Annotation model training method, cell type annotation method and related equipment

Abstract

The embodiment of the invention provides an annotation model training method, a cell type annotation method and related equipment, and belongs to the technical field of biological information analysis. The method comprises the following steps: inputting acquired spatial transcriptome data and single cell sequencing data into an initial annotation model; determining reference annotation information corresponding to each target single cell in each spatial site of the sample tissue data according to the single cell sequencing data by taking the spatial transcriptome data as a constraint condition; determining prediction annotation information corresponding to each target single cell based on the reference annotation information and the spatial transcriptome data; performing joint annotation processing on each target single cell according to the reference annotation information and the predicted annotation information to obtain a sample annotation result corresponding to the sample tissue data; and adjusting model parameters of the initial annotation model according to a sample annotation result to obtain a trained annotation model. The cell type annotation accuracy of the biological sample tissue can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of bioinformatics analysis, and in particular to a training method for an annotation model, a cell type annotation method, and related equipment. Background Art

[0002] Cell Type Annotation is the process of identifying and classifying different cell types in biological samples. Cell type annotation helps to study cells and promote the development of biotechnology.

[0003] In the related art, the spatial transcriptome information is used to obtain the cell position of each spatial site of the biological sample tissue, and the corresponding cell type annotation of the biological sample tissue is determined based on the spatial position and single-cell sequencing data. However, due to the limitation of equipment resources, the resolution of the spatial transcriptome information obtained is usually not high; when the cells in the biological sample tissue aggregate to form a high-density cell aggregation area, there will be multiple cells at the same spatial site. In this case, it is difficult to accurately map the single-cell sequencing data with the spatial transcriptome information, resulting in low accuracy of the cell type annotation of the biological sample tissue in the related art. Summary of the invention

[0004] The main purpose of the embodiments of the present application is to propose an annotation model training method, a cell type annotation method and related equipment, aiming to improve the accuracy of cell type annotation of biological sample tissues.

[0005] To achieve the above-mentioned purpose, a first aspect of an embodiment of the present application proposes a training method for an annotation model, the method comprising:

[0006] Obtaining sample tissue data, as well as spatial transcriptome data and single-cell sequencing data corresponding to the sample tissue data, and inputting the spatial transcriptome data and single-cell sequencing data into an initial annotation model;

[0007] Using spatial transcriptome data as a constraint, the corresponding reference annotation information of each target single cell in each spatial site of the sample tissue data is determined based on the single-cell sequencing data;

[0008] Based on the reference annotation information and spatial transcriptome data, the corresponding multimodal prior information of each target single cell is determined, and the corresponding predicted annotation information of the target single cell is determined according to the multimodal prior information;

[0009] According to the reference annotation information and the predicted annotation information, each target single cell is jointly annotated to obtain the sample annotation result corresponding to the sample tissue data;

[0010] The model parameters of the initial annotation model are adjusted according to the sample annotation results to obtain the trained annotation model.

[0011] In some embodiments, the spatial transcriptome data is used as a constraint condition, and the reference annotation information corresponding to each target single cell in each spatial site of the sample tissue data is determined according to the single-cell sequencing data, including:

[0012] Use the preset segmentation tool to deconstruct the spatial transcriptome data to obtain the cell number value representing the number of single cells at each spatial site of the sample tissue data, as well as the cell coordinate position representing the spatial position of each target single cell;

[0013] Using the single-cell sequencing data, determining the first gene expression information of the corresponding target single cell at each cell coordinate position, and determining a first logarithmic operation result according to the cell quantity value and the first gene expression information;

[0014] The first logarithmic operation result is added to the preset batch variable value and the preset noise variable value to obtain the reference annotation information of the target single cell.

[0015] In some embodiments, based on the reference annotation information and the spatial transcriptome data, determining the multimodal prior information corresponding to each target single cell includes:

[0016] Determine the cell spatial distribution characteristics of the target single cell at the corresponding spatial site according to the reference annotation information corresponding to the target single cell and the reference annotation information corresponding to the neighboring single cells of the target single cell;

[0017] Using the preset recognition tool, the morphological recognition processing is performed on the corresponding target single cell at the cell coordinate position to obtain the cell morphological characteristics of the target single cell;

[0018] Based on the spatial distribution characteristics and cell morphological characteristics of cells, the corresponding multimodal prior information of each single cell is determined.

[0019] In some embodiments, determining the prediction annotation information corresponding to the target single cell according to the multimodal prior information includes:

[0020] Predict the cell type of the target single cell based on the cell spatial distribution characteristics to obtain the first prior information;

[0021] The cell type of the target single cell is predicted based on the cell morphology characteristics to obtain the second prior information;

[0022] Calculate the product of the first prior information and the second prior information to obtain multimodal product data;

[0023] Using the multimodal product data, determining the second gene expression information corresponding to the target single cell at each cell coordinate position;

[0024] Based on the second gene expression information, the corresponding prediction annotation information of the target single cell is determined.

[0025] In some embodiments, determining the cell spatial distribution characteristics of the target single cell at the corresponding spatial site according to the reference annotation information corresponding to the target single cell and the reference annotation information corresponding to the neighboring single cells of the target single cell includes:

[0026] If the reference annotation information corresponding to the target single cell is different from the reference annotation information corresponding to the neighboring single cell, a first constraint value is obtained;

[0027] If the reference annotation information corresponding to the target single cell is the same as the reference annotation information corresponding to the neighboring single cell, the second constraint value is obtained;

[0028] Based on the first constraint value or the second constraint value, a cell spatial distribution characteristic characterizing the target single cell at the corresponding spatial location is obtained.

[0029] In some embodiments, joint annotation processing is performed on each target single cell according to the reference annotation information and the predicted annotation information to obtain a sample annotation result corresponding to the sample tissue data, including:

[0030] Based on a preset first weight parameter, the reference annotation information is updated, and based on a preset second weight parameter, the predicted annotation information is updated;

[0031] According to the updated reference annotation information and the updated predicted annotation information, each target single cell is jointly annotated to obtain the sample cell annotation results corresponding to the sample tissue data.

[0032] In some embodiments, the second weight parameter is obtained by the following steps, comprising:

[0033] Obtaining sample verification results corresponding to sample tissue data, as well as initial second weight parameters and initial adjustment ratios;

[0034] If the sample annotation result is different from the sample verification result, the initial adjustment ratio is updated to obtain the adjustment ratio;

[0035] The initial second weight parameter is updated based on the adjustment ratio to obtain the second weight parameter.

[0036] To achieve the above-mentioned purpose, the second aspect of the embodiment of the present application proposes a cell type annotation method, the method comprising:

[0037] Obtaining sample tissue data, as well as spatial transcriptome data and single-cell sequencing data corresponding to the sample tissue data, and inputting the spatial transcriptome data and single-cell sequencing data into an initial annotation model;

[0038] Using spatial transcriptome data as a constraint, the corresponding reference annotation information of each target single cell in each spatial site of the sample tissue data is determined based on the single-cell sequencing data;

[0039] Based on the reference annotation information and spatial transcriptome data, the corresponding multimodal prior information of each target single cell is determined, and the corresponding predicted annotation information of the target single cell is determined according to the multimodal prior information;

[0040] According to the reference annotation information and the predicted annotation information, each target single cell is jointly annotated to obtain the sample annotation result corresponding to the sample tissue data;

[0041] Adjust the model parameters of the initial annotation model according to the sample annotation results to obtain a trained annotation model;

[0042] Obtaining target tissue data, as well as target spatial transcriptome data and target single-cell sequencing data corresponding to the target tissue data, and inputting the target spatial transcriptome data and the target single-cell sequencing data into the trained annotation model;

[0043] According to the target spatial transcriptome data and target single-cell sequencing data, the target tissue data is jointly annotated to obtain the target cell annotation results corresponding to the target tissue data.

[0044] To achieve the above-mentioned purpose, a third aspect of an embodiment of the present application proposes a cell type annotation device, the device comprising:

[0045] An acquisition module is used to acquire sample tissue data, as well as spatial transcriptome data and single-cell sequencing data corresponding to the sample tissue data, and input the spatial transcriptome data and single-cell sequencing data into an initial annotation model;

[0046] A reference annotation module is used to determine the corresponding reference annotation information of each target single cell in each spatial site of the sample tissue data based on the single-cell sequencing data, taking the spatial transcriptome data as a constraint condition;

[0047] A prediction annotation module, used to determine the multimodal prior information corresponding to each target single cell based on the reference annotation information and the spatial transcriptome data, and to determine the prediction annotation information corresponding to the target single cell based on the multimodal prior information;

[0048] An annotation processing module is used to perform joint annotation processing on each target single cell according to the reference annotation information and the predicted annotation information to obtain a sample annotation result corresponding to the sample tissue data;

[0049] A training module is used to adjust the model parameters of the initial annotation model according to the sample annotation results to obtain a trained annotation model;

[0050] A first target processing module is used to obtain target tissue data, and target spatial transcriptome data and target single-cell sequencing data corresponding to the target tissue data, and input the target spatial transcriptome data and target single-cell sequencing data into the trained annotation model;

[0051] The second target processing module is used to perform joint annotation processing on the target tissue data according to the target spatial transcriptome data and the target single-cell sequencing data to obtain the target cell annotation results corresponding to the target tissue data.

[0052] To achieve the above-mentioned purpose, the fourth aspect of an embodiment of the present application proposes an electronic device, the electronic device comprising a memory and a processor, the memory storing a computer program, and the processor implementing the method of the first aspect or the method of the second aspect when executing the computer program.

[0053] To achieve the above objectives, the fifth aspect of an embodiment of the present application proposes a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the method of the first aspect or the method of the second aspect.

[0054] The cell type annotation method and device, electronic device and storage medium proposed in the present application obtain sample tissue data, as well as spatial transcriptome data and single-cell sequencing data corresponding to the sample tissue data, and input the spatial transcriptome data and single-cell sequencing data into the initial annotation model; using the spatial transcriptome data as a constraint condition, determine the reference annotation information corresponding to each target single cell in each spatial site of the sample tissue data according to the single-cell sequencing data; based on the reference annotation information and the spatial transcriptome data, determine the multimodal prior information corresponding to each target single cell, and determine the predicted annotation information corresponding to the target single cell according to the multimodal prior information; the multimodal prior information can further reveal the cell morphological characteristics of each target single cell in the spatial site, as well as the target single cell The cellular spatial characteristics of cells and adjacent single cells are connected so that the annotation device can have a deeper understanding of each target single cell; according to the reference annotation information and the predicted annotation information, each target single cell is jointly annotated to obtain the sample annotation result corresponding to the sample tissue data; through the mutual complementation of the reference annotation information and the predicted annotation information, the reference annotation information can be better mapped with the spatial transcriptome data, thereby improving the accuracy and comprehensiveness of the annotation model for the target single cell; according to the sample annotation results, the model parameters of the initial annotation model are adjusted to obtain the trained annotation model, which can perform cell annotation processing on the input target tissue data at the single-cell resolution accuracy, thereby improving the accuracy of the cell type annotation of the biological sample tissue. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1It is a schematic diagram of an application scenario of a training device for an annotation model provided in an embodiment of the present application;

[0056] Figure 2 It is an optional flowchart of the training method of the annotation model provided in the embodiment of the present application;

[0057] Figure 3 is an optional flow chart of the cell type annotation method provided in the embodiment of the present application;

[0058] Figure 4 This is a schematic diagram of an optional functional module of the cell type annotation device provided in an embodiment of the present application;

[0059] Figure 5 It is a schematic diagram of the hardware structure of the electronic device provided in the embodiment of the present application. DETAILED DESCRIPTION

[0060] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0061] It should be noted that, although the functional modules are divided in the device schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first", "second", etc. in the specification, claims and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0062] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0063] First, some nouns involved in this application are analyzed:

[0064] Spatial transcriptomics (ST) sequencing technology aims to measure gene expression within cells while retaining the precise spatial positioning of cells. It is of great significance to the study of growth and development and the occurrence and development of diseases.

[0065] The indicator function is usually defined as a function f(x), where x is a point in the set or region. When x belongs to the set or region, the value of f(x) is 1; otherwise, the value of f(x) is 0.

[0066] The normalization constant is a concept from mathematical probability theory. It refers to a constant contained in any interval of any non-negative function, which makes the integral of the function over a specific interval exactly equal to 1. This concept is usually introduced to transform the function into a probability density function or probability mass function.

[0067] The Potts model is an important theoretical model in statistical physics and condensed matter physics. It is mainly used to describe the interaction between particles or spins in lattice systems and their macroscopic properties.

[0068] The Maximum A Posteriori (MAP) method describes the relationship between prior probability, likelihood function and posterior probability, and estimates model parameters by maximizing the posterior probability, where the posterior probability refers to the probability distribution of model parameters after given observation data. The goal of the MAP method is to find an optimal model parameter value so that the posterior probability of the parameter value is maximized under given observation data.

[0069] Cell Type Annotation is the process of identifying and classifying different cell types in biological samples. Cell type annotation helps to study cells and promote the development of biotechnology.

[0070] In the related art, the spatial transcriptome information is used to obtain the cell position of each spatial site of the biological sample tissue, and the corresponding cell type annotation of the biological sample tissue is determined based on the spatial position and single-cell sequencing data. However, due to the limitation of equipment resources, the resolution of the spatial transcriptome information obtained is usually not high; when the cells in the biological sample tissue aggregate to form a high-density cell aggregation area, there will be multiple cells at the same spatial site. In this case, it is difficult to accurately map the single-cell sequencing data with the spatial transcriptome information, resulting in low accuracy of the cell type annotation of the biological sample tissue in the related art.

[0071] Based on this, the embodiments of the present application provide an annotation model training method, a cell type annotation method, and related equipment, aiming to improve the accuracy of cell type annotation of biological sample tissues.

[0072] It should be noted that in the embodiments of the present application, when it comes to information related to user basic information, user identity or characteristics, the user's permission or consent will be obtained first, and the collection, use and processing of these data will comply with relevant laws, regulations and standards. In addition, when the embodiments of the present application need to obtain sensitive personal information of the user, the user's separate permission or separate consent will be obtained first. After the separate permission or separate consent of the user is clearly obtained, the necessary data for the normal operation of the embodiments of the present application will be obtained. For example, when obtaining sample tissue containing user sample tissue data to execute the training method or cell type annotation method of the annotation model proposed in the embodiments of the present application, the consent of the relevant user will be obtained, otherwise the sample tissue data used in the embodiments of the present application cannot be obtained.

[0073] The training method of the annotation model, the cell type annotation method and related equipment provided in the embodiments of the present application are specifically described through the following embodiments. First, the application scenario of the cell type annotation device in the embodiments of the present application is described, wherein the training method of the annotation model provided in the embodiments of the present application (for the convenience of description, it may also be referred to as "the training method" below) can be applied to the cell type annotation device (for the convenience of description, it may also be referred to as "the annotation device" below), and the annotation device is provided with the initial annotation model proposed in the embodiments of the present application. After the initial annotation model is trained by the training method, the annotation model can be obtained. Figure 1 As shown, Figure 1 : is a schematic diagram of an application scenario of the training device of the annotation model provided in the embodiment of the present application. First, the target spatial transcriptome data and target single-cell sequencing data corresponding to the target sample tissue are obtained, wherein: Figure 1 A small circle in the target space transcriptome data represents a spatial site, and a spatial site usually includes multiple single cells. However, due to various factors such as limited equipment resources, the acquired target space transcriptome data usually has a low resolution. Then, the target space transcriptome data and the target single cell sequencing data are input into the annotation device. The annotation device processes the target space transcriptome data and the target single cell sequencing data, and outputs the target cell annotation results that characterize the cell type results of each target single cell in the target sample tissue. In the target cell annotation results, multiple single cells at each spatial site can be labeled. Compared with the results obtained by cell annotation using related technical methods, the cell type annotation results corresponding to each single cell in the target sample tissue obtained in the embodiment of the present application are more accurate.

[0074] Next, the training method of the annotation model in the embodiment of the present application is specifically described through the following examples.

[0075] In the embodiment of the present application, the description will be made from the perspective of an annotation device, which can be integrated into a computer device, such as a server. Figure 2 As shown, Figure 2 is an optional flowchart of the training method of the annotation model provided in the embodiment of the present application, Figure 2 The method may include but is not limited to steps 101 to 105. When the annotation device executes the training method of the annotation model, the specific process is as follows. It should be noted that the present embodiment Figure 2 The order of step 101 to step 105 is not specifically limited, and the order of steps can be adjusted or some steps can be reduced or added according to actual needs.

[0076] Step 101, obtaining sample tissue data, as well as spatial transcriptome data and single-cell sequencing data corresponding to the sample tissue data, and inputting the spatial transcriptome data and single-cell sequencing data into an initial annotation model.

[0077] The following is a detailed description of step 101.

[0078] Among them, the sample tissue data is obtained from the sample tissue, and the sample tissue is a slice structure obtained based on a specific tissue or organ in the organism; the organism can be an individual with life activities including but not limited to microorganisms, plants, animals, etc. When the organism is an animal, the specific tissue can be epithelial tissue, muscle tissue, nervous tissue, etc., which are structures formed by cells with similar morphology, structure and function and intercellular matrix, and the organ can be a heart, lung, liver and other structures; after obtaining the sample tissue, the relevant biological equipment is used to convert the content in the sample tissue into sample tissue data that can be recognized and processed by a computer. The sample tissue data can be data in possible forms such as matrix form data, list form data or array form data; in addition, the sample tissue data can be converted into images according to actual needs, so that the annotation device or other related equipment can better process the data.

[0079] It should be noted that the embodiments of the present application do not limit the source of the sample tissue, as long as the sample tissue includes cells that need to be identified and annotated. At the same time, the embodiments of the present application do not limit the sample tissue data obtained based on the sample tissue, and the specific settings can be made according to actual conditions.

[0080] Among them, spatial transcriptome data is obtained through spatial transcriptomics technology, which includes rough gene spatial expression data and staining image data. Gene spatial expression data can reflect the average gene expression of each spatial site in the sample tissue; staining image data is the result of staining the sample tissue according to the gene spatial expression data to highlight different cell types or structures.

[0081] Among them, spatial transcriptomics technology refers to the use of specific probes (such as spatial barcode probes in the 10x Genomics Visium platform) to capture the gene expression corresponding to each spatial site in the sample tissue; spatial sites refer to different spatial regions in the sample tissue data. It is understandable that in fact each spatial site includes multiple single cells, while the cell annotation results obtained in the relevant technology are the average gene expression of each spatial site. In other words, the spatial transcriptome data obtained by the relevant technology can retain the spatial location information of the cell and the corresponding average gene expression information to a certain extent. The average gene expression information is the average gene expression of all single cells in the spatial site, but it is not accurate to the gene expression of a single cell, so the resolution is low.

[0082] Among them, single cell RNA sequencing (scRNA-seq) achieves sequencing by transcribing a single cell, which can provide a deeper understanding of the type and expression pattern of a single cell in a tissue, but scRNA-seq cannot obtain the original spatial location information of a single cell. In addition, in order to reveal the heterogeneous expression of cells in sample tissue data, single cell sequencing data usually has multiple data sets, which are used to transcribe single cells of different types.

[0083] Therefore, in order to enable the annotation device to ultimately output target cell annotation results at high resolution, the embodiment of the present application first needs to obtain the spatial transcriptome data and single-cell sequencing data corresponding to the sample tissue, wherein the annotation device can be provided with a device capable of processing the sample tissue to obtain the spatial transcriptome data and the single-cell sequencing data, or can be connected to relevant equipment through an interface to obtain the spatial transcriptome data and the single-cell sequencing data corresponding to the sample tissue. The specific settings can be made according to actual conditions, and the embodiment of the present application does not impose any restrictions on this; then, the acquired spatial transcriptome data and single-cell sequencing data are input into the initial annotation model, wherein the initial annotation model refers to a model that has not been trained by the training method proposed in the embodiment of the present application.

[0084] Step 102, using the spatial transcriptome data as a constraint condition, determining the reference annotation information corresponding to each target single cell in each spatial site of the sample tissue data according to the single-cell sequencing data.

[0085] The following is a detailed description of step 102.

[0086] In some embodiments, the spatial site of the sample tissue data is determined based on the spatial transcriptome data, and multiple target single cells in the spatial site are transcribed based on scRNA-seq to obtain reference annotation information corresponding to all target single cells in each spatial site. It is understandable that the reference annotation information can be accurate to the single cell level with high resolution. The reference annotation information corresponding to each target single cell is used to jointly process the target single cell with the predicted annotation information in the subsequent step to finally obtain the sample annotation result.

[0087] Among them, the embodiment of the present application refers to the single cells that need to be annotated during the training of the annotation model as target single cells; the target single cells can be all single cells in all spatial sites in the sample tissue data, or all single cells in some spatial sites in the sample tissue data, and the embodiment of the present application does not make specific restrictions on this.

[0088] In some embodiments, using spatial transcriptome data as a constraint condition, determining reference annotation information corresponding to each target single cell in each spatial site of sample tissue data according to single-cell sequencing data includes the following steps 201 to 203:

[0089] Step 201, using a preset segmentation tool to deconstruct the spatial transcriptome data, to obtain a cell number value representing the number of single cells in each spatial site of the sample tissue data, and a cell coordinate position representing the spatial position of each target single cell.

[0090] Step 202, using the single-cell sequencing data, determining the first gene expression information of the corresponding target single cell at each cell coordinate position, and determining a first logarithmic operation result according to the cell quantity value and the first gene expression information.

[0091] Step 203, adding the first logarithmic operation result to the preset batch variable value and the preset noise variable value to obtain reference annotation information of the target single cell.

[0092] Steps 201 to 203 are described in detail below.

[0093] Among them, the segmentation tool can be a deep learning tool for cell nucleus segmentation, such as a star segmentation tool (StarDist), a universal segmentation tool (Cellpose), etc. Of course, the segmentation tool can be selected according to actual conditions, and this application does not impose specific restrictions on this.

[0094] Exemplarily, after obtaining the spatial transcriptome data, the embodiment of the present application uses StarDist to perform structural processing on the spatial transcriptome data, and obtains a cell number value of 2 at spatial site A, indicating that there are 2 single cells in spatial site A; and, since the staining image data is static data, the cell coordinate positions of these 2 single cells can be further obtained, such as the cell coordinate position of the target single cell a1 in spatial site A is (x1, y1, z1), and the cell coordinate position of the target single cell a2 is (x2, y2, z2), where xyz is used to represent the three different dimensions of spatial coordinates.

[0095] Furthermore, after obtaining the number of target single cells contained in each spatial site and the corresponding cell coordinate information of each target single cell, the single-cell sequencing data is used to transcribe the target single cell at each cell coordinate position to obtain the first gene expression information; then, based on the first gene expression, the reference annotation information y of the target single cell is further obtained according to the following formula (1): i,g :

[0096]

[0097] Among them, y i,g is the observed gene expression of the g-th gene in the ith spatial site, N i is the total gene expression observed at the ith spatial site, λ i,g The relative gene expression of the g-th gene in the i-th spatial position; Poisson is the Poisson distribution, which represents y i,g It follows Poisson distribution; represents the gene expression of the gth gene in the i-th spatial site obtained by transcription of the k-th scRNA-seq; M represents the number of cells in the spatial site; is the result of the first logarithmic operation; γ g is the batch variable value, which is used to explain the batch effect of the g-th gene expression between scRNA-seq and spatial transcriptome data, so as to more flexibly handle the complex heterogeneity between spatial transcriptome data obtained from different platforms, and γ g It is obtained randomly and can be set according to the actual situation; ∈ g The noise variable value is introduced to more reasonably evaluate the similarities and differences between different target single cells, thereby promoting the effective integration of data in the processing process.

[0098] Step 103, based on the reference annotation information and the spatial transcriptome data, determine the multimodal prior information corresponding to each target single cell, and determine the predicted annotation information corresponding to the target single cell according to the multimodal prior information.

[0099] The following is a detailed description of step 103.

[0100] In some embodiments, since the reference annotation information is obtained by transcription based on the cell number value and the cell coordinate position, it is impossible to determine which target single cell in the spatial site the first gene expression information obtained by transcription from scRNA-seq, that is, the first gene expression information cannot be well mapped with the spatial transcriptome data, so the reference annotation information is obtained. The average gene expression of all target single cells in the spatial site has a low resolution. The embodiments of this application will further process the spatial transcriptome data based on the reference annotation information to obtain predictive annotation information for enhancing the resolution of the final output result.

[0101] In some embodiments, based on the reference annotation information and the spatial transcriptome data, determining the multimodal prior information corresponding to each target single cell includes the following steps 301 to 303:

[0102] Step 301, determining the cell spatial distribution characteristics of the target single cell at the corresponding spatial site according to the reference annotation information corresponding to the target single cell and the reference annotation information corresponding to the neighboring single cells of the target single cell.

[0103] Step 302, using a preset recognition tool, performs morphological recognition processing on the corresponding target single cell at the cell coordinate position to obtain the cell morphological characteristics of the target single cell.

[0104] Step 303: Based on the cell spatial distribution characteristics and cell morphology characteristics, the multimodal prior information corresponding to each single cell is determined.

[0105] Steps 301 to 303 are described in detail below.

[0106] In some embodiments, in order to further improve the resolution of target single cell annotation, it is also necessary to determine the multimodal prior information corresponding to the target single cell based on the reference annotation information and the spatial transcriptome data. The multimodal prior information can further reveal the cell morphological characteristics of each target single cell in the spatial site, as well as the cell spatial characteristic connection between the target single cell and the adjacent single cells, so that the annotation device can have a deeper understanding of each target single cell, thereby improving the accuracy of cell annotation.

[0107] Furthermore, the different cell morphological characteristics of each single cell can reflect its biological classification, cell type, physiological function, reproduction mode, disease information, etc. For example, in cancer sample tissue data, tumor cells usually have more irregular shapes, larger nuclear / cytoplasmic ratios, and more abundant organelles than normal cells, which all reflect the malignant transformation process of tumor cells. In this way, the different cell morphological characteristics of different cells can be used to improve the annotation ability of the annotation device for cells.

[0108] The neighboring single cells of the target single cell refer to other single cells that are adjacent or close to the target single cell in spatial position, and the spatial position here includes two-dimensional space and three-dimensional space.

[0109] Furthermore, after understanding the spatial distribution of cells at each spatial site based on the reference annotation information, morphological recognition processing is performed on each target single cell in the spatial site to obtain the cell morphological characteristics of each target single cell, so that the subsequent annotation device can better map the results obtained by transcribing the single-cell sequencing data with multiple target single cells in the spatial site based on the cell morphological characteristics, so as to improve the accuracy of cell annotation.

[0110] Among them, the recognition tool can be a classic cell profile tool (cell profile), a tool obtained through convolutional neural network training, etc. The recognition tool can be specifically selected according to actual conditions, and the embodiments of the present application do not limit this.

[0111] In some embodiments, determining the cell spatial distribution characteristics of the target single cell at the corresponding spatial site according to the reference annotation information corresponding to the target single cell and the reference annotation information corresponding to the neighboring single cells of the target single cell includes the following steps 401 to 403:

[0112] Step 401: If the reference annotation information corresponding to the target single cell is different from the reference annotation information corresponding to the neighboring single cell, a first constraint value is obtained.

[0113] Step 402: If the reference annotation information corresponding to the target single cell is the same as the reference annotation information corresponding to the neighboring single cell, a second constraint value is obtained.

[0114] Step 403: Based on the first constraint value or the second constraint value, obtain the cell spatial distribution characteristics characterizing the target single cell at the corresponding spatial location.

[0115] Steps 401 to 403 are described in detail below.

[0116] In some embodiments, after determining the reference annotation information corresponding to the current target single cell and the reference annotation information corresponding to the neighboring single cells of the target single cell, the cell spatial distribution feature v of the target single cell at the corresponding spatial location is determined based on the following formula (2):

[0117]

[0118] in, is the indicator function, recording the reference annotation information k corresponding to the target single cell i,m , the reference annotation information of the mth single cell at the ith spatial site is recorded as k i′,m′ ; when k i,m =k i′m′ When k i,m ≠k i′m′ When , the second constraint value is obtained, and the second constraint value can be 0; the cell spatial distribution feature ν is used to represent the spatial smoothness of the target single cell at the corresponding spatial site. The larger ν is, the smoother the distribution of each target single cell in the spatial site is; the cell distribution feature can be realized by the Potts model set in the annotation device.

[0119] Furthermore, for example, in the central nervous system, neurons of the same type are often arranged continuously along specific anatomical paths to form nerve nuclei or nerve fiber clusters. This orderly spatial distribution is closely related to the relevant functions of neurons. Determining the spatial distribution characteristics of cells helps the annotation model to more finely understand the shared features between each target single cell, thereby helping to improve the ability to annotate cells in the future.

[0120] In some embodiments, determining the prediction annotation information corresponding to the target single cell according to the multimodal prior information includes the following steps 501 to 505:

[0121] Step 501, predicting the cell type of the target single cell based on the cell spatial distribution characteristics to obtain first prior information.

[0122] Step 502: predict the cell type of the target single cell based on the cell morphology characteristics to obtain second prior information.

[0123] Step 503, calculating the product of the first priori information and the second priori information to obtain multimodal product data.

[0124] Step 504, using the multimodal product data, determining the second gene expression information corresponding to the target single cell at each cell coordinate position;

[0125] Step 505: Determine the corresponding prediction annotation information of the target single cell based on the second gene expression information.

[0126] Steps 501 to 505 are described in detail below.

[0127] In some embodiments, the cell type of the target single cell is predicted based on the multimodal prior information to obtain multiple prior information; the prior information is an important reference for determining the cell type of the target single cell, which can help the annotation device enhance the accuracy of the cell type annotation.

[0128] Furthermore, the cell type of the target single cell is predicted by the following formula (3) to obtain the first prior information:

[0129]

[0130] in, is the spatial distribution characteristic of cells; Z is the normalization constant; is the set of neighboring points of the mth cell at the ith spatial position; exp() is an exponential operation with the natural logarithm base e as the base.

[0131] Furthermore, the classifier set in the annotation device is used to predict the cell type of the target single cell based on the cell morphology characteristics to obtain the second prior information p(k i,m ∣E i,m ), where E i,m ∈R H ,H is the dimension of cell morphological characteristics. Among them, the second prior information p(k i,m ∣E i,m ) essentially corresponds to the conditional distribution probability under the classification problem. Therefore, the constructed classifier is trained using the pre-collected data set to approximate the expected result p(k i,m ∣E i,m ), the classifier can be

[0132] Among them, the classifier can be a support vector machine, a random forest classifier, a neural network classifier, etc. The specific type of the classifier can be set according to the actual situation, and the embodiment of the present application does not limit this. The classifier can be set in the initial annotation model to be jointly trained with the initial annotation model during the training process to improve the generalization ability of the annotation model obtained after training.

[0133] Furthermore, the multimodal product data is obtained as p(k i,m ∣k -{i,m} ,E i,m )=p(k i,m ∣E i,m )×p(k i,m ∣k -{i,m} Then, the second gene expression information p(ki,m ∣k -{i,m} ,E i,m ):

[0134] p(k i,m ∣k -{i,m} ,E i,m )=p(k i,m ∣E i,m )×p(k i,m ∣k -{i,m} )×const (4)

[0135] Among them, const is the error constant, which can be adaptively adjusted according to actual conditions.

[0136] Further, in the case of determining the second gene expression information, the logarithmic operation result of the second gene expression information is determined as the prediction annotation information logp(k i,m ∣k -{i,m} ,E i,m ).

[0137] Step 104 , performing joint annotation processing on each target single cell according to the reference annotation information and the predicted annotation information, and obtaining a sample annotation result corresponding to the sample tissue data.

[0138] In some embodiments, after obtaining reference annotation information and predicted annotation information, each target single cell in the spatial site is jointly annotated using the reference annotation information and the predicted annotation information. By complementing the reference annotation information and the predicted annotation information, the reference annotation information can be better mapped with the spatial transcriptome data, thereby improving the accuracy and comprehensiveness of the annotation model's annotation of the target single cells.

[0139] In some embodiments, joint annotation processing is performed on each target single cell according to the reference annotation information and the predicted annotation information to obtain a sample annotation result corresponding to the sample tissue data, including the following steps 601 to 602:

[0140] Step 601: Update reference annotation information based on a preset first weight parameter, and update predicted annotation information based on a preset second weight parameter.

[0141] Step 602 : performing joint annotation processing on each target single cell according to the updated reference annotation information and the updated predicted annotation information to obtain a sample cell annotation result corresponding to the sample tissue data.

[0142] Steps 601 to 602 are described in detail below.

[0143] In some embodiments, the reference annotation information and the predicted annotation information are jointly processed using the maximum a posteriori estimation method to determine the sample cell annotation result k i,m , where i=1,2,…I, m=1,2,…,M.

[0144]

[0145] Among them, logp(k i,m ) and logp(k i,m ∣k -{i,m} ,E i,m ) is equivalent to log(λ i,g )and equivalent; argmax is used to indicate the processing result with the largest value returned; specifically, logp(k i，m ) is used as the prior term of formula (5), As the likelihood term of formula (5), the prior information is incorporated into the annotation model estimation process, and the sample cell annotation results are obtained by maximizing the posterior method.

[0146] Furthermore, the annotation model can also set a first weight parameter such as α and a second weight parameter β for the reference annotation information. In another embodiment, the sample annotation result can also be determined by the following formula (6):

[0147]

[0148] in, is the updated reference annotation information, βlogp(k i,m ) is the updated prediction annotation information; then, each target single cell is jointly annotated using the same method as formula (5) to obtain the sample cell annotation result corresponding to the sample tissue data

[0149] It should be noted that the first weight parameter and the second weight parameter can be continuously and adaptively adjusted during the training process until the trained annotation model is obtained, that is, the embodiment of the present application does not limit the specific values ​​of the first weight parameter and the second weight parameter.

[0150] In some embodiments, the second weight parameter is obtained by following steps 701 to 703:

[0151] Step 701, obtaining the sample verification result corresponding to the sample tissue data, as well as the initial second weight parameter and the initial adjustment ratio.

[0152] Step 702: If the sample annotation result is different from the sample verification result, the initial adjustment ratio is updated to obtain an adjustment ratio.

[0153] Step 703: Update the initial second weight parameter based on the adjustment ratio to obtain the second weight parameter.

[0154] Steps 701 to 703 are described in detail below.

[0155] In some embodiments, since the second weight parameter is related to the predicted annotation result, the second weight parameter can be quickly adjusted by comparing the difference between the predicted annotation information and the reference annotation information to improve the efficiency of obtaining the second weight parameter during the training process.

[0156] Furthermore, an initial second weight parameter and an initial adjustment ratio characterizing the adjustment rate of the initial second weight parameter are assigned to the annotation model at the initial stage of training; then, after the annotation model completes the output of a sample annotation result, it can be compared with the sample verification result obtained in advance. If the sample tissue data is different from the sample verification result, or the difference between the sample tissue data and the sample verification result exceeds a preset threshold, the initial adjustment ratio is updated to obtain the adjustment ratio; then, the adjustment ratio is multiplied by the initial second weight parameter to obtain the second weight parameter.

[0157] It is understandable that if the sample tissue data is different from the sample verification result, it means that the annotation model is still not accurate enough in mapping the annotation result to the target single cell when performing joint processing based on the predicted annotation result and the reference annotation result. When the reference annotation result is accurately obtained from the single-cell sequencing data, the accuracy of the annotation result mapping is improved by adjusting the annotation model's attention to the predicted annotation result (the second weight parameter); and, since the adjustment ratio is introduced, the updated second weight parameter can be obtained in parallel while determining the first weight parameter, which speeds up the annotation model's next output result, thereby improving the overall training efficiency of the annotation model.

[0158] It should be noted that the initial second weight parameter and the initial adjustment ratio can be set according to actual conditions, and the embodiments of the present application do not limit this.

[0159] Step 105: adjust the model parameters of the initial annotation model according to the sample annotation results to obtain a trained annotation model.

[0160] The following is a detailed description of step 105.

[0161] In some embodiments, in order to better implement the training of the initial annotation model, the sample verification results corresponding to the sample tissue data can be obtained in advance. The sample verification results are used to characterize the output results of the initial annotation model under ideal conditions after the sample tissue data is input into the initial annotation model; the sample verification results are compared with the actual output sample annotation results, and the loss value calculated based on the sample verification results and the sample annotation results can be used to adjust the model parameters of the initial annotation model based on the loss value. On the basis of continuously iterating the sample annotation results, the model parameters are continuously updated until the preset training conditions are reached to obtain the trained annotation model.

[0162] Among them, the model parameters can be the learning rate, the number and / or size of hidden layers of the convolutional network, the convolution kernel size, the regularization parameter, etc. The model parameters that need to be adjusted for the annotation model can be set according to actual conditions, and this application does not impose any restrictions on this.

[0163] Among them, the training condition can be reaching a preset training time, or reaching a preset number of training times, etc. The embodiment of the present application does not limit this and can be adjusted according to actual conditions.

[0164] In addition, the annotation device can also execute the cell type annotation method proposed in the embodiment of the present application, such as Figure 3 As shown, Figure 3 is an optional flow chart of the cell type annotation method provided in the embodiment of the present application, Figure 3 The method may include but is not limited to steps 801 to 807. When the annotation device executes the cell type annotation method, the specific process is as follows. It should be noted that this embodiment Figure 3 The order of step 801 to step 807 is not specifically limited, and the order of steps can be adjusted or some steps can be reduced or added according to actual needs.

[0165] Step 801, obtaining sample tissue data, as well as spatial transcriptome data and single-cell sequencing data corresponding to the sample tissue data, and inputting the spatial transcriptome data and single-cell sequencing data into an initial annotation model.

[0166] Step 802, using the spatial transcriptome data as a constraint condition, determining the reference annotation information corresponding to each target single cell in each spatial site of the sample tissue data according to the single-cell sequencing data.

[0167] Step 803, based on the reference annotation information and the spatial transcriptome data, determine the multimodal prior information corresponding to each target single cell, and determine the predicted annotation information corresponding to the target single cell according to the multimodal prior information.

[0168] Step 804 , performing joint annotation processing on each target single cell according to the reference annotation information and the predicted annotation information, and obtaining a sample annotation result corresponding to the sample tissue data.

[0169] Step 805: adjust the model parameters of the initial annotation model according to the sample annotation results to obtain a trained annotation model.

[0170] Step 806, obtaining target tissue data, and target spatial transcriptome data and target single-cell sequencing data corresponding to the target tissue data, and inputting the target spatial transcriptome data and target single-cell sequencing data into the trained annotation model.

[0171] Step 807, performing joint annotation processing on the target tissue data according to the target spatial transcriptome data and the target single-cell sequencing data, to obtain target cell annotation results corresponding to the target tissue data.

[0172] Steps 801 to 807 are described in detail below.

[0173] Among them, the method of step 801 to step 805 is similar to that of step 101 to step 105, which will not be repeated here. After obtaining the trained annotation model, the target spatial transcriptome data and target single-cell sequencing data of the acquired target tissue data are input into the annotation model, and then the annotation model jointly processes the input data based on the adjusted model parameters, and outputs the target cell annotation results corresponding to the target tissue data.

[0174] Among them, the output target cell annotation result can be expressed in the form of matrix data, and can be converted into image expression and other forms according to actual needs. The embodiment of the present application does not limit this.

[0175] It can be understood that the annotation model can improve the accuracy of the mapping between the reference annotation information transcribed from the target single-cell sequencing data and the target single cell based on the multimodal prior information obtained from the target spatial transcriptome data, thereby enabling the resolution of the target cell annotation results to be accurate to each target single cell in each spatial site, thereby improving the accuracy of cell type annotation of sample tissue data and the quality of the output target cell annotation results.

[0176] like Figure 4 As shown, Figure 4 : is an optional functional module diagram of a cell type annotation device provided in an embodiment of the present application, wherein the cell type annotation device may include:

[0177] The acquisition module 901 is used to acquire sample tissue data, as well as spatial transcriptome data and single-cell sequencing data corresponding to the sample tissue data, and input the spatial transcriptome data and single-cell sequencing data into the initial annotation model.

[0178] The reference annotation module 902 is used to determine the reference annotation information corresponding to each target single cell in each spatial site of the sample tissue data based on the single-cell sequencing data, taking the spatial transcriptome data as a constraint condition.

[0179] The prediction annotation module 903 is used to determine the multimodal prior information corresponding to each target single cell based on the reference annotation information and the spatial transcriptome data, and determine the prediction annotation information corresponding to the target single cell according to the multimodal prior information.

[0180] The annotation processing module 904 is used to perform joint annotation processing on each target single cell according to the reference annotation information and the predicted annotation information to obtain a sample annotation result corresponding to the sample tissue data.

[0181] The training module 905 is used to adjust the model parameters of the initial annotation model according to the sample annotation results to obtain a trained annotation model.

[0182] The first target processing module 906 is used to obtain target tissue data, and target spatial transcriptome data and target single-cell sequencing data corresponding to the target tissue data, and input the target spatial transcriptome data and target single-cell sequencing data into the trained annotation model.

[0183] The second target processing module 907 is used to perform joint annotation processing on the target tissue data according to the target spatial transcriptome data and the target single-cell sequencing data to obtain the target cell annotation results corresponding to the target tissue data.

[0184] The specific implementation of the cell type annotation device is basically the same as the specific embodiment of the above-mentioned cell type annotation method, and will not be repeated here.

[0185] Furthermore, the cell type annotation device provided in the present application can be used in downstream analysis scenarios such as bioinformatics, and the downstream analysis scenarios include but are not limited to cell-cell type spatial distribution scenarios, immune microenvironment scenarios, cell-cell interaction scenarios, cell-environment interaction scenarios, etc.; of course, downstream analysis scenarios can also include more aspects, and the embodiments of the present application do not limit this.

[0186] The embodiment of the present application also provides an electronic device, the electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the above-mentioned cell type annotation method when executing the computer program. The electronic device can be any intelligent terminal including a tablet computer, a car computer, etc.

[0187] like Figure 5 As shown, Figure 5 : is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present application, the electronic device includes:

[0188] The processor 1001 may be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present application;

[0189] The memory 1002 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 1002 can store an operating system and other application programs. When the technical solution provided in the embodiments of this specification is implemented by software or firmware, the relevant program code is stored in the memory 1002, and the processor 1001 calls and executes the cell type annotation method of the embodiment of this application;

[0190] Input / output interface 1003, used to implement information input and output;

[0191] The communication interface 1004 is used to realize the communication interaction between the device and other devices. The communication can be realized through a wired manner (such as USB, network cable, etc.) or a wireless manner (such as mobile network, WIFI, Bluetooth, etc.);

[0192] A bus 1005 , which transmits information between various components of the device (e.g., the processor 1001 , the memory 1002 , the input / output interface 1003 , and the communication interface 1004 );

[0193] The processor 1001 , the memory 1002 , the input / output interface 1003 and the communication interface 1004 are connected to each other in communication within the device via the bus 1005 .

[0194] An embodiment of the present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the above-mentioned cell type annotation method is implemented.

[0195] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0196] The embodiments described in the embodiments of the present application are intended to more clearly illustrate the technical solutions of the embodiments of the present application and do not constitute a limitation on the technical solutions provided in the embodiments of the present application. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0197] Those skilled in the art will appreciate that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.

[0198] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0199] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices may be implemented as software, firmware, hardware, or a suitable combination thereof.

[0200] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0201] It should be understood that in the present application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0202] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the above units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0203] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0204] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0205] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including multiple instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, referred to as ROM), random access memory (Random Access Memory, referred to as RAM), disk or optical disk and other media that can store programs.

[0206] The preferred embodiments of the present invention are described above with reference to the accompanying drawings, but the scope of the rights of the present invention is not limited thereto. Any modification, equivalent substitution and improvement made by a person skilled in the art without departing from the scope and essence of the present invention should be within the scope of the rights of the present invention.

Claims

1. A training method for an annotation model, characterized in that: The method comprises: Acquire sample tissue data, as well as spatial transcriptome data and single-cell sequencing data corresponding to the sample tissue data, and input the spatial transcriptome data and the single-cell sequencing data into an initial annotation model; Taking the spatial transcriptome data as a constraint condition, determining the reference annotation information corresponding to each target single cell in each spatial site of the sample tissue data according to the single-cell sequencing data; Based on the reference annotation information and the spatial transcriptome data, determining the multimodal prior information corresponding to each of the target single cells, and determining the predicted annotation information corresponding to the target single cell according to the multimodal prior information; According to the reference annotation information and the predicted annotation information, joint annotation processing is performed on each of the target single cells to obtain a sample annotation result corresponding to the sample tissue data; The model parameters of the initial annotation model are adjusted according to the sample annotation results to obtain a trained annotation model.

2. The method according to claim 1, characterized in that The method of determining the reference annotation information corresponding to each target single cell in each spatial site of the sample tissue data based on the single-cell sequencing data using the spatial transcriptome data as a constraint condition includes: Deconstructing the spatial transcriptome data using a preset segmentation tool to obtain a cell number value representing the number of single cells in each of the spatial sites of the sample tissue data, and a cell coordinate position representing the spatial position of each of the target single cells; Determine the first gene expression information of the corresponding target single cell at each cell coordinate position using the single cell sequencing data, and determine a first logarithmic operation result according to the cell quantity value and the first gene expression information; The first logarithmic operation result is added to a preset batch variable value and a preset noise variable value to obtain reference annotation information of the target single cell.

3. The method according to claim 2, characterized in that The step of determining the multimodal prior information corresponding to each of the target single cells based on the reference annotation information and the spatial transcriptome data includes: Determine the cell spatial distribution characteristics of the target single cell at the corresponding spatial site according to the reference annotation information corresponding to the target single cell and the reference annotation information corresponding to the neighboring single cells of the target single cell; Using a preset recognition tool, performing morphological recognition processing on the target single cell corresponding to the cell coordinate position to obtain the cell morphological characteristics of the target single cell; Based on the cell spatial distribution characteristics and the cell morphological characteristics, the multimodal prior information corresponding to each of the single cells is determined.

4. The method according to claim 3, characterized in that Determining the prediction annotation information corresponding to the target single cell according to the multimodal prior information includes: Predicting the cell type of the target single cell based on the cell spatial distribution characteristics to obtain first prior information; Predicting the cell type of the target single cell based on the cell morphology characteristics to obtain second prior information; Calculating the product of the first priori information and the second priori information to obtain multimodal product data; Determining second gene expression information corresponding to the target single cell at each of the cell coordinate positions using the multimodal product data; Based on the second gene expression information, the corresponding prediction annotation information of the target single cell is determined.

5. The method according to claim 3, characterized in that: The step of determining the cell spatial distribution characteristics of the target single cell at the corresponding spatial site according to the reference annotation information corresponding to the target single cell and the reference annotation information corresponding to the neighboring single cells of the target single cell comprises: If the reference annotation information corresponding to the target single cell is different from the reference annotation information corresponding to the neighbor single cell, obtaining a first constraint value; If the reference annotation information corresponding to the target single cell is the same as the reference annotation information corresponding to the neighbor single cell, a second constraint value is obtained; Based on the first constraint value or the second constraint value, a cell spatial distribution characteristic characterizing the target single cell at the corresponding spatial location is obtained.

6. The method according to claim 5, characterized in that The step of performing joint annotation processing on each of the target single cells according to the reference annotation information and the predicted annotation information to obtain a sample annotation result corresponding to the sample tissue data includes: Based on a preset first weight parameter, updating the reference annotation information, and based on a preset second weight parameter, updating the predicted annotation information; According to the updated reference annotation information and the updated predicted annotation information, joint annotation processing is performed on each of the target single cells to obtain a sample cell annotation result corresponding to the sample tissue data.

7. The method according to claim 6, characterized in that The second weight parameter is obtained by the following steps, which include: Obtaining a sample verification result corresponding to the sample tissue data, as well as an initial second weight parameter and an initial adjustment ratio; If the sample annotation result is different from the sample verification result, updating the initial adjustment ratio to obtain an adjustment ratio; The initial second weight parameter is updated based on the adjustment ratio to obtain a second weight parameter.

8. A cell type annotation method, characterized in that: The method comprises: Acquire sample tissue data, as well as spatial transcriptome data and single-cell sequencing data corresponding to the sample tissue data, and input the spatial transcriptome data and the single-cell sequencing data into an initial annotation model; Taking the spatial transcriptome data as a constraint condition, determining the reference annotation information corresponding to each target single cell in each spatial site of the sample tissue data according to the single-cell sequencing data; Based on the reference annotation information and the spatial transcriptome data, determining the multimodal prior information corresponding to each of the target single cells, and determining the predicted annotation information corresponding to the target single cell according to the multimodal prior information; According to the reference annotation information and the predicted annotation information, joint annotation processing is performed on each of the target single cells to obtain a sample annotation result corresponding to the sample tissue data; Adjusting the model parameters of the initial annotation model according to the sample annotation result to obtain a trained annotation model; Acquire target tissue data, and target spatial transcriptome data and target single-cell sequencing data corresponding to the target tissue data, and input the target spatial transcriptome data and target single-cell sequencing data into the trained annotation model; According to the target spatial transcriptome data and the target single-cell sequencing data, the target tissue data is jointly annotated to obtain a target cell annotation result corresponding to the target tissue data.

9. A cell type annotation device, characterized in that: The device comprises: An acquisition module, used to acquire sample tissue data, and spatial transcriptome data and single-cell sequencing data corresponding to the sample tissue data, and input the spatial transcriptome data and the single-cell sequencing data into an initial annotation model; A reference annotation module, used to determine the reference annotation information corresponding to each target single cell in each spatial site of the sample tissue data according to the single-cell sequencing data, taking the spatial transcriptome data as a constraint condition; A prediction annotation module, used to determine the multimodal prior information corresponding to each of the target single cells based on the reference annotation information and the spatial transcriptome data, and determine the prediction annotation information corresponding to the target single cell according to the multimodal prior information; An annotation processing module, used to perform joint annotation processing on each of the target single cells according to the reference annotation information and the predicted annotation information, to obtain a sample annotation result corresponding to the sample tissue data; A training module, used for adjusting the model parameters of the initial annotation model according to the sample annotation result to obtain a trained annotation model; A first target processing module is used to obtain target tissue data, and target spatial transcriptome data and target single-cell sequencing data corresponding to the target tissue data, and input the target spatial transcriptome data and target single-cell sequencing data into the trained annotation model; The second target processing module is used to perform joint annotation processing on the target tissue data according to the target spatial transcriptome data and the target single-cell sequencing data to obtain target cell annotation results corresponding to the target tissue data.

10. An electronic device, characterized in that: The electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the training method of the annotation model described in any one of claims 1 to 7 or the cell type annotation method described in claim 8 when executing the computer program.

11. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, it implements the training method of the annotation model described in any one of claims 1 to 7 or the cell type annotation method described in claim 8.

Images