second label graph; Sampling the mapping graph to obtain a maximum weighted fully connected subgraph of the molecule pair; According to the maximum weighted fully connected subgraph of each of the molecular pairs, the molecule to be screened having the greatest similarity to the reference molecule is screened out from the set of molecules to be screened as the target molecule.

2. The method according to claim 1, wherein: Obtaining a label map corresponding to any molecule, wherein the any molecule is the molecule to be screened or the reference molecule, comprises: According to the molecular structure of any one of the molecules, an undirected graph of the any one of the molecules is generated, wherein the vertices in the undirected graph correspond to single atoms or rings containing multiple atoms of the molecule, and the edges in the undirected graph correspond to chemical bonds in the molecule; Extracting features of each atom in any one of the molecules, and generating a feature tag set of any one of the molecules based on the extracted features, wherein the feature tag set includes features of the single atom and features of the ring, and the features of the ring are obtained by aggregating features of multiple atoms contained in the ring; A label graph corresponding to any one of the molecules is generated according to the undirected graph and the feature label set of any one of the molecules, wherein any one of the molecules is the molecule to be screened, the corresponding label graph is the first label graph, any one of the molecules is the reference molecule, and the corresponding label graph is the second label graph.

3. The method according to claim 2, wherein: The step of generating a feature tag set of any molecule based on the extracted features comprises: Identifying a feature type of the extracted feature, and determining a key feature tag of the extracted feature based on the feature type, wherein the key feature tag indicates whether the feature is a key feature; The feature tag set is generated based on the extracted features, the feature types of the features and the key feature tags of the features.

4. The method according to claim 3, wherein: The molecule is a drug molecule, and the feature types include additive chemical features, non-additive chemical features and pharmaceutical features, the non-additive chemical features and pharmaceutical features are key features, and the additive chemical features are non-key features.

5. The method according to any one of claims 1 to 4, wherein: The step of acquiring, for each of the molecular pairs, mappings between vertices of the first label graph and the second label graph and conflict information between the mappings, and generating a mapping graph of the molecular pairs based on the mappings and the conflict information between the mappings comprises: Match each first vertex in the first label graph with each second vertex in the second label graph for key features, so as to obtain a mapping between the first vertex and the second vertex; Obtain two non-conflicting mappings from all the mappings to form a mapping pair, and form an edge between the two mappings included in the mapping pair; Obtaining a weight corresponding to the mapping; The mapping graph is generated with the mappings as vertices of the mapping graph, the edges between the conflict-free mapping pairs, and the weights of the mappings.

6. The method according to claim 5, wherein: The step of obtaining two non-conflicting mappings from all the mappings to form a mapping pair includes: Select two mappings from all the mappings, and determine the first vertex and the second vertex corresponding to the two mappings respectively; Obtaining the first atom or the first ring represented by the first vertices corresponding to the two mappings respectively; Obtaining a second atom or a second ring represented by a second vertex corresponding to each of the two mappings; In response to no atom conflict and no distance conflict between the two first atoms or first rings, and no atom conflict and no distance conflict between the two second atoms or second rings, the two mappings are determined to be two non-conflicting mappings, and the mapping pair is formed.

7. The method according to claim 6, wherein: The determination of the atomic conflict comprises: Performing an atomic alignment on the two first atoms or first rings, and performing an atomic alignment on the two second atoms or second rings; In response to the same atom not existing in two of the first atoms or the first ring, and the same atom not existing in two of the second atoms or the second ring, it is determined that there is no atom conflict between the two mappings.

8. The method according to claim 6, wherein: The determination of the distance conflict includes: Obtaining a first distance between two of the first atoms or first rings; obtaining a second distance between two of the second atoms or second rings; In response to a difference between the first distance and the second distance being less than or equal to a preset distance, it is determined that there is no distance conflict between the two mappings.

9. The method according to claim 5, wherein: The process of generating the mapping based on key feature matching includes: Obtaining a first key feature of one of the first vertices and a second key feature of one of the second vertices; Determining whether the attributes of the first key feature and the second key feature are the same; If there is at least one key feature with the same attribute, the mapping between the one of the first vertices and the one of the second vertices is generated.

10. The method according to claim 5, wherein: The obtaining the weight corresponding to the mapping includes: For each of the mappings, a first vertex and a second vertex corresponding to the mapping are determined, and a weight of the mapping is generated according to the number of atoms and features of each of the first vertex and the second vertex, where the features are features of atoms or rings corresponding to the vertex.

11. The method according to claim 10, wherein: The step of generating the weight of the mapping according to the number and characteristics of atoms of the first vertex and the second vertex respectively includes: Obtaining an average value of the number of atoms at the first vertex and the number of atoms at the second vertex; Obtain the number of features having the same attributes or values ​​between the first vertex and the second vertex; A weight of the mapping between the first vertex and the second vertex is determined according to the average value and the number of features having the same attribute or value.

12. The method according to claim 1, wherein: The step of sampling the mapping graph to obtain a maximum weighted fully connected subgraph of the molecule pair includes: The mapping graph is sampled based on a Gaussian boson sampling algorithm to obtain the maximum weighted fully connected subgraph.

13. The method according to claim 1, wherein: The step of screening out the molecule to be screened having the greatest similarity to the reference molecule from the set of molecules to be screened according to the maximum weighted fully connected subgraph of each molecule pair as the target molecule comprises: Based on the maximum fully connected graph of the molecule pair, obtaining the similarity between the label graphs of the molecule to be screened and the reference molecule in the molecule pair; The molecule to be screened having the greatest similarity to the reference molecule is screened out from the set of molecules to be screened as the target molecule.

14. The method according to claim 13, wherein: The obtaining the similarity between the label graphs of the molecule to be screened and the reference molecule in the molecule pair based on the maximum fully connected graph of the molecule pair includes: Obtaining a first total weight of the first label graph; Obtaining a second total weight of the second label graph; Determining a third total weight of the maximum weight fully connected subgraph based on the weights of the vertices included in the maximum weight fully connected subgraph; The similarity between the first label and the second label graph is determined according to the first total weight, the second total weight and the third total weight.

15. A molecular screening device, comprising: An acquisition module, used to acquire a first label graph of the to-be-screened molecule in the to-be-screened molecule set and a second label graph of the reference molecule, wherein each label graph is an undirected graph including features of the molecule corresponding to the label graph, a vertex in the undirected graph corresponds to a single atom of the molecule or a ring containing multiple atoms, and an edge in the undirected graph corresponds to a chemical bond in the molecule; A mapping module, used for each of the molecules to be screened and the reference molecule to form a molecule pair, for each of the molecule pairs, obtaining the mapping between the vertices of the first label map and the second label map and the conflict information between the mappings, and generating a mapping map of the molecule pair based on the mapping and the conflict information between the mappings; wherein the mapping between the vertices is obtained by matching key features between the vertices; the conflict information between the mappings is determined by the atomic conflict and distance conflict between the first atoms or the first rings represented by the first vertices corresponding to the two mappings, and the atomic conflict and distance conflict between the second atoms or the second rings represented by the second vertices corresponding to the two mappings, the first vertex is a vertex in the first label map, and the second vertex is a vertex in the second label map; A sampling module, used for sampling the mapping graph to obtain a maximum weighted fully connected subgraph of the molecule pair; The screening module is used to screen out the molecule to be screened with the greatest similarity to the reference molecule from the molecule set to be screened according to the maximum weighted fully connected subgraph of each molecule pair as the target molecule.

16. The device according to claim 15, wherein: The acquisition module comprises: A first generating submodule is used to generate an undirected graph of any molecule according to the molecular structure of any molecule, wherein the vertices in the undirected graph correspond to single atoms of the molecule or a ring containing multiple atoms, and the edges in the undirected graph correspond to chemical bonds in the molecule, and the any molecule is the molecule to be screened or the reference molecule; A second generation submodule extracts features of each atom in any one of the molecules, and generates a feature tag set of any one of the molecules based on the extracted features, wherein the feature tag set includes features of the single atom and features of the ring, and the features of the ring are obtained by aggregating features of multiple atoms contained in the ring; The third generating submodule is used to generate a label graph corresponding to any molecule according to the undirected graph and feature label set of any molecule, wherein any molecule is the molecule to be screened, the corresponding label graph is the first label graph, and any molecule is the reference molecule, and the corresponding label graph is the second label graph.

17. The device according to claim 16, wherein: The second generation submodule includes: A first determining unit, configured to identify a feature type of the extracted feature, and determine a key feature tag of the extracted feature based on the feature type, wherein the key feature tag indicates whether the feature is a key feature; The second generating unit is used to generate the feature tag set based on the extracted features, feature types of the features and key feature tags of the features.

18. The device according to claim 17, wherein: The molecule is a drug molecule, and the feature types include additive chemical features, non-additive chemical features and pharmaceutical features, the non-additive chemical features and pharmaceutical features are key features, and the additive chemical features are non-key features.

19. The device according to any one of claims 15 to 18, wherein: The mapping module comprises: A feature matching submodule, used for matching each first vertex in the first label graph with each second vertex in the second label graph by key features, so as to obtain a mapping between the first vertex and the second vertex; A fourth generating submodule, configured to obtain two non-conflicting mappings from all the mappings to form a mapping pair, and to form an edge between the two mappings included in the mapping pair; A first acquisition submodule, used to acquire a weight corresponding to the mapping; A fifth generating submodule is used to generate the mapping graph with the mappings as vertices of the mapping graph, the edges between the conflict-free mapping pairs and the weights of the mappings.

20. The device according to claim 19, wherein The fourth generation submodule comprises: A second determining unit, configured to select two mappings from all the mappings, and determine first vertices and second vertices corresponding to the two mappings respectively; A first acquisition unit, used for acquiring a first atom or a first ring represented by a first vertex corresponding to each of the two mappings; A second acquisition unit, used for acquiring second atoms or second rings represented by second vertices corresponding to the two mappings; The third determination unit is used to determine that the two mappings are two non-conflicting mappings and form the mapping pair in response to the fact that there is no atomic conflict and distance conflict between the two first atoms or first rings and there is no atomic conflict and distance conflict between the two second atoms or second rings.

21. The device according to claim 20, wherein: The third determining unit includes: an atomic alignment subunit, used for atomically aligning two of the first atoms or first rings, and atomically aligning two of the second atoms or second rings; The second determining subunit is configured to determine that there is no atomic conflict between the two mappings in response to the fact that the same atom does not exist in the two first atoms or the first ring and the same atom does not exist in the two second atoms or the second ring.

22. The device according to claim 20, wherein: The third determining unit includes: A first acquisition subunit, used for acquiring a first distance between two of the first atoms or first rings; A second acquisition subunit, used for acquiring a second distance between two of the second atoms or second rings; The second determining subunit is configured to determine that there is no distance conflict between the two mappings in response to a difference between the first distance and the second distance being less than or equal to a preset distance.

23. The device according to claim 19, wherein: The feature matching submodule includes: A third acquisition unit, used to acquire a first key feature of one of the first vertices and a second key feature of one of the second vertices; A judging unit, configured to judge whether the attributes of the first key feature and the second key feature are the same; The second generating unit is configured to generate the mapping between the one of the first vertices and the one of the second vertices if there is at least one key feature with the same attribute.

24. The device according to claim 19, wherein The acquisition submodule includes: The third generating unit is used to determine, for each of the mappings, a first vertex and a second vertex corresponding to the mapping, and generate a weight of the mapping according to the number of atoms and features of each of the first vertex and the second vertex, where the features are features of the atoms or rings corresponding to the vertex.

25. The device according to claim 24, wherein: The third generating unit comprises: A third acquisition subunit, used to acquire an average value of the number of atoms at the first vertex and the number of atoms at the second vertex; a fourth acquisition subunit, configured to acquire the number of features having the same attributes or values ​​between the first vertex and the second vertex; The third determining unit is used to determine the weight of the mapping between the first vertex and the second vertex according to the average value and the number of features having the same attribute or value.

26. The device according to claim 15, wherein: The sampling module comprises: The sampling submodule is used to sample the mapping graph based on a Gaussian boson sampling algorithm to obtain the maximum weight fully connected subgraph.

27. The device according to claim 15, wherein: The screening module comprises: A second acquisition submodule is used to acquire the similarity between the label graphs of the molecule to be screened and the reference molecule in the molecule pair based on the maximum fully connected graph of the molecule pair; The screening submodule is used to screen out the molecule to be screened with the greatest similarity to the reference molecule from the set of molecules to be screened as the target molecule.

28. The device according to claim 27, wherein The second acquisition submodule includes: A fourth acquisition unit, configured to acquire a first total weight of the first label graph; A fifth acquiring unit, configured to acquire a second total weight of the second label graph; A third determining unit, configured to determine a third total weight of the maximum weight fully connected subgraph based on the weights of the vertices included in the maximum weight fully connected subgraph; A fourth determining unit is used to determine the similarity between the first label and the second label map according to the first total weight, the second total weight and the third total weight.

29. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 14.

30. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-14.

31. A computer program product comprising a computer program, which, when executed by a processor, implements the steps of the method according to any one of claims 1 to 14.

Images