Graph neural network construction method and device, electronic equipment and storage medium

Abstract

The invention provides a graph neural network construction method and device, electronic equipment and a storage medium, and relates to the technical field of biological information. The method comprises the following steps: encoding single cell data in a single cell data set into a low-dimensional submerged space to obtain a first feature vector which retains key features and / or type specificity information of the single cell data; obtaining a constructed gene regulatory network, and encoding the gene regulatory network into low-dimensional-factor nodes; under the condition that the global network structure of the gene regulation and control network is reserved, extracting local topological features of the low-dimensional factor nodes to obtain a second feature vector; the second feature vector is used for reflecting interaction among all genes in the single cell data set; constructing a graph neural network based on the first feature vector and the second feature vector; and the graph neural network is used for automatically labeling the single cell data. The problems that cell labeling is inaccurate, time and labor are wasted, and the cell labeling is limited by the batch effect in the prior art are solved.

Description

Technical Field

[0001] The present application relates to the field of bioinformatics, and more specifically, to a graph neural network construction method, device, electronic device and storage medium. Background Art

[0002] Cell annotation is a key step in processing single-cell data. Accurate cell type annotation is the basis for cell communication, type-specific expression patterns, differentiation trajectories, and other downstream analyses.

[0003] At present, cell annotation is based on single-cell transcriptome sequencing. Single-cell transcriptome sequencing (scRNA-seq) quantifies gene expression profiles at high resolution, which is conducive to a deeper understanding of the molecular basis of cell heterogeneity and tissue phenotypic differences, allowing us to identify various cell types at single-cell resolution and analyze differences in cell expression profiles. However, there are still technical problems such as inaccurate cell annotation, time-consuming and labor-intensive, and limited by batch effects.

[0004] As can be seen from the above, the prior art has technical problems such as inaccurate cell labeling, time-consuming and labor-intensive, and limited by batch effects. Summary of the invention

[0005] The present application provides a graph neural network construction method, device, electronic device and storage medium, which can solve the problems of inaccurate cell labeling, time-consuming and labor-intensive, and limited by batch effects in related technologies. The technical solution is as follows:

[0006] According to one aspect of the present application, a method for constructing a graph neural network includes: encoding single-cell data in a single-cell dataset into a low-dimensional latent space to obtain a first eigenvector that retains key features and / or type-specific information of the single-cell data; obtaining a constructed gene regulatory network, and encoding the gene regulatory network into low-dimensional gene nodes; while retaining the global network structure of the gene regulatory network, extracting local topological features of the low-dimensional gene nodes to obtain a second eigenvector; the second eigenvector is used to reflect the interactions between all genes in the single-cell dataset; based on the first eigenvector and the second eigenvector, constructing a graph neural network; the graph neural network is used to automatically annotate single-cell data.

[0007] According to one aspect of the present application, a graph neural network construction device includes: a single-cell data encoding module, which is used to encode single-cell data in a single-cell data set into a low-dimensional latent space to obtain a first feature vector that retains the key features and / or type-specific information of the single-cell data; a gene regulatory network encoding module, which is used to obtain a constructed gene regulatory network and encode the gene regulatory network into a low-dimensional gene node; a local topological feature extraction module, which is used to extract local topological features of the low-dimensional gene nodes while retaining the global network structure of the gene regulatory network to obtain a second feature vector; the second feature vector is used to reflect the interactions between all genes in the single-cell data set; a graph neural network construction module, which is used to construct a graph neural network based on the first feature vector and the second feature vector; the graph neural network is used to automatically annotate single-cell data.

[0008] In an exemplary embodiment, the graph neural network construction module includes: a gene node initial feature acquisition unit, used to use the second feature vector as the initial feature of the gene node in the graph neural network; a cell node initial feature acquisition unit, used to use the first feature vector as the initial feature of the cell node in the graph neural network; an edge construction unit, used to construct the edge of the graph neural network between the gene node and the cell node; and a graph neural network acquisition unit, used to obtain the graph neural network from the edges between the gene nodes and the cell nodes.

[0009] In an exemplary embodiment, the edges of the graph neural network include at least one of the following: gene regulation edges connecting gene nodes; self-feedback edges pointing from gene nodes or cell nodes to themselves; and gene expression edges connecting between gene nodes and cell nodes.

[0010] In an exemplary embodiment, the local topological feature extraction module includes: a second feature vector acquisition unit, used to determine the graph adjacency matrix of the gene regulatory network; based on the graph adjacency matrix, the connected low-dimensional gene nodes are made to have similar embeddings through an autoencoder to capture the local topological features of the low-dimensional gene nodes, and the second feature vector is obtained while maintaining the global network structure by making the low-dimensional gene nodes with similar neighbor structures have similar embeddings; the second feature vector includes information about the global network structure and the local topological features of each of the low-dimensional gene nodes.

[0011] In an exemplary embodiment, the gene regulatory network encoding module includes: a gene relationship collection unit, used to collect directional gene relationships; a gene regulatory relationship library calling unit, used to store the collected gene relationships in a gene regulatory relationship library, and obtain the gene regulatory network by calling the gene regulatory relationship library.

[0012] In an exemplary embodiment, the graph neural network construction device also includes: a graph neural network training module, which is used to train the graph neural network.

[0013] In an exemplary embodiment, the graph neural network training module includes: a reference data set acquisition unit, used to acquire a reference data set; the reference data set includes single-cell data carrying a label; the label indicates the cell type corresponding to the single-cell data; a prediction result acquisition unit, used to train the graph neural network based on the single-cell data in the reference data set to obtain a prediction result; the prediction result indicates the cell type corresponding to the single-cell data; a loss function calculation unit, used to calculate the loss function based on the prediction result and the label, and use the loss function to optimize the parameters of the graph neural network; a repetition unit, used to repeat the above steps based on other single-cell data in the reference data set until the loss function meets a specific condition, and the training of the graph neural network is completed.

[0014] In an exemplary embodiment, the graph neural network construction device also includes: a single-cell data acquisition module, used to acquire newly sequenced single-cell data; a probability prediction module, used to input the single-cell data into the graph neural network, so that the cell nodes of the last layer of the graph neural network predict the probability that the single-cell data belongs to different cell types through a linear classifier; and a labeling result acquisition module, used to obtain the labeling results of the single-cell data based on the predicted probability.

[0015] According to one aspect of the present application, an electronic device includes at least one processor and at least one memory, wherein the memory stores computer-readable instructions; the computer-readable instructions are executed by one or more of the processors, so that the electronic device implements the graph neural network construction method as described above.

[0016] According to one aspect of the present application, a storage medium stores computer-readable instructions thereon, and the computer-readable instructions are executed by one or more processors to implement the graph neural network construction method as described above.

[0017] According to one aspect of the present application, a computer program product includes computer-readable instructions, which are stored in a storage medium. One or more processors of an electronic device read the computer-readable instructions from the storage medium, load and execute the computer-readable instructions, so that the electronic device implements the graph neural network construction method as described above.

[0018] The beneficial effects of the technical solution provided by this application are:

[0019] In the above technical scheme, the first eigenvector and the second eigenvector are obtained according to the single-cell data and the gene regulatory network in the single-cell data set, and a graph neural network is constructed based on the first eigenvector and the second eigenvector. The obtained graph neural network can be used to automatically annotate the newly sequenced single-cell data. When constructing the graph neural network, the large-scale gene regulatory relationship (gene regulatory network) is integrated with the single-cell RNA sequencing data (single-cell data), which can not only reflect the information flow between genes and cells, but also reflect the influence between genes and genes caused by the gene regulatory relationship, that is, the gene regulatory relationship is introduced, and the influence between genes is considered, so that the constructed graph neural network can more accurately perform cell annotation, whether it is cell annotation within the same data set, or between different data sets, tissues, sequencing platforms and even species. The present invention can achieve higher single-cell data annotation accuracy, that is, overcome the batch effect; in addition, the present invention comes with a constructed gene regulatory network, which saves the time of manually collecting relevant information from various databases, and is more conducive to the user's operation, saving time and effort. The present invention effectively solves the problems of inaccurate cell annotation, time-consuming and labor-intensive, and limited by batch effects in the related art. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments of the present application. Obviously, the drawings described below are only some embodiments of the present application, and those skilled in the art can obtain other drawings based on these drawings without creative work.

[0021] Figure 1 It is a schematic diagram of the implementation environment involved in this application;

[0022] Figure 2 is a flowchart of a method for constructing a graph neural network according to an exemplary embodiment;

[0023] Figure 3 yes Figure 2 The flowchart of an embodiment after step 260 in the corresponding embodiment;

[0024] Figure 4 yes Figure 2 A flowchart of an embodiment corresponding to step 260 in an embodiment;

[0025] Figure 5 yes Figure 2 A flowchart of an embodiment corresponding to step 240 in an embodiment;

[0026] Figure 6 yes Figure 2A flowchart of an embodiment corresponding to step 220 of “obtaining a constructed gene regulatory network”;

[0027] Figure 7 yes Figure 2 The flowchart of an embodiment after step 260 in the corresponding embodiment;

[0028] Figure 8 It is a schematic diagram of a specific implementation of a graph neural network construction method in an application scenario;

[0029] Fig. 9 is a schematic diagram of the performance comparison between the present invention (scHGR) and other annotation tools within a data set;

[0030] Fig.10 It is a schematic diagram of the performance evaluation of the present invention (scHGR) and other annotation tools on cross-sequencing platform data sets;

[0031] Fig.11 is a schematic diagram of the performance evaluation of the present invention (scHGR) and other annotation tools on cross-species datasets;

[0032] Fig.12 is a structural block diagram of a graph neural network construction device according to an exemplary embodiment;

[0033] Fig.13 The present invention is a structural block diagram of an electronic device according to an exemplary embodiment. DETAILED DESCRIPTION

[0034] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be interpreted as limiting the present application.

[0035] It will be understood by those skilled in the art that, unless expressly stated, the singular forms "one", "said", and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present disclosure refers to the presence of the features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or there may be intermediate elements. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling. The term "and / or" used herein includes all or any unit and all combinations of one or more associated listed items.

[0036] The following is an introduction and explanation of several terms involved in this application:

[0037] Single-cell transcriptome sequencing, or single-cell RNA sequencing (scRNA-seq), is a high-throughput gene sequencing technology used to study the gene expression patterns of individual cells. It can help us understand the differences between different cell types, changes in gene expression during cell development and differentiation, cell functions and metabolic pathways, etc. Single-cell transcriptome sequencing can obtain RNA sequencing data of individual cells.

[0038] The main steps of single-cell RNA sequencing technology are: the first step is to extract a part of cells from the tissue; the second step is to separate this group of extracted cells into individual cells; the third step is to dissolve the cells to obtain the genetic material in each cell; the fourth step is to synthesize complementary DNA (cDNA) from the RNA in a single cell through the action of reverse transcriptase; the fifth step is to further culture the synthesized cDNA through in vitro transcription (IVT) or polymerase chain reaction (PCR) technology to obtain a large amount of cDNA (i.e., amplification and multiplication); the sixth step is to prepare a gene sequencing library and sequence the amplified cDNA to obtain gene expression values ​​(sequencing data).

[0039] Gene expression profile refers to the activity level of all genes in an organism during transcription and translation under specific conditions. By analyzing the gene expression profile, we can understand the activity of different genes in different tissues, different developmental stages or different environmental conditions, thereby revealing the function and regulatory mechanism of genes. Gene expression profiles can be determined through RNA sequencing data: by sequencing the RNA of a single cell, we can determine which genes in the cell are transcribed into RNA, thereby obtaining the gene expression profile of the cell.

[0040] DeepWalk is an unsupervised learning method based on random walks for learning node embeddings. DeepWalk generates node sequences by performing random walks in a graph and converts these node sequences into embedding vectors using the Word2Vec algorithm. DeepWalk can learn node embeddings without the need for prior knowledge and can be applied to large-scale graph data.

[0041] Node2Vec is a supervised learning method based on random walks for learning node embedding. Node2Vec treats node embedding as a classification problem and uses a deep learning model to learn node embedding. Node2Vec can capture the structural information between nodes and can be applied to large-scale graph data.

[0042] Hybrid Graph Neural Network (HGNN) is a hybrid model that combines Graph Neural Network (GNN) and other neural network models. GNN is a type of neural network model specifically designed to process graph structured data. It can learn feature representations for nodes and edges and capture information in graph structures through message passing and aggregation. However, GNN may face problems such as high computational complexity and limited information dissemination when processing large-scale graph data.

[0043] As mentioned above, the prior art has technical problems such as inaccurate cell labeling, time-consuming and labor-intensive, and limited by batch effects.

[0044] One cell annotation scheme is to rely solely on cell clustering technology of newly sequenced single-cell data, but in this scheme, different algorithms may lead to different cell subpopulations, thus causing deviations in subsequent studies. In addition, it is very time-consuming and labor-intensive to collect verified cell type-specific marker genes from databases and literature, or to further annotate clusters with the help of experts.

[0045] Another cell labeling solution relies on automatic labeling technology of cell atlases. With the rapid accumulation of single-cell omics data, large-scale cell atlases are easily obtained. Based on large-scale cell atlases, artificial intelligence models can be trained to establish mapping relationships from gene expression profiles to specific cell types. This can yield models that have the ability to determine cell types based on gene expression profiles. This allows for automatic labeling at the single-cell level in a reliable and repeatable manner. However, this technology does not take into account information propagation between genes and gene regulatory relationships.

[0046] From the above, it can be seen that the relevant technology still has defects such as inaccurate cell labeling, time-consuming and labor-intensive, and limited by batch effects.

[0047] To this end, the graph neural network construction method provided in the present application can construct a graph neural network that can accurately and automatically annotate single-cell data. Accordingly, the graph neural network construction method is suitable for a graph neural network construction device, which can be deployed in an electronic device. The electronic device can be a computer device configured with a von Neumann architecture, for example, the computer device includes a desktop computer, a laptop computer, a server, etc.; the electronic device can also refer to a portable and mobile electronic device, for example, the electronic device includes a smart phone, a tablet computer, etc.

[0048] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0049] Figure 1The schematic diagram is a diagram of an implementation environment involved in a graph neural network construction method. The implementation environment includes a terminal 110 and a server 130.

[0050] Specifically, the terminal 110 may be an electronic device such as a desktop computer, a laptop computer, a tablet computer, a smart phone, etc., which is not limited here.

[0051] The server side 130 can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers. It can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, as well as big data and artificial intelligence platforms.

[0052] The server 130 establishes a communication connection with the terminal 110 in advance through wired or wireless means.

[0053] In one application scenario, a single-cell data set is input into the terminal 110, and the terminal 110 uploads the single-cell data set to the server 130, so that the server 130 builds a graph neural network based on the single-cell data set and the gene regulatory network. After the server 130 builds and trains the graph neural network, the user can upload the newly sequenced single-cell data to be annotated to the server 130 through the terminal 110, obtain the annotation results through the graph neural network of the server 130, and transmit them to the terminal 110 for feedback to the user.

[0054] See also Figure 2 , the embodiment of the present application provides a method for constructing a graph neural network, the method is applicable to an electronic device, the electronic device may be Figure 1 The server 130 in the implementation environment shown may also be Figure 1 Terminal 110 is shown in an implementation context.

[0055] In the following method embodiments, for ease of description, the execution subject of each step of the method is taken as an electronic device as an example for illustration, but this does not constitute a specific limitation.

[0056] like Figure 2 As shown, the method may include the following steps:

[0057] Step 200, encoding the single-cell data in the single-cell data set into a low-dimensional latent space to obtain a first feature vector that retains key features and / or type-specific information of the single-cell data.

[0058] The single-cell data may be RNA sequencing data or gene expression profile, which is not specifically limited here.

[0059] Cell data has the characteristics of high dimension and sparseness, so the cell data can be preprocessed first, for example, by performing dimensionality reduction on the cell data.

[0060] In a possible implementation, the dimensionality reduction technique PCA (Principal Component Analysis) is used to encode the cell data into a low-dimensional latent space to obtain a first eigenvector.

[0061] Step 220, obtaining the constructed gene regulatory network, and encoding the gene regulatory network into low-dimensional gene nodes.

[0062] A gene regulatory network is a network with genes as nodes, which describes the interactions and regulatory relationships between genes. Each gene is regarded as a node in the network, and the interactions and regulatory relationships between genes are represented by the edges connecting the nodes.

[0063] In one possible implementation, the gene regulatory network is constructed based on public databases and platforms.

[0064] Gene regulatory networks are complex and intricate. In order to better understand and analyze gene regulatory networks and reduce their complexity, they can be encoded into low-dimensional gene nodes. By encoding gene regulatory networks into low-dimensional gene nodes, each gene can be represented as a vector with a lower dimension, thereby reducing the complexity of calculation and storage and making it easier to visualize and analyze.

[0065] In one possible implementation, graph embedding technology can be used in the encoding process. Graph embedding technology can map complex graph structures into low-dimensional vector spaces, preserving the relationships and similarities between nodes. By applying graph embedding technology, gene nodes in the gene regulatory network are encoded into low-dimensional vectors while preserving their positions and relationships in the global network.

[0066] Step 240 , extracting local topological features of low-dimensional gene nodes while retaining the global network structure of the gene regulatory network to obtain a second feature vector.

[0067] The second eigenvector is used to reflect the interactions between all genes in the single-cell dataset.

[0068] During the encoding process, we hope to preserve the global network structure, that is, to maintain the interaction relationship between genes. At the same time, we also hope to extract the local topological features of each gene to reflect the interaction between genes.

[0069] By extracting the local topological features of each gene, we can understand the neighbor nodes and interactions of each gene in the network. Local topological features can reflect the interactions between genes and help us understand the mechanism of gene regulation. By integrating the interactions between all genes, we can more comprehensively analyze and understand the interaction relationship between genes in the entire single-cell dataset (sequencing dataset).

[0070] In a possible implementation, the second eigenvector is obtained using a graph embedding technology, and the graph embedding technology may be SDNE (Structural Deep Network Embedding), DeepWalk, Node2Vec, etc., which is not specifically limited here.

[0071] Take SDNE as an example. SDNE is an unsupervised learning method based on deep learning to embed nodes in a graph into a low-dimensional space. SDNE uses an autoencoder to learn the embedding of nodes. An autoencoder is a neural network that can encode input data into a low-dimensional vector and then decode it back to the original data. SDNE uses an autoencoder to learn the embedding of nodes. It takes nodes as input and tries to reconstruct them so that the nodes are embedded in a low-dimensional space. SDNE can capture the structural information between nodes and can be applied to large-scale graph data.

[0072] Step 260: construct a graph neural network based on the first eigenvector and the second eigenvector.

[0073] In a possible implementation, each cell and gene is used as a node in the graph neural network, and the first eigenvector and the second eigenvector are used as features of the node.

[0074] The constructed graph neural network extracts the nonlinear correlation between single-cell data and cell types, so that newly sequenced single-cell data can be automatically annotated, such as Figure 3 As shown, after step 260, the method may further include the following steps:

[0075] Step 300, obtaining newly sequenced single-cell data.

[0076] Single-cell data can be single-cell RNA sequencing data or gene expression profiles. Single-cell data can be obtained through single-cell RNA sequencing (scRNA-seq).

[0077] Step 320, input the single-cell data into the graph neural network, so that the cell nodes of the last layer of the graph neural network pass through the linear classifier to predict the probability that the single-cell data belongs to different cell types.

[0078] In one possible implementation, single-cell data is encoded into vectors and input into a graph neural network, which processes and transmits the input vectors through a multi-layer neural network model.

[0079] Step 340, obtaining the labeling result of the single cell data based on the predicted probability.

[0080] In one possible implementation, the cell type with the highest probability is used as the annotation result of the single-cell data.

[0081] Through the above process, the first eigenvector and the second eigenvector are obtained according to the single-cell data and the gene regulatory network in the single-cell data set, and a graph neural network is constructed based on the first eigenvector and the second eigenvector. The obtained graph neural network can be used to automatically annotate the newly sequenced single-cell data. When constructing the graph neural network, the large-scale gene regulatory relationship (gene regulatory network) is integrated with the single-cell RNA sequencing data (single-cell data), which can not only reflect the information flow between genes and cells, but also reflect the influence between genes and genes caused by the gene regulatory relationship, that is, introduce the gene regulatory relationship, consider the influence between genes, so that the constructed graph neural network can more accurately perform cell annotation, whether it is to perform cell annotation within the same data set, or between different data sets, tissues, sequencing platforms and even species. The present invention can achieve higher single-cell data annotation accuracy, that is, overcome the batch effect; in addition, the present invention comes with a constructed gene regulatory network, which saves the time of manually collecting relevant information from various databases, and is also more conducive to the user's operation, saving time and effort. The present invention effectively solves the problems of inaccurate cell annotation, time-consuming and labor-intensive, and limited by batch effects in the related art.

[0082] See also Figure 4 In an exemplary embodiment, step 260 may include the following steps:

[0083] Step 400, using the second feature vector as the initial feature of the gene node in the graph neural network.

[0084] When constructing a graph neural network, each cell and gene is used as a node in the graph neural network, and the characteristics of the nodes of the graph neural network are determined.

[0085] Step 420: Use the first feature vector as the initial feature of the cell node in the graph neural network.

[0086] Step 440, constructing edges of the graph neural network between the gene nodes and the cell nodes.

[0087] The edges of a graph neural network are used to determine the connection relationship between nodes. The construction of the edges of a graph neural network can be based on the attribute similarity, spatial proximity, common attributes, etc. between nodes.

[0088] In a possible implementation, if the expression level of a gene in a cell is higher than 0, there is a bidirectional edge between the gene and the cell node, that is, a gene expression edge connecting the gene node and the cell node.

[0089] In one possible implementation, if one gene regulates another gene, there is a unidirectional edge between the two genes, namely, a gene regulation edge connecting between the gene nodes.

[0090] In one possible implementation, each node has an edge pointing to itself, that is, a self-feedback edge from a gene node or a cell node pointing to itself.

[0091] Step 460, obtaining a graph neural network from the edges between gene nodes and cell nodes.

[0092] In the above embodiment, the second eigenvector reflecting the interactions between all genes in the single-cell data set is used to construct a graph neural network. Compared with the prior art that ignores the influence between genes, the present application introduces gene regulatory relationships into single-cell data annotation, which can improve the accuracy of cell annotation.

[0093] See also Figure 5 In an exemplary embodiment, step 240 may include the following steps:

[0094] Step 500, determining the graph adjacency matrix of the gene regulatory network.

[0095] A graph adjacency matrix is ​​a data structure that represents a graph. It is a two-dimensional matrix in which rows and columns represent vertices in the graph, and the elements in the matrix represent the connection relationship between vertices.

[0096] The graph adjacency matrix of the gene regulatory network is used to represent the connection relationship between the nodes of the gene regulatory network.

[0097] Step 520, based on the graph adjacency matrix, the connected low-dimensional gene nodes are made to have similar embeddings through the autoencoder to capture the local topological features of the low-dimensional gene nodes, and the second feature vector is obtained while maintaining the global network structure by making the low-dimensional gene nodes with similar neighbor structures have similar embeddings.

[0098] The second eigenvector includes information about the global network structure and the local topological features of each low-dimensional gene node.

[0099] As mentioned above, the graph adjacency matrix of the gene regulatory network is used to represent the connection relationship between the nodes of the gene regulatory network. Therefore, based on the graph adjacency matrix of the gene regulatory network, the connected low-dimensional gene nodes and the low-dimensional gene nodes with similar neighbor structures can be determined.

[0100] Under the action of the above-mentioned embodiments, the gene regulatory network is reduced to a continuous vector to serve as the initial feature of the graph neural network, thereby integrating the complex relationships between genes into the gene nodes, facilitating the information propagation between cell nodes and gene nodes in the graph neural network, and making the single-cell data annotation of the graph neural network more accurate.

[0101] See also Figure 6 In an exemplary embodiment, step 220 of "obtaining a constructed gene regulatory network" may include the following steps:

[0102] Step 600, collecting directional gene relationships.

[0103] Gene relationship refers to the interaction and influence between genes. Different relationships can exist between genes, including interaction, complementary effect, antagonism, etc., which are not intended to constitute specific limitations here.

[0104] In a possible implementation, the source of gene relationships includes public gene regulation information databases and platforms, such as GREDB, BioGRID, TRRUST, and RegNetwork, etc., which are not specifically limited here.

[0105] Among them, TRRUST is a manually annotated database that effectively extracts regulatory relationships from more than 20 million published articles, involving human and mouse data.

[0106] RegNetwork was developed based on 25 databases, integrating regulatory relationships within the database and potential relationships inferred based on transcription factor binding sites.

[0107] BioGRID is a biomedical interactive repository that includes both focused low-throughput studies and large high-throughput datasets covering 80,848 papers and multiple species.

[0108] GREDB is a publicly available, manually curated biological database that preserves human gene expression regulatory relationships.

[0109] Step 620, storing the collected gene relationships in a gene regulation relationship library, and obtaining a gene regulation network by calling the gene regulation relationship library.

[0110] The gene regulatory relationship library covers a variety of gene regulatory relationships and is the basic data for building and training graph neural networks.

[0111] In the above embodiments, the present invention constructs a large-scale gene regulation relationship library, which saves the time of manually collecting relevant information from various databases, saves time and labor, and is easy to use.

[0112] See also Figure 7 In an exemplary embodiment, after step 260, the method may further include training the graph neural network, and the training process includes the following steps:

[0113] Step 700: Obtain a reference data set.

[0114] The reference data set includes single-cell data carrying labels, where the labels indicate the cell type corresponding to the single-cell data, and the single-cell data refers to expression data (such as RNA sequencing data or gene expression profiles).

[0115] With the rapid accumulation of single-cell omics data, large amounts of detailed labeling datasets (reference datasets) are readily available.

[0116] Step 720: Train the graph neural network based on the single-cell data in the reference data set to obtain a prediction result.

[0117] The prediction results indicate the cell type to which the single-cell data corresponds.

[0118] Step 740, calculate the loss function based on the prediction results and labels, and use the loss function to optimize the parameters of the graph neural network.

[0119] In one possible implementation, the parameters are optimized by minimizing the gap between the predictions and the labels (the loss function).

[0120] Step 760, repeat the above steps based on other single-cell data in the reference data set until the loss function meets specific conditions, and the training of the graph neural network is completed.

[0121] The specific condition may be that the loss function reaches a specific range, or that the loss function converges, which is not specifically limited here.

[0122] Through the above-mentioned embodiments, the graph neural network is trained to establish the relationship between the gene expression level and the cell type in each cell.

[0123] Figure 8 This is a schematic diagram of a specific implementation method of a graph neural network construction method in an application scenario.

[0124] In this application scenario, A represents input. Figure 8 In the algorithm, the input includes a reference dataset and a query dataset. The reference dataset contains annotated gene expression profiles (single-cell data), and the query dataset includes newly sequenced gene expression profiles (unlabeled).

[0125] For each reference data set, cell populations with less than 10 cells were first removed to ensure that the hybrid graph neural network could be fully trained. In terms of gene screening, different strategies were adopted depending on the data distribution. All genes were first sorted according to dispersion, which is the quotient of the mean and variance. If the reference data and query data distributions are similar, the top 1000 genes are selected and the raw reads are used as input for the technique. Otherwise, the top 5000 genes are retained and the gene reads need to be log-normalized.

[0126] B is a gene regulation relationship library, involving gene relationship records of humans and mice, covering multiple gene regulation information databases, including GREDB, BioGRID, TRRUST and RegNetwork. It promotes cell recognition by integrating signal propagation between genes. In this application scenario, only gene regulation data with experimental evidence are collected.

[0127] C is a gene encoder that generates the initial embedding of gene nodes using a graph embedding algorithm using the relevant gene information obtained from the gene regulatory relationship library B. Gene encoder C uses the classic graph embedding technology SDNE to encode the intricate gene regulatory network into low-dimensional gene nodes, while retaining the global network structure and extracting the characteristics of the local topology of each gene to reflect the interactions between all genes in the sequencing data set.

[0128] D is a cell encoder that encodes single-cell data (gene expression profiles) into low-dimensional embeddings of cell nodes. Cell encoder D uses nonlinear dimensionality reduction technology PCA to encode high-dimensional and sparse original gene expression profiles into low-dimensional latent space, retaining the key features and type-specific information in the original data.

[0129] E is a hybrid graph neural network. In this application scenario, each cell and gene is used as a node in the hybrid graph neural network, and the initial features are low-dimensional vectors constructed by the cell encoder D and the gene encoder C. The hybrid graph neural network E includes six types of edges, including "intracellular gene expression", "gene regulation" and "self-feedback". Figure 8 The figure shows five types of node relationships, a, b, c, d, e, etc. The hybrid graph neural network E extracts the nonlinear correlation between expression patterns (gene expression profiles) and cell types, and updates the node embeddings so that the cell nodes of the same type are similar and the cell nodes of different types are significantly different. The hybrid graph neural network E optimizes the parameters and completes the training by minimizing the gap between the predicted results and the reference labels.

[0130] F is the output of the hybrid graph neural network E, including automatic cell annotation, differentially expressed gene analysis (DEG analysis) based on propagation scores, revealing new cell subtypes with biological significance, and gene ontology enrichment (GO enrichment) for major diseases.

[0131] In this application scenario, a single-cell data automatic annotation tool based on a deep learning framework is provided. The tool uses existing large-scale single-cell data (single-cell atlas) as a reference, and explores the relationship between gene expression levels and cell categories in each cell through graph embedding technology, dimensionality reduction technology, and hybrid graph neural network technology. The cell type of the reference data set is migrated to the newly sequenced query data set to achieve the function of automatic annotation. Moreover, even if the reference data set and the query data set come from different batches, different sequencing technologies, different tissues, or even species, through the hybrid graph neural network, the tool can still achieve the cell type transfer between them, that is, overcome the batch effect between large-span data sets, and the tool can achieve accurate annotation of single-cell data across tissues, sequencing platforms, and even species. While annotating cell types, the present invention can screen key genes and perform downstream work such as differential gene analysis.

[0132] The present invention is verified by the following experiments.

[0133] like Fig. 9 As shown, it is a schematic diagram showing the performance comparison of scHGR (the present invention) and other annotation techniques within the dataset.

[0134] The annotation effects of the present invention (scHGR) and other annotation technologies (scHPL, scGAC, scDeepSort, scVI) were compared, and human peripheral blood mononuclear cell datasets (PBMC-FACS) and mouse brain datasets (AMB) were used for experimental verification. In order to avoid experimental evaluation bias, a five-fold cross-validation strategy was used for each dataset. Fig. 9 The values ​​shown are all average values ​​of evaluation indicators. Fig. 9 As shown in the figure, multiple indicators (Accuracy, Precision, Recall, F1-Score, MCC) are used to comprehensively quantify the effectiveness of different annotation techniques. For PBMC-FACS, all five indicators came to the same conclusion: scHGR (the present invention) outperforms other state-of-the-art technologies. In particular, the accuracy is 6% higher than the second place, which illustrates the strength of scHGR in avoiding false positive samples. scVI is second only to scHGR because its generation module alleviates the inherent limitation of high sparsity of raw sequencing data.

[0135] In AMB, the prediction of scHGR (the present invention) is almost perfect, with an average precision and Matthews correlation coefficient (MCC) of up to 99%. It is obvious that the precision, recall and F1-Score of scHGR (the present invention) are 20% higher than those of other methods, indicating that it has advantages in overcoming false alarm rates and annotating large-scale data sets. On the other hand, the error band at the top of the bar graph shows that scHRG (the present invention) has the lowest error and is a reliable tool with strong stability and balanced performance.

[0136] like Fig.10 As shown, it shows a schematic diagram of the performance evaluation of the present invention (scHGR) and other annotation tools on a cross-sequencing platform data set.

[0137] Among them, the F1-Score and Accuracy heatmaps obtained through cross-technology experiments, where the rows represent annotation tools and the columns represent data sets from different sequencing platforms. The box plot summarizes the maximum, minimum, median and quantile accuracy of different annotation tools, as well as the F1-Score. The median is marked with a specific value above the corresponding box.

[0138] Among them, the effectiveness of the annotation of the present invention on the data sets of 6 different sequencing platforms was measured. The experimental results on the cross-sequencing platform data sets showed significant differences: almost perfect annotation was achieved on PBMC1-SM2 (scHGR: F1-Score = 94%), but the annotation results of PBMC1-SW were generally poor (except scHGR, the F1-Score of all methods was 60%). Except for scHPL, the accuracy of each method was not significantly different, all exceeding 70%. In contrast, the F1-Score of scHGR was significantly higher than that of other technologies (scHPL: 26%, scGAC: 58%, scDeepSort: 63%, scVI: 67%, scHGR: 80%), proving that the present application (scHGR) has cross-technical identifiability (that is, it overcomes the batch effect between large-span data sets and can achieve accurate annotation of single-cell data across sequencing platforms). The gene regulation considered by scHGR is not constrained by the sequencing platform, which is also the key reason for its high robustness.

[0139] In order to prove that the present application scheme can achieve accurate annotation of single-cell data across species, the inventors used the human dataset as a reference to annotate the mouse dataset, and used the mouse dataset as a reference to annotate the human dataset.

[0140] like Fig.11As shown, it shows a schematic diagram of the performance evaluation of different annotation tools on cross-species datasets. The bar graph shows the accuracy of all annotation tools in two cases, and the specific numerical values ​​are marked above the corresponding bar graphs.

[0141] Among them, the accuracy of the present invention and the existing technology in automatically annotating cross-species single-cell sequencing data is compared. Overall, scDeepSort and scVI showed unstable performance, with accuracy dropping by more than 20%, indicating that their effectiveness depends largely on the quality and scale of the reference dataset. In contrast, the reason why scHGR remains stable and effective is that the hybrid graph neural network forms a cross-species information flow through homologous genes. For human-mouse, scGAC, scDeepSort, scVI, and scHGR perform significantly better than scHPL, with scHGR slightly outperforming other technologies. However, in the mouse-human scenario, the prediction results are often unsatisfactory due to the insufficient number of reference cells. In this case of insufficient training, only scHGR maintains a high degree of reliability.

[0142] Combination Fig. 9 , Fig.10 and Fig.11 It can be seen that the method of the present application (scHGR) has achieved better results than other annotation tools, whether within a single dataset or across sequencing platform datasets or species datasets, and has effectively solved the problems of inaccurate cell annotation, time-consuming and labor-intensive, and limited by batch effects in related technologies.

[0143] The following is an embodiment of the device of the present application, which can be used to execute the method for constructing a graph neural network involved in the present application. For details not disclosed in the embodiment of the device of the present application, please refer to the method embodiment of the method for constructing a graph neural network involved in the present application.

[0144] See also Fig.12 In an embodiment of the present application, a graph neural network construction device 900 is provided, including but not limited to: a cell data encoding module 910, a gene regulatory network encoding module 930, a local topological feature extraction module 950 and a graph neural network construction module 970.

[0145] Among them, the cell data encoding module 910 is used to encode the single-cell data in the single-cell data set into a low-dimensional latent space to obtain a first feature vector that retains the key features and / or type-specific information of the single-cell data.

[0146] The gene regulatory network encoding module 930 is used to obtain the constructed gene regulatory network and encode the gene regulatory network into low-dimensional gene nodes.

[0147] The local topological feature extraction module 950 is used to extract the local topological features of the low-dimensional gene nodes while retaining the global network structure of the gene regulatory network to obtain a second feature vector; the second feature vector is used to reflect the interaction between all genes in the single-cell dataset.

[0148] The graph neural network construction module 970 is used to construct a graph neural network based on the first eigenvector and the second eigenvector; the graph neural network is used to automatically annotate single-cell data.

[0149] It should be noted that the graph neural network construction device provided in the above embodiment only uses the division of the above-mentioned functional modules as an example when constructing the graph neural network. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the graph neural network construction device will be divided into different functional modules to complete all or part of the functions described above.

[0150] In addition, the graph neural network construction device and the graph neural network construction method provided in the above embodiments belong to the same concept, and the specific manner in which each module performs operations has been described in detail in the method embodiments and will not be repeated here.

[0151] See also Fig.13 In an embodiment of the present application, an electronic device 4000 is provided, and the electronic device 4000 may include: a desktop computer, a laptop computer, a server, etc.

[0152] exist Fig.13 In the embodiment, the electronic device 4000 includes at least one processor 4001 and at least one memory 4003.

[0153] The data interaction between the processor 4001 and the memory 4003 can be realized through at least one communication bus 4002. The communication bus 4002 may include a path for transmitting data between the processor 4001 and the memory 4003. The communication bus 4002 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The communication bus 4002 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Fig.13 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0154] Optionally, the electronic device 4000 may further include a transceiver 4004, which may be used for data interaction between the electronic device and other electronic devices, such as data transmission and / or data reception. It should be noted that in actual applications, the transceiver 4004 is not limited to one, and the structure of the electronic device 4000 does not constitute a limitation on the embodiments of the present application.

[0155] Processor 4001 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It may implement or execute various exemplary logic blocks, modules and circuits described in conjunction with the disclosure of this application. Processor 4001 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0156] The memory 4003 can be a ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical disk storage, optical disk storage (including compressed optical disk, laser disk, optical disk, digital versatile disk, Blu-ray disk, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program instructions or codes in the form of instructions or data structures and can be accessed by the electronic device 400, but is not limited to this.

[0157] Computer-readable instructions are stored in the memory 4003 , and the processor 4001 can read the computer-readable instructions stored in the memory 4003 through the communication bus 4002 .

[0158] The computer-readable instructions are executed by one or more processors 4001 to implement the graph neural network construction method in the above-mentioned embodiments.

[0159] In addition, an embodiment of the present application provides a storage medium on which computer-readable instructions are stored, and the computer-readable instructions are executed by one or more processors to implement the graph neural network construction method as described above.

[0160] A computer program product is provided in an embodiment of the present application. The computer program product includes computer-readable instructions. The computer-readable instructions are stored in a storage medium. One or more processors of an electronic device read the computer-readable instructions from the storage medium, load and execute the computer-readable instructions, so that the electronic device implements the graph neural network construction method as described above.

[0161] Compared with the related art, this application can produce the following beneficial effects:

[0162] 1. Obtain the first eigenvector and the second eigenvector according to the single-cell data and the gene regulatory network in the single-cell data set, and construct a graph neural network based on the first eigenvector and the second eigenvector. The obtained graph neural network can be used to automatically annotate the newly sequenced single-cell data. When constructing the graph neural network, the large-scale gene regulatory relationship (gene regulatory network) is integrated with the single-cell RNA sequencing data (single-cell data), which can not only reflect the information flow between genes and cells, but also reflect the influence between genes and genes caused by the gene regulatory relationship, that is, introduce the gene regulatory relationship, consider the influence between genes, so that the constructed graph neural network can more accurately perform cell annotation, whether it is cell annotation within the same data set, or between different data sets, tissues, sequencing platforms and even species. The present invention can achieve higher single-cell data annotation accuracy, that is, overcome the batch effect; in addition, the present invention comes with a constructed gene regulatory network, which saves the time of manually collecting relevant information from various databases, and is more conducive to the user's operation, saving time and effort. The present invention effectively solves the problems of inaccurate cell annotation, time-consuming and labor-intensive, and limited by batch effects in the related art.

[0163] 2. It provides an automatic annotation tool for single-cell data based on a deep learning framework. The tool uses existing large-scale single-cell data (single-cell atlas) as a reference, and explores the relationship between gene expression levels and cell categories in each cell through graph embedding technology, dimensionality reduction technology, and hybrid graph neural network technology. It migrates the cell types of the reference data set to the newly sequenced query data set to achieve the function of automatic annotation. In addition, by overcoming the batch effect between large-span data sets through hybrid graph neural networks, the tool can accurately annotate single-cell data across tissues, sequencing platforms, and even species.

[0164] 3. Gene expression patterns are highly correlated with cell differentiation and biological functions. The present application introduces the key factor affecting gene expression levels - gene regulation relationships - into the task of cell identity annotation. Specifically, directed edges of gene regulation are added to the graph neural network to achieve information exchange between cells and genes, gene to gene, cell to cell, and gene itself and cell itself, thereby promoting the accuracy of single-cell data annotation. Whether cell annotation is performed within a data set or between different data sets, tissues, sequencing platforms, or even species, the present invention can achieve higher accuracy.

[0165] It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0166] The above description is only a partial implementation method of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for constructing a graph neural network, characterized in that: The method comprises: Encoding the single-cell data in the single-cell data set into a low-dimensional latent space to obtain a first feature vector that retains key features and / or type-specific information of the single-cell data; Acquire the constructed gene regulatory network, and encode the gene regulatory network into low-dimensional gene nodes; While retaining the global network structure of the gene regulatory network, extracting local topological features of the low-dimensional gene nodes to obtain a second feature vector; the second feature vector is used to reflect the interaction between all genes in the single-cell dataset; A graph neural network is constructed based on the first eigenvector and the second eigenvector; the graph neural network is used to automatically annotate single-cell data.

2. The method according to claim 1, characterized in that The constructing a graph neural network based on the first feature vector and the second feature vector includes: Using the second feature vector as the initial feature of the gene node in the graph neural network; Using the first feature vector as the initial feature of the cell node in the graph neural network; Constructing edges of the graph neural network between the gene nodes and the cell nodes; The graph neural network is obtained from the edges between gene nodes and cell nodes.

3. The method according to claim 2, characterized in that The edges of the graph neural network include at least one of the following: Gene regulatory edges connecting gene nodes; The self-feedback edge of a gene node or cell node pointing to itself; Gene expression edges connecting gene nodes and cell nodes.

4. The method according to claim 1, characterized in that The step of extracting local topological features of the low-dimensional gene nodes to obtain a second feature vector while retaining the global network structure of the gene regulatory network comprises: determining a graph adjacency matrix of the gene regulatory network; Based on the graph adjacency matrix, the connected low-dimensional gene nodes are made to have similar embeddings through an autoencoder to capture the local topological features of the low-dimensional gene nodes, and the second feature vector is obtained while maintaining the global network structure by making the low-dimensional gene nodes with similar neighbor structures have similar embeddings; the second feature vector includes information about the global network structure and the local topological features of each of the low-dimensional gene nodes.

5. The method according to claim 1, characterized in that The step of obtaining the constructed gene regulatory network comprises: Collect directional genetic relationships; The collected gene relationships are stored in a gene regulation relationship library, and the gene regulation network is obtained by calling the gene regulation relationship library.

6. The method according to claim 1, characterized in that After constructing the graph neural network based on the first feature vector and the second feature vector, the method further includes: Training the hybrid graph neural network; The training process includes: Acquire a reference data set; the reference data set includes single-cell data carrying a label; the label indicates the cell type corresponding to the single-cell data; Training the graph neural network based on the single cell data in the reference data set to obtain a prediction result; the prediction result indicates the cell type corresponding to the single cell data; Calculating a loss function according to the prediction result and the label, and optimizing the parameters of the graph neural network using the loss function; The above steps are repeated based on other single-cell data in the reference data set until the loss function meets specific conditions, and the training of the graph neural network is completed.

7. The method according to any one of claims 1 to 6, characterized in that: After constructing the graph neural network based on the first feature vector and the second feature vector, the method further includes: Obtain newly sequenced single-cell data; Inputting the single cell data into the graph neural network so that the cell nodes of the last layer of the graph neural network predict the probability that the single cell data belongs to different cell types through a linear classifier; Based on the predicted probability, the annotation result of the single cell data is obtained.

8. A graph neural network construction device, characterized in that: include: A single-cell data encoding module, used to encode the single-cell data in the single-cell data set into a low-dimensional latent space to obtain a first feature vector that retains key features and / or type-specific information of the single-cell data; A gene regulatory network encoding module, used to obtain the constructed gene regulatory network and encode the gene regulatory network into low-dimensional gene nodes; A local topological feature extraction module, used to extract local topological features of the low-dimensional gene nodes while retaining the global network structure of the gene regulatory network to obtain a second feature vector; the second feature vector is used to reflect the interaction between all genes in the single-cell dataset; A graph neural network construction module is used to construct a graph neural network based on the first eigenvector and the second eigenvector; the graph neural network is used to automatically annotate single-cell data.

9. An electronic device, characterized in that: include: at least one processor and at least one memory, wherein: The memory has computer-readable instructions stored thereon; The computer-readable instructions are executed by one or more of the processors, so that the electronic device implements the graph neural network construction method as described in any one of claims 1 to 7.

10. A storage medium having computer-readable instructions stored thereon, characterized in that: The computer-readable instructions are executed by one or more processors to implement the graph neural network construction method as described in any one of claims 1 to 7.

Images