Regulation and control element identification method and device, equipment, storage medium and program product

Abstract

The invention relates to a regulation and control element recognition method and device, equipment, a storage medium and a program product, relates to the technical field of biological information, and can improve the recognition accuracy and efficiency of a regulation and control element in a complex regulation and control scene. The method comprises the following steps: acquiring single cell chromatin accessibility sequencing data and single cell expression quantity sequencing data, and acquiring a regulation relationship set to be adjusted; according to the regulatory relation set and the single cell chromatin accessibility sequencing data, obtaining expression quantity prediction results of the multiple genes, and according to the regulatory relation set and the single cell chromatin accessibility sequencing data, obtaining accessibility prediction results of the multiple regulatory elements; adjusting the regulation relation set according to the expression quantity prediction result, the single cell expression quantity sequencing data, the accessibility prediction result and the single cell chromatin accessibility sequencing data to obtain an adjusted regulation relation set; and determining a regulation and control element identification result according to the adjusted regulation and control relation set.

Description

Technical Field

[0001] This application relates to the field of bioinformatics, and in particular to a method, apparatus, computer device, computer-readable storage medium, and computer program product for identifying regulatory elements. Background Technology

[0002] Regulatory elements are sequence segments in the genome that do not directly encode proteins or RNA, but can regulate the spatiotemporal patterns and intensity of gene expression by interacting with specific molecules. Taking cis-regulatory elements (CREs) as an example, as a core component of the gene expression regulatory network, cis-regulatory elements can precisely regulate the spatiotemporal expression patterns of genes by specifically binding to transcription factors and other regulatory proteins.

[0003] Among related technologies, techniques such as Signac (a toolkit for single-cell epigenetic regulation analysis) and ArchR (a tool for analyzing single-cell chromatin accessibility data) can be used to mine regulatory elements based on multi-omics data. However, these methods mainly rely on correlation analysis between genes and regulatory elements, are limited to simple association models, and are unable to fully characterize the regulatory relationships in complex regulatory networks. Moreover, they mainly focus on gene neighborhoods and fail to effectively include distant regulatory elements, making it difficult to resolve regulatory mechanisms that span long distances across the genome. Therefore, these technologies struggle to efficiently and accurately identify the complex and multidimensional regulatory relationships between regulatory elements and genes. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, apparatus, computer device, computer-readable storage medium, and computer program product for identifying control elements in response to the above-mentioned technical problems.

[0005] In a first aspect, this application provides a method for identifying control elements, including:

[0006] Single-cell chromatin accessibility sequencing data and single-cell expression level sequencing data are obtained to obtain a set of regulatory relationships to be adjusted; the single-cell chromatin accessibility sequencing data includes sequencing data reflecting the chromatin accessibility of multiple regulatory elements, the single-cell expression level sequencing data includes sequencing data reflecting the expression of multiple genes on chromosomes, and the parameter value corresponding to each regulatory relationship in the set of regulatory relationships characterizes the regulatory strength of the regulatory element on the expression of the corresponding gene.

[0007] Based on the set of regulatory relationships and the single-cell chromatin accessibility sequencing data, the expression level prediction results of multiple genes are obtained, and based on the set of regulatory relationships and the single-cell expression level sequencing data, the accessibility prediction results of multiple regulatory elements are obtained.

[0008] Based on the expression level prediction results, the single-cell expression level sequencing data, the accessibility prediction results, and the single-cell chromatin accessibility sequencing data, the set of regulatory relationships is adjusted to obtain an adjusted set of regulatory relationships.

[0009] Based on the adjusted set of control relationships, the identification result of the control element is determined.

[0010] In one embodiment, adjusting the set of regulatory relationships based on the expression level prediction results, the single-cell expression level sequencing data, the accessibility prediction results, and the single-cell chromatin accessibility sequencing data includes:

[0011] The first loss value is determined based on the gene expression level difference between the predicted expression level and the single-cell expression level sequencing data;

[0012] A second loss value is determined based on the accessibility difference between the accessibility prediction results and the single-cell chromatin accessibility sequencing data;

[0013] The weighted regularization loss value is determined based on the parameter value corresponding to each regulation relationship in the set of regulation relationships;

[0014] The target loss value is determined based on the first loss value, the second loss value, and the weighted regularization loss value;

[0015] The set of control relationships is adjusted based on the target loss value.

[0016] In one embodiment, the regulatory element includes a cis-regulatory element, and the single-cell expression sequencing data includes sequencing data reflecting the expression of multiple genes on the same chromosome; the differences in gene expression levels are determined by the following steps:

[0017] Based on the single-cell expression sequencing data, the first expression level of each of the multiple genes on the same chromosome is determined, and based on the expression level prediction results, the second expression level of each of the multiple genes on the same chromosome is determined;

[0018] For each of the genes on the same chromosome in the same cell, a first difference between the first expression level and the second expression level is obtained;

[0019] The differences in gene expression levels are determined based on each of the first differences.

[0020] In one embodiment, the regulatory element includes a cis-regulatory element, and the single-cell chromatin accessibility sequencing data includes sequencing data reflecting the chromatin accessibility of each of the multiple cis-regulatory elements on the same chromosome; the accessibility differences are determined by the following steps:

[0021] Based on the single-cell chromatin accessibility sequencing data, the first chromatin accessibility of each of the multiple cis-regulatory elements on the same chromosome is determined, and based on the accessibility prediction results, the second chromatin accessibility of each of the multiple cis-regulatory elements on the same chromosome is determined.

[0022] For each of the cis-regulatory elements on the same chromosome of the same cell, a second difference between the first chromatin accessibility and the second chromatin accessibility is determined;

[0023] The accessibility difference is determined based on each of the second differences.

[0024] In one embodiment, obtaining the expression level prediction results of multiple genes based on the set of regulatory relationships and the single-cell chromatin accessibility sequencing data includes:

[0025] For each of the genes on the same chromosome in the same cell, the regulatory strength of the multiple regulatory elements on the same chromosome on the expression of the gene is determined according to the set of regulatory relationships, and the first chromatin accessibility of the multiple regulatory elements on the same chromosome in the same cell is determined according to the single-cell chromatin accessibility sequencing data.

[0026] Based on the regulatory strength of multiple regulatory elements on the same chromosome on the expression of the gene and the accessibility of multiple regulatory elements on the same chromosome in the first chromatin of the same cell, the predicted regulatory contribution value of each regulatory element to the gene is determined; the predicted regulatory contribution value is used to indicate the likelihood that the regulatory element will regulate the expression of the gene.

[0027] The predicted expression level of the gene is determined based on the predicted contribution of each regulatory element to the regulation of the gene.

[0028] In one embodiment, obtaining accessibility prediction results for multiple regulatory elements based on the set of regulatory relationships and the single-cell expression level sequencing data includes:

[0029] For each of the regulatory elements on the same chromosome in the same cell, the regulatory strength of the regulatory elements on the expression of multiple genes on the same chromosome is determined according to the set of regulatory relationships, and the first expression level of each of the multiple genes on the same chromosome is determined according to the single-cell expression level sequencing data.

[0030] Based on the regulatory strength of the regulatory elements on the expression of multiple genes on the same chromosome and the first expression level of each of the multiple genes on the same chromosome, a regulatory association prediction value for each of the genes is determined; the regulatory association prediction value is used to indicate the association between the regulatory element and the regulation of gene expression.

[0031] Based on the predicted regulatory association values ​​of the regulatory elements with each of the genes, the accessibility prediction results of the regulatory elements are determined.

[0032] In one embodiment, determining the control element identification result based on the adjusted control relationship set includes:

[0033] Based on the adjusted set of regulatory relationships, multiple regulatory relationship pairs are determined; the regulatory relationship pairs are used to associate genes and regulatory elements that have a regulatory relationship with the genes.

[0034] The significance assessment result of the regulatory relationship pair is determined based on the first significance probability and the second significance probability of the regulatory relationship pair; the first significance probability reflects the significance of the gene in each regulatory element that has a regulatory relationship with the gene, and the second significance probability reflects the significance of the regulatory element in each gene that has a regulatory relationship with the regulatory element.

[0035] For each gene in the adjusted set of regulatory relationships, based on the significance evaluation results of each pair of regulatory relationships, a target regulatory element is selected from each regulatory element that has a regulatory relationship with the gene, and the regulatory element identification result of the gene is obtained.

[0036] Secondly, this application also provides a control element identification device, comprising:

[0037] The first acquisition module is used to acquire single-cell chromatin accessibility sequencing data and single-cell expression level sequencing data, and to acquire a set of regulatory relationships to be adjusted; the single-cell chromatin accessibility sequencing data includes sequencing data reflecting the chromatin accessibility of multiple regulatory elements, the single-cell expression level sequencing data includes sequencing data reflecting the expression of multiple genes on chromosomes, and the parameter value corresponding to each regulatory relationship in the set of regulatory relationships characterizes the regulatory strength of the regulatory element on the expression of the corresponding gene;

[0038] The second acquisition module is used to acquire the expression level prediction results of multiple genes based on the set of regulatory relationships and the single-cell chromatin accessibility sequencing data, and to acquire the accessibility prediction results of multiple regulatory elements based on the set of regulatory relationships and the single-cell expression level sequencing data.

[0039] An adjustment module is used to adjust the set of regulatory relationships based on the expression level prediction results, the single-cell expression level sequencing data, the accessibility prediction results, and the single-cell chromatin accessibility sequencing data, to obtain an adjusted set of regulatory relationships.

[0040] The determination module is used to determine the identification result of the control element based on the adjusted control relationship set.

[0041] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0042] Single-cell chromatin accessibility sequencing data and single-cell expression level sequencing data are obtained to obtain a set of regulatory relationships to be adjusted; the single-cell chromatin accessibility sequencing data includes sequencing data reflecting the chromatin accessibility of multiple regulatory elements, the single-cell expression level sequencing data includes sequencing data reflecting the expression of multiple genes on chromosomes, and the parameter value corresponding to each regulatory relationship in the set of regulatory relationships characterizes the regulatory strength of the regulatory element on the expression of the corresponding gene.

[0043] Based on the set of regulatory relationships and the single-cell chromatin accessibility sequencing data, the expression level prediction results of multiple genes are obtained, and based on the set of regulatory relationships and the single-cell expression level sequencing data, the accessibility prediction results of multiple regulatory elements are obtained.

[0044] Based on the expression level prediction results, the single-cell expression level sequencing data, the accessibility prediction results, and the single-cell chromatin accessibility sequencing data, the set of regulatory relationships is adjusted to obtain an adjusted set of regulatory relationships.

[0045] Based on the adjusted set of control relationships, the identification result of the control element is determined.

[0046] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0047] Single-cell chromatin accessibility sequencing data and single-cell expression level sequencing data are obtained to obtain a set of regulatory relationships to be adjusted; the single-cell chromatin accessibility sequencing data includes sequencing data reflecting the chromatin accessibility of multiple regulatory elements, the single-cell expression level sequencing data includes sequencing data reflecting the expression of multiple genes on chromosomes, and the parameter value corresponding to each regulatory relationship in the set of regulatory relationships characterizes the regulatory strength of the regulatory element on the expression of the corresponding gene.

[0048] Based on the set of regulatory relationships and the single-cell chromatin accessibility sequencing data, the expression level prediction results of multiple genes are obtained, and based on the set of regulatory relationships and the single-cell expression level sequencing data, the accessibility prediction results of multiple regulatory elements are obtained.

[0049] Based on the expression level prediction results, the single-cell expression level sequencing data, the accessibility prediction results, and the single-cell chromatin accessibility sequencing data, the set of regulatory relationships is adjusted to obtain an adjusted set of regulatory relationships.

[0050] Based on the adjusted set of control relationships, the identification result of the control element is determined.

[0051] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0052] Single-cell chromatin accessibility sequencing data and single-cell expression level sequencing data are obtained to obtain a set of regulatory relationships to be adjusted; the single-cell chromatin accessibility sequencing data includes sequencing data reflecting the chromatin accessibility of multiple regulatory elements, the single-cell expression level sequencing data includes sequencing data reflecting the expression of multiple genes on chromosomes, and the parameter value corresponding to each regulatory relationship in the set of regulatory relationships characterizes the regulatory strength of the regulatory element on the expression of the corresponding gene.

[0053] Based on the set of regulatory relationships and the single-cell chromatin accessibility sequencing data, the expression level prediction results of multiple genes are obtained, and based on the set of regulatory relationships and the single-cell expression level sequencing data, the accessibility prediction results of multiple regulatory elements are obtained.

[0054] Based on the expression level prediction results, the single-cell expression level sequencing data, the accessibility prediction results, and the single-cell chromatin accessibility sequencing data, the set of regulatory relationships is adjusted to obtain an adjusted set of regulatory relationships.

[0055] Based on the adjusted set of control relationships, the identification result of the control element is determined.

[0056] The aforementioned regulatory element identification method, apparatus, computer equipment, computer-readable storage medium, and computer program product acquire single-cell chromatin accessibility sequencing data and single-cell expression level sequencing data, acquire a set of regulatory relationships to be adjusted, wherein the single-cell chromatin accessibility sequencing data includes sequencing data reflecting the chromatin accessibility of multiple regulatory elements, and the single-cell expression level sequencing data includes sequencing data reflecting the expression status of multiple genes on chromosomes, wherein the parameter value corresponding to each regulatory relationship in the regulatory relationship set characterizes the regulatory strength of the regulatory element on the expression of the corresponding gene, and, based on the regulatory relationship set and the single-cell chromatin accessibility sequencing data, acquire the expression level prediction results of multiple genes, and, based on the regulatory relationship set and the single-cell expression level sequencing data, acquire the accessibility prediction results of multiple regulatory elements, adjust the regulatory relationship set based on the expression level prediction results, the single-cell expression level sequencing data, the accessibility prediction results, and the single-cell chromatin accessibility sequencing data to obtain an adjusted regulatory relationship set, and determine the regulatory element identification result based on the adjusted regulatory relationship set. Compared to related technologies in this field, this application characterizes the regulatory strength of regulatory elements on the expression of corresponding genes by using the parameter values ​​corresponding to each regulatory relationship in the regulatory relationship set. Based on the regulatory relationship set, sequencing data reflecting the chromatin accessibility of multiple regulatory elements, and sequencing data reflecting the expression of multiple genes on chromosomes, the application determines the predicted expression levels of multiple genes and the predicted accessibility of multiple regulatory elements. This allows for the capture of the synergistic or antagonistic regulatory effects of multiple regulatory elements on multiple genes in complex "many-to-many" regulatory scenarios where "one gene may be regulated by multiple regulatory elements, and one regulatory element may also regulate multiple genes." It accurately uncovers the complex regulatory relationships between regulatory elements and genes. Furthermore, this application integrates all genes and regulatory elements at the chromosome level, effectively identifying and incorporating remote regulatory elements, and fully covering long-distance regulatory pathways across the genome, thereby improving the accuracy of regulatory element identification. Attached Figure Description

[0057] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0058] Figure 1 This is a flowchart illustrating a method for identifying a control element in one embodiment;

[0059] Figure 2 This is a flowchart illustrating another method for identifying control elements in one embodiment;

[0060] Figure 3 This is a schematic diagram of a verification result in one embodiment;

[0061] Figure 4 This is a schematic diagram illustrating another verification result in one embodiment;

[0062] Figure 5 This is a schematic diagram of a third verification result in one embodiment;

[0063] Figure 6 This is a schematic diagram of the fourth verification result in one embodiment;

[0064] Figure 7 This is a schematic diagram of the fifth verification result in one embodiment;

[0065] Figure 8 This is a structural block diagram of the control element identification device in one embodiment;

[0066] Figure 9 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0067] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0068] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various objects, but these objects are not limited by these terms. These terms are only used to distinguish the first object from the second object. The term "comprising" and any variations thereof, as used in this application, are intended to cover non-exclusive inclusion. The term "multiple" as used in this application refers to two or more.

[0069] To enable those skilled in the art to better understand this application, the relevant technologies are first introduced below.

[0070] Regulatory elements are sequence segments in the genome that do not directly encode proteins or RNA, but can regulate the spatiotemporal patterns and intensity of gene expression through interactions with specific molecules. Cis-regulatory elements (CREs), as core components of the gene expression regulatory network, can precisely regulate the spatiotemporal expression patterns of genes by specifically binding to transcription factors and other regulatory proteins. Genetic variations associated with common and complex diseases are significantly enriched in non-coding regions, particularly distal CREs (such as enhancers). This breakthrough discovery expands the focus of disease mechanism research from protein-coding regions to non-coding regulatory elements, marking a major paradigm shift. Systematic and in-depth analysis of regulatory elements holds promise for elucidating the core gene regulatory networks driving developmental processes, cell differentiation, and species' environmental adaptation.

[0071] However, in the face of the massive number of regulatory elements in the genome (such as more than a million CREs), identifying regulatory elements with pathological significance still faces the following dual challenges: (1) Traditional population-level sequencing technology is difficult to resolve the functional heterogeneity of CREs at the single-cell level; (2) CREs have cell type-specific activation characteristics, and current studies on their functional annotation only cover a limited number of tissue types.

[0072] In related technologies, the identification of CREs is achieved by introducing technologies such as Signac, ArchR, SnapATAC (Single Nucleus Analysis Pipeline for ATAC, a single-cell ATAC data analysis tool), SCARlink (Single-Cell ATAC and RNALink, a combined analysis of single-cell ATAC and RNA), and DIRECT-NET.

[0073] Signac and ArchR are currently the mainstream technologies for mining CREs based on multi-omics data. Specifically, Signac identifies potential regulatory CREs by integrating gene expression levels and the chromatin openness of neighboring CREs. ArchR uses cell type information to achieve more refined CRE annotation, and then uses the MetaCells strategy to improve the stability of gene expression levels and chromatin openness signals, thereby enhancing the reliability of regulatory association inferences. However, Signac and ArchR mainly rely on correlation analysis between genes and CREs, limited to a simple "one-to-one" association model, which is difficult to fully characterize the essence of regulatory relationships. Considering that there are generally complex "many-to-many" regulatory networks between CREs and target genes, that is, the expression of the same gene can be regulated by multiple regulatory elements and the expression of multiple genes can be regulated by the same regulatory element, this single-dimensional analysis method still has certain limitations in revealing the overall dynamic characteristics of the regulatory system and its potential action patterns.

[0074] Furthermore, unlike the aforementioned similarity-based methods, SnapATAC employs a predictive modeling strategy to identify CREs with regulatory functions. This method integrates the chromatin accessibility features of multiple CREs within a gene's neighborhood and uses a logistic regression model to quantitatively predict gene expression levels. SCARlink introduces a regularized Poisson regression model, which significantly improves the robustness and interpretability of multivariate association analysis; while DIRECT-NET constructs a multivariate regression framework to systematically evaluate the contribution of each CRE to the expression of the target gene. These prediction-based models demonstrate good performance on specific datasets, effectively enhancing the accuracy of regulatory relationship inference. However, methods such as SnapATAC, SCARlink, and DIRECT-NET still primarily focus on gene neighborhood regions (such as promoters or proximal enhancers), failing to fully incorporate distant regulatory elements (such as distant enhancers or long-range interactions within topologically associated domains), making it difficult to comprehensively reveal long-distance regulatory networks across the genome and limiting the systematic analysis of complex gene regulatory mechanisms.

[0075] It is evident that the relevant technologies have difficulty accurately identifying control elements in long-distance and complex control networks.

[0076] Therefore, it is necessary to provide a method, apparatus, computer device, computer-readable storage medium, and computer program product for identifying control elements in response to the above-mentioned technical problems.

[0077] In one embodiment, such as Figure 1 As shown, a method for identifying control elements is provided. This embodiment illustrates the application of this method to a server. It is understood that this method can also be applied to a terminal, and further to a system including both a terminal and a server, and is implemented through interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0078] S101: Obtain single-cell chromatin accessibility sequencing data and single-cell expression level sequencing data to obtain the set of regulatory relationships to be adjusted; single-cell chromatin accessibility sequencing data includes sequencing data reflecting the chromatin accessibility of multiple regulatory elements; single-cell expression level sequencing data includes sequencing data reflecting the expression of multiple genes on chromosomes; parameter values ​​corresponding to each regulatory relationship in the set of regulatory relationships characterize the regulatory strength of the regulatory elements on the expression of the corresponding genes.

[0079] Regulatory elements (CREs) are specific DNA segments in the genome that participate in regulating gene expression activity. They do not directly encode proteins but rather regulate the spatiotemporal patterns and intensity of gene expression through interactions with specific molecules. CREs can include those located within the gene's vicinity and those located distally, as well as those on different chromosomes. In some examples, CREs within the gene's vicinity may include gene promoter regions (CREs); distal CREs are those located far from the coding region of the target gene. These distal CREs achieve physical proximity to the gene's promoter region through folding and interactions of chromatin higher-order structures, thereby regulating gene expression.

[0080] Regulatory relationships can characterize the interaction between regulatory elements (such as enhancers and promoters) and genes. Each regulatory relationship is characterized by specific parameters to represent the regulatory strength of the regulatory element on the expression of the corresponding gene. The set of regulatory relationships to be adjusted is a dataset containing multiple regulatory relationships. Its initial regulatory strength can be randomly generated and then optimized by combining accessibility and expression data.

[0081] Chromatin accessibility reflects whether chromatin regions are in a loose state. Loose chromatin allows regulatory elements to bind to transcription factors, thereby affecting gene expression. Single-cell chromatin accessibility sequencing data can be datasets obtained through specific sequencing technologies, reflecting the degree of chromatin accessibility or chromatin openness of multiple regulatory elements. Higher data values ​​generally indicate stronger chromatin accessibility of the corresponding regulatory element. In some embodiments, single-cell chromatin accessibility sequencing data can be obtained through experimental sequencing or public databases. After relevant preprocessing, final data containing multiple regulatory elements and their corresponding accessibility values ​​can be obtained.

[0082] Gene expression can reflect the amount of mRNA transcribed from a gene in a specific biological sample, reflecting the activity level of the gene; single-cell expression sequencing data is a dataset obtained through sequencing technology that reflects the expression levels of multiple genes on a chromosome.

[0083] For example, after identifying the target, such as the tissue structure that needs to be identified as a regulatory element, the scRNA-seq (i.e., single-cell expression sequencing data) and scATAC-seq (i.e., single-cell chromatin accessibility sequencing data) of the target are obtained. Based on the scRNA-seq and scATAC-seq of the target, a connection matrix that can characterize the regulatory relationship (i.e., a set of regulatory relationships containing multiple regulatory relationships) is obtained.

[0084] S102, based on the regulatory relationship set and single-cell chromatin accessibility sequencing data, obtains the expression level prediction results of multiple genes, and based on the regulatory relationship set and single-cell expression level sequencing data, obtains the accessibility prediction results of multiple regulatory elements.

[0085] The expression level prediction result can be a numerical value reflecting the gene expression level, calculated using a pre-set model or algorithm based on input data (including the set of regulatory relationships and single-cell chromatin accessibility sequencing data). The accessibility prediction result refers to a numerical value reflecting the degree of chromatin accessibility of regulatory elements, calculated using a pre-set model or algorithm based on input data (including the set of regulatory relationships and single-cell expression level sequencing data).

[0086] For example, to facilitate data analysis, the following variables are defined based on the regulatory relationship set, single-cell chromatin accessibility sequencing data, and single-cell expression level sequencing data: Represents n cells on chromosome θ scRNA-seq data for each gene; Represents n cells on the same chromosome θ scATAC-seq data for 1 CRE; It refers to the cis-regulatory linkage matrix between genes on chromosome θ and CREs, used to characterize the regulatory strength of CREs on the expression of the corresponding genes.

[0087] In some embodiments, for chromosome θ, by and This method predicts the expression levels of multiple genes on chromosome θ in corresponding cells, obtaining the predicted expression levels of multiple genes on chromosome θ. Through... and The chromatin accessibility of multiple regulatory elements on chromosome θ is predicted in the corresponding cells, and the accessibility prediction results of multiple regulatory elements on chromosome θ are obtained.

[0088] S103. Based on the expression level prediction results, single-cell expression level sequencing data, accessibility prediction results, and single-cell chromatin accessibility sequencing data, the set of regulatory relationships is adjusted to obtain the adjusted set of regulatory relationships.

[0089] In the specific implementation, we will take chromosome θ as an example for explanation. Characterizes the actual expression levels of multiple genes on chromosome θ. Characterizing the actual chromatin accessibility of multiple genes on chromosome θ This step characterizes the actual regulatory relationships between genes and regulatory elements on chromosome θ. For ease of description, the predicted expression levels of multiple genes on chromosome θ are referred to as... The prediction results of the accessibility of multiple regulatory elements on chromosome θ are called .

[0090] To optimize the set of regulatory relationships, ensuring that these relationships closely reflect reality and avoiding the impact of a few extreme values ​​on the overall optimization effect, we can base our approach on " close to , close to , The optimization objective is to minimize the weighted regularization loss value. The set of regulatory relationships is then adjusted to obtain the adjusted set of regulatory relationships.

[0091] S104. Based on the adjusted set of control relationships, determine the identification result of the control element.

[0092] In practical applications, based on the adjusted set of regulatory relationships, the significance of each regulatory relationship in the set is evaluated among all regulatory elements associated with its corresponding gene, yielding a first significance evaluation result for each regulatory relationship. The significance of each regulatory relationship in the set is then evaluated among all genes associated with its corresponding regulatory element, yielding a second significance evaluation result for each regulatory relationship. For each gene in the adjusted set of regulatory relationships, the first and second significance evaluation results for each regulatory relationship are corrected, resulting in corrected first and second significance evaluation results for each regulatory relationship. These corrected first and second significance evaluation results are then combined to obtain a comprehensive significance evaluation result for each regulatory relationship. Based on the comprehensive significance evaluation result for each regulatory relationship, a target regulatory element is selected from all regulatory elements associated with that gene. The regulatory relationship pair consisting of the target regulatory element and the gene is then used as the regulatory element identification result for that gene.

[0093] In the above-mentioned method for identifying regulatory elements, single-cell chromatin accessibility sequencing data and single-cell expression level sequencing data are obtained to acquire a set of regulatory relationships to be adjusted. The single-cell chromatin accessibility sequencing data includes sequencing data reflecting the chromatin accessibility of multiple regulatory elements, and the single-cell expression level sequencing data includes sequencing data reflecting the expression status of multiple genes on the chromosome. The parameter value corresponding to each regulatory relationship in the set of regulatory relationships characterizes the regulatory strength of the regulatory element on the expression of the corresponding gene. Based on the set of regulatory relationships and the single-cell chromatin accessibility sequencing data, the expression level prediction results of multiple genes are obtained. Also, based on the set of regulatory relationships and the single-cell expression level sequencing data, the accessibility prediction results of multiple regulatory elements are obtained. Based on the expression level prediction results, single-cell expression level sequencing data, accessibility prediction results, and single-cell chromatin accessibility sequencing data, the set of regulatory relationships is adjusted to obtain an adjusted set of regulatory relationships. Based on the adjusted set of regulatory relationships, the regulatory element identification result is determined. Compared to related technologies in this field, this embodiment characterizes the regulatory strength of regulatory elements on the expression of corresponding genes by using the parameter values ​​corresponding to each regulatory relationship in the regulatory relationship set. Based on the regulatory relationship set, sequencing data reflecting the chromatin accessibility of multiple regulatory elements, and sequencing data reflecting the expression of multiple genes on chromosomes, the expression level prediction results of multiple genes and the accessibility prediction results of multiple regulatory elements are determined. This can capture the synergistic or antagonistic regulatory effects of multiple regulatory elements on multiple genes in the complex "many-to-many" regulatory scenario where "one gene may be regulated by multiple regulatory elements, and one regulatory element may also regulate multiple genes". It can accurately explore the complex regulatory relationship between regulatory elements and genes. Moreover, this embodiment can effectively identify and include remote regulatory elements through global correlation analysis of single-cell chromatin accessibility sequencing data and single-cell expression level sequencing data, fully covering long-distance regulatory pathways across the genome, thereby improving the accuracy of regulatory element identification.

[0094] In an exemplary embodiment, in step S103, the regulatory relationship set is adjusted based on the expression level prediction results, single-cell expression level sequencing data, accessibility prediction results, and single-cell chromatin accessibility sequencing data to obtain an adjusted regulatory relationship set, which may include the following steps:

[0095] A first loss value is determined based on the difference in gene expression levels between the predicted expression levels and single-cell expression sequencing data; a second loss value is determined based on the difference in accessibility between the predicted accessibility levels and single-cell chromatin accessibility sequencing data; a weighted regularization loss value is determined based on the parameter values ​​corresponding to each regulatory relationship in the regulatory relationship set; a target loss value is determined based on the first loss value, the second loss value, and the weighted regularization loss value; and the regulatory relationship set is adjusted based on the target loss value.

[0096] Specifically, taking chromosome θ as an example, the expression level prediction results for chromosome θ include the expression level prediction results of multiple genes on chromosome θ, and the accessibility prediction results for chromosome θ include the accessibility prediction results of multiple CREs on chromosome θ. In this step, for ease of description, the single-cell expression level sequencing data of chromosome θ will be referred to as... The predicted expression level of chromosome θ is called The first loss value of chromosome θ is called The single-cell chromatin accessibility sequencing data of chromosome θ is called... The accessibility prediction results of chromosome θ are called The second loss value of chromosome θ is called .

[0097] In some exemplary embodiments, see Equation (1) for analyzing expression level prediction results. and single-cell expression level sequencing data The difference in gene expression levels between them yields the first loss value. See Equation (2) for analysis of accessibility prediction results. Single-cell chromatin accessibility sequencing data The accessibility difference between them yields the second loss value. See equation (3) for analysis of the connection matrix. L2 regularization loss (That is, the weighted regularization loss value); see Equation (4), determine the target loss value based on the first loss value, the second loss value and the weighted regularization loss value, adjust the control relationship set with the goal of minimizing the target loss value and update the target loss value based on the adjusted control relationship set.

[0098] (1)

[0099] (2)

[0100] (3)

[0101] (4)

[0102] in, The expression level of the i-th gene on chromosome θ in cell k is obtained from single-cell expression sequencing data of chromosome θ. The expression level of the i-th gene on chromosome θ in cell k is obtained from the expression level prediction results of chromosome θ. The first chromatin accessibility of the j-th CRE on chromosome θ in cell k is obtained from single-cell chromatin accessibility sequencing data of chromosome θ. The accessibility of the j-th CRE on chromosome θ to the second chromatin in cell k is obtained from the accessibility prediction results of chromosome θ. The regulatory strength between the i-th gene and the j-th CRE on chromosome θ; For hyperparameters; In some embodiments, to prevent gradient vanishing due to an excessive number of parameters during the minimization of the target loss value, the weight matrix can be initialized using He, i.e. p represents the number of CREs; the L-BFGS algorithm is used to minimize the loss function, with epochs set to 500 and a learning rate of 0.5.

[0103] In this embodiment, the first loss value, determined based on the difference in gene expression levels between the expression level prediction results and single-cell expression level sequencing data, can intuitively measure the degree of deviation of the current regulatory relationship set in predicting gene expression levels. The second loss value, determined based on the difference in accessibility between the accessibility prediction results and single-cell chromatin accessibility sequencing data, can intuitively reflect the degree of deviation of the current regulatory relationship set in predicting the accessibility of regulatory elements. By adjusting the regulatory relationship set based on the target loss value determined by the first and second loss values, it is ensured that the adjustment of the regulatory relationship set simultaneously fits both biological evidence of gene expression and chromatin accessibility, thereby more accurately identifying regulatory elements and their regulatory strength.

[0104] In one exemplary embodiment, the regulatory element includes a cis-regulatory element, and the single-cell expression sequencing data includes sequencing data reflecting the expression of multiple genes on the same chromosome; differences in gene expression levels are determined through the following steps:

[0105] Based on single-cell expression sequencing data, the first expression level of multiple genes on the same chromosome is determined, and the second expression level of multiple genes on the same chromosome is determined based on the expression level prediction results; for each gene on the same chromosome in the same cell, the first difference between the first expression level and the second expression level is obtained; based on each first difference, the gene expression level difference is determined.

[0106] Among them, the first expression level of multiple genes on the same chromosome refers to the actual expression level of multiple genes on the same chromosome; the second expression level of multiple genes on the same chromosome refers to the predicted expression level of multiple genes on the same chromosome; and the first difference between the first expression level and the second expression level is used to reflect the deviation between the first expression level and the second expression level.

[0107] In practical applications, for the i-th gene on chromosome θ, the first difference in expression level of the i-th gene on chromosome θ in cell k can be determined based on the difference between the first expression level of the i-th gene on chromosome θ in cell k and the second expression level of the i-th gene on chromosome θ in cell k. Furthermore, based on the difference in expression level of the i-th gene on chromosome θ... The first difference in gene expression levels across n cells is used to determine the gene expression differences on chromosome θ.

[0108] In this embodiment, since cis-regulatory elements (such as enhancers and promoters) typically only affect genes on the same chromosome as the cis-regulatory element and do not function across chromosomes, focusing on the first difference between the first expression level (measured expression level) and the second expression level (predicted expression level) of multiple genes on the same chromosome can more accurately reflect the actual regulatory effect of cis-regulatory elements on gene expression. Furthermore, by analyzing the expression level differences of genes on the same chromosome, the influence of the local chromosomal environment on gene expression can be better captured. Compared to overall gene expression level comparisons that do not distinguish chromosomes, this embodiment also determines the first difference between the first expression level and the second expression level for each gene on the same chromosome in the same cell. The gene expression level difference is determined based on the first difference corresponding to each gene. This gene-by-gene comparison method can avoid interference caused by differences in gene expression characteristics on different chromosomes, thereby allowing for a more detailed analysis of the deviation between the actual and predicted gene expression.

[0109] In one exemplary embodiment, the regulatory element includes a cis-regulatory element, and the single-cell chromatin accessibility sequencing data includes sequencing data reflecting the chromatin accessibility of each of the multiple cis-regulatory elements on the same chromosome; accessibility differences are determined by the following steps:

[0110] Based on single-cell chromatin accessibility sequencing data, the first chromatin accessibility of each of the multiple cis-regulatory elements on the same chromosome is determined, and based on the accessibility prediction results, the second chromatin accessibility of each of the multiple cis-regulatory elements on the same chromosome is determined; for each cis-regulatory element on the same chromosome in the same cell, a second difference between the first chromatin accessibility and the second chromatin accessibility is determined; based on each second difference, accessibility differences are determined.

[0111] The first chromatin accessibility of multiple cis-regulatory elements on the same chromosome refers to the actual chromatin accessibility of each of the multiple cis-regulatory elements on the same chromosome; the second chromatin accessibility of multiple cis-regulatory elements on the same chromosome refers to the predicted chromatin accessibility of each of the multiple cis-regulatory elements on the same chromosome; and the second difference between the first chromatin accessibility and the second chromatin accessibility is used to reflect the deviation between the first chromatin accessibility and the second chromatin accessibility.

[0112] In specific implementation, for the j-th cis-regulatory element on chromosome θ, the second difference corresponding to the chromatin accessibility of the j-th cis-regulatory element on chromosome θ in cell k can be determined based on the difference between the first chromatin accessibility of the j-th cis-regulatory element on chromosome θ in cell k and the second chromatin accessibility of the j-th cis-regulatory element on chromosome θ in cell k. Furthermore, based on the difference between the first chromatin accessibility of the j-th cis-regulatory element on chromosome θ and the second chromatin accessibility of the j-th cis-regulatory element on chromosome θ in cell k, the second difference can be determined. The second difference in chromatin accessibility of each cis-regulatory element in n cells determines the accessibility difference of chromosome θ.

[0113] In this embodiment, since cis-regulatory elements (such as enhancers and promoters) typically only affect genes on the same chromosome as the cis-regulatory element and do not function across chromosomes, focusing on the second difference between the first chromatin accessibility (measured chromatin accessibility) and the second chromatin accessibility (predicted chromatin accessibility) of multiple cis-regulatory elements on the same chromosome can more accurately reflect the interaction between cis-regulatory elements and genes. Furthermore, by analyzing the accessibility differences of cis-regulatory elements on the same chromosome, the influence of the local chromosomal environment on the accessibility of regulatory elements can be better captured. Compared to a comprehensive accessibility comparison that does not distinguish between chromosomes and regulatory element types, this embodiment also determines the second difference between the first chromatin accessibility and the second chromatin accessibility for each cis-regulatory element on the same chromosome in the same cell. The accessibility difference is determined based on the second difference corresponding to each cis-regulatory element. This method of comparing cis-regulatory elements one by one can avoid interference from accessibility differences between different chromosomes and different types of regulatory elements, thereby allowing for a more detailed analysis of the deviation between the actual and predicted accessibility of regulatory elements.

[0114] In an exemplary embodiment, step S102, obtaining the expression level prediction results of multiple genes based on the regulatory relationship set and single-cell chromatin accessibility sequencing data, may include the following steps:

[0115] For each gene on the same chromosome in the same cell, the regulatory strength of multiple regulatory elements on gene expression is determined based on the set of regulatory relationships. Furthermore, the first chromatin accessibility of multiple regulatory elements on the same chromosome in the same cell is determined based on single-cell chromatin accessibility sequencing data. Based on the regulatory strength of multiple regulatory elements on gene expression and the first chromatin accessibility of multiple regulatory elements on the same chromosome in the same cell, the predicted regulatory contribution of each regulatory element to the gene is determined. The predicted regulatory contribution indicates the likelihood that a regulatory element will regulate gene expression. Based on the predicted regulatory contribution of each regulatory element to the gene, the predicted gene expression level is determined.

[0116] Among them, the predicted contribution of regulatory elements to gene regulation can characterize the potential contribution of regulatory elements to genes, that is, whether the regulation of gene expression by CREs is "active". For example, a certain regulatory element has significant regulation of both genes A and B, but the predicted contribution of gene B to the regulatory element is significantly lower than that of gene A. Therefore, the regulatory element is more likely to regulate the expression of gene A.

[0117] Specifically, taking chromosome θ as an example, for the i-th gene on chromosome θ in cell k, based on the set of regulatory relationships, the regulatory strength of multiple regulatory elements on chromosome θ on the expression of the i-th gene is determined. Furthermore, based on single-cell chromatin accessibility sequencing data, the first chromatin accessibility of multiple regulatory elements on chromosome θ in cell k is determined. Based on the regulatory strength of multiple regulatory elements on chromosome θ on the expression of the i-th gene and the first chromatin accessibility of multiple regulatory elements on chromosome θ in cell k, the predicted regulatory contribution of each regulatory element to the i-th gene is determined. ;j is the sequence number of the regulatory element on chromosome θ; based on the predicted contribution of the p regulatory elements on chromosome θ to the regulation of the i-th gene, the predicted expression level of the i-th gene is determined. .

[0118] In this embodiment, by comprehensively considering the regulatory strength of multiple regulatory elements on the same chromosome on gene expression and the first chromatin accessibility of multiple regulatory elements on the same chromosome in the same cell, it is possible to avoid erroneously identifying regulatory elements with low regulatory strength or low chromatin accessibility for genes. This avoids the error caused by applying a static regulatory relationship network to all cell types, thereby more accurately analyzing the predicted value of the regulatory contribution of each regulatory element to the gene.

[0119] In an exemplary embodiment, step S102, obtaining accessibility prediction results for multiple regulatory elements based on the regulatory relationship set and single-cell expression level sequencing data, may include the following steps:

[0120] For each regulatory element on the same chromosome in the same cell, the regulatory strength of the regulatory element on the expression of multiple genes on the same chromosome is determined based on the set of regulatory relationships. Furthermore, the first expression level of each of the multiple genes on the same chromosome is determined based on single-cell expression sequencing data. Based on the regulatory strength of the regulatory element on the expression of multiple genes on the same chromosome and the first expression level of each of the multiple genes on the same chromosome, the predicted regulatory association value of the regulatory element on each gene is determined. The predicted regulatory association value is used to indicate the association between the regulatory element and the regulation of gene expression. Based on the predicted regulatory association value of the regulatory element on each gene, the accessibility prediction result of the regulatory element is determined.

[0121] Among them, the predictive value of the regulatory element's association with gene regulation can characterize the relationship between the regulatory element and the regulation of gene expression. For example, a very low predictive value of the regulatory element's association with gene regulation may indicate that the regulatory element is not the driver of the regulatory operation of gene expression.

[0122] Specifically, taking chromosome θ as an example, for the j-th regulatory element on chromosome θ in cell k, based on the set of regulatory relationships, the regulatory strength of the j-th regulatory element on chromosome θ for the expression of multiple genes is determined. Furthermore, based on single-cell expression sequencing data, the first expression level of multiple genes on chromosome θ in cell k is determined. Finally, based on the regulatory strength of multiple regulatory elements on chromosome θ for the expression of multiple genes and the first expression level of multiple genes on chromosome θ in cell k, the predicted regulatory association values ​​of the j-th regulatory element for each gene are determined. ; i is the gene number on chromosome θ; based on the predicted contribution of the j-th regulatory element on chromosome θ to the regulation of g genes, the accessibility prediction result of the j-th regulatory element is determined. .

[0123] In this embodiment, by comprehensively considering the regulatory strength of regulatory elements on the expression of multiple genes on the same chromosome and the first expression level of each of the multiple genes on the same chromosome, it is possible to avoid erroneously identifying regulatory elements with low regulatory strength or low expression levels of the regulated genes, thereby discovering potentially active regulatory elements. This allows for a more accurate analysis of the regulatory association prediction values ​​of each regulatory element on the genes, and based on this discovery, it is possible to verify whether there are false positives in the regulatory relationship and to verify the functionality of the regulatory relationship.

[0124] In an exemplary embodiment, step S104, determining the identification result of the control element based on the adjusted control relationship set, may include the following steps:

[0125] Based on the adjusted set of regulatory relationships, multiple regulatory relationship pairs are identified. These pairs are used to associate genes with regulatory elements that have a regulatory relationship with the genes. The significance assessment results of the regulatory relationship pairs are determined based on the first and second significance probabilities. The first significance probability reflects the significance of the gene among the various regulatory elements that have a regulatory relationship with the gene, and the second significance probability reflects the significance of the regulatory element among the various genes that have a regulatory relationship with the regulatory element. For each gene in the adjusted set of regulatory relationships, the target regulatory element is selected from the various regulatory elements that have a regulatory relationship with the gene based on the significance assessment results of each regulatory relationship pair, thus obtaining the gene's regulatory element identification results.

[0126] For example, determining the significance evaluation result of the regulatory relationship pair based on the first significance probability and the second significance probability of the regulatory relationship pair may include the following steps: correcting the first significance probability and the second significance probability of the regulatory relationship pair to obtain corrected first significance probability and corrected second significance probability; merging the corrected first significance probability and the corrected second significance probability to obtain the significance evaluation result of the regulatory relationship pair.

[0127] Specifically, since a regulatory relationship involves two elements, a gene and a regulatory element, this embodiment evaluates its statistical significance from two dimensions: (1) in the context of the target gene: assessing whether the regulatory strength of the regulatory element significantly deviates from the background distribution among all regulatory elements used to regulate the gene; (2) in the context of the target regulatory element: assessing whether the regulatory strength of the gene significantly deviates from the background distribution among all genes acted upon by the regulatory element. Based on this, the association strength between the target gene context and the target regulatory element context is transformed by Z-score. Assuming that the regulatory weights approximately follow a normal distribution, the p-value is calculated through the Z-distribution to obtain two sets of p-values, namely the first significance probability and the second significance probability. Since the calculation here involves a large number of genes and regulatory elements, in order to control the false positive risk caused by multiple testing, this embodiment performs Bonferroni correction on the two sets of p-values. After the above processing, each regulatory relationship pair will have two p-values. Then, the Pearson merging method in the SciPy package is used to integrate the two corrected p-values ​​into a comprehensive significance score P (that is, the significance evaluation result of the regulatory relationship pair). Finally, for each gene, regulatory pairs with a P-value less than 10^-7 are screened from all regulatory pairs associated with that gene; among them, the regulatory elements associated with each regulatory pair that satisfies a P-value less than 10^-7 are the target regulatory elements.

[0128] In this embodiment, by introducing a fine screening framework with dual significance tests centered on genes and regulatory elements, highly specific and reliable regulatory element-gene pairs can be extracted from the adjusted set of regulatory relationships. This effectively solves the ambiguity problem in complex regulatory networks and ensures the accuracy, reliability, and direct usability of the regulatory element identification results.

[0129] To enable those skilled in the art to better understand the above steps, an example is provided below to illustrate the embodiments of this application. In this example, a cis-regulatory element is used as the regulating element. However, it should be understood that the embodiments of this application are not limited thereto.

[0130] like Figure 2 As shown, this example includes steps S201 to S215.

[0131] S201: Obtain single-cell chromatin accessibility sequencing data and single-cell expression level sequencing data to obtain the set of regulatory relationships to be adjusted.

[0132] Among them, single-cell chromatin accessibility sequencing data includes sequencing data reflecting the chromatin accessibility of multiple cis-regulatory elements; single-cell expression level sequencing data includes sequencing data reflecting the expression of multiple genes on chromosomes; and parameter values ​​corresponding to each regulatory relationship in the set of regulatory relationships are used to characterize the regulatory strength of cis-regulatory elements on the expression of the corresponding genes.

[0133] S202, for each gene on the same chromosome in the same cell, determines the regulatory strength of multiple cis-regulatory elements on gene expression based on the set of regulatory relationships, and determines the first chromatin accessibility of multiple cis-regulatory elements on the same chromosome in the same cell based on single-cell chromatin accessibility sequencing data.

[0134] S203, based on the regulatory strength of multiple cis-regulatory elements on gene expression on the same chromosome and the accessibility of multiple cis-regulatory elements on the first chromatin of the same chromosome in the same cell, determines the predicted contribution of each cis-regulatory element to gene regulation.

[0135] Among them, the regulatory contribution prediction value is used to indicate the likelihood that cis-regulatory elements will regulate gene expression.

[0136] S204 determines the predicted gene expression level based on the predicted contribution of each cis-regulatory element to gene regulation.

[0137] S205, based on single-cell expression sequencing data, determines the first expression level of multiple genes on the same chromosome, and based on expression level prediction results, determines the second expression level of multiple genes on the same chromosome.

[0138] S206, for each gene on the same chromosome in the same cell, obtain the first difference between the first expression level and the second expression level; determine the gene expression level difference based on each first difference; determine the first loss value based on the gene expression level difference between the expression level prediction result and the single-cell expression level sequencing data.

[0139] S207, for each cis-regulatory element on the same chromosome in the same cell, determines the regulatory strength of the cis-regulatory element on the expression of multiple genes on the same chromosome based on the set of regulatory relationships, and determines the first expression level of each of the multiple genes on the same chromosome based on single-cell expression sequencing data.

[0140] S208, based on the regulatory strength of cis-regulatory elements on the expression of multiple genes on the same chromosome and the first expression level of each of the multiple genes on the same chromosome, determines the predicted regulatory association value of cis-regulatory elements on each gene.

[0141] Among them, the regulatory association prediction value is used to indicate the association between cis-regulatory elements and the regulation of gene expression.

[0142] S209. Based on the predicted regulatory association values ​​of cis-regulatory elements with each gene, the accessibility prediction results of cis-regulatory elements are determined.

[0143] S210, based on single-cell chromatin accessibility sequencing data, determines the first chromatin accessibility of multiple cis-regulatory elements on the same chromosome, and based on accessibility prediction results, determines the second chromatin accessibility of multiple cis-regulatory elements on the same chromosome.

[0144] S211, for each cis-regulatory element on the same chromosome in the same cell, determine a second difference between the first chromatin accessibility and the second chromatin accessibility; determine the accessibility difference based on each second difference; determine a second loss value based on the accessibility difference between the accessibility prediction results and the single-cell chromatin accessibility sequencing data.

[0145] S212, determine the weighted regularization loss value based on the parameter value corresponding to each regulation relationship in the regulation relationship set.

[0146] S213, determine the target loss value based on the first loss value, the second loss value, and the weighted regularization loss value; adjust the set of control relationships based on the target loss value.

[0147] S214, Based on the adjusted set of regulatory relationships, determine multiple regulatory relationship pairs.

[0148] Among them, regulatory relationships are used to associate genes and cis-regulatory elements that have a regulatory relationship with genes.

[0149] S215. Determine the significance assessment result of the regulatory relationship pair based on the first and second significance probabilities of the regulatory relationship pair.

[0150] The first significance probability reflects the significance of a gene among all cis-regulatory elements that have a regulatory relationship with it, while the second significance probability reflects the significance of a cis-regulatory element among all genes that have a regulatory relationship with it. Specifically, the significance of the regulatory relationship among all CREs associated with the gene and among all genes associated with the CRE can be assessed using z-scores to obtain the first and second significance probabilities of the regulatory relationship pair.

[0151] S216: For each gene in the adjusted set of regulatory relationships, based on the significance evaluation results of each regulatory relationship pair, the target cis-regulatory element is selected from each cis-regulatory element that has a regulatory relationship with the gene, and the regulatory element identification result of the gene is obtained.

[0152] Compared to related technologies, this embodiment, through global association analysis of single-cell chromatin accessibility sequencing data and single-cell expression level sequencing data, can effectively identify and include distal cis-regulatory elements, fully covering long-distance regulatory pathways across the genome. Furthermore, by using the parameter values ​​corresponding to each regulatory relationship in the set of regulatory relationships, the regulatory strength of cis-regulatory elements on the expression of corresponding genes is characterized. Combining sequencing data reflecting the chromatin accessibility of multiple cis-regulatory elements and sequencing data reflecting the expression of multiple genes on chromosomes, the potential associations between all genes on each chromosome and cis-regulatory elements are systematically analyzed. A bidirectional regression model is used to construct a regulatory network between genes and cis-regulatory elements, and the weight distribution characteristics of each connection in the network are evaluated from a global perspective to identify significant regulatory relationships. This achieves precise analysis of the complex regulatory effects between cis-regulatory elements and target genes, thus overcoming the limitations of related technologies in terms of regulatory distance, topological structure, and many-to-many mapping relationships. Based on the framework provided in this application, the inventors conducted related experiments:

[0153] In Experiment 1, the inventors verified the prediction results of this application using an enhancer activity reporter system and a cis-regulatory element knockout system in pluripotent stem cells. Specifically, this included the following steps:

[0154] ① Based on the method provided in this application, CREs were predicted using multi-omics data from key stages of mouse gastrulation (E6.5–E7.5), with Mesp1, a core regulatory gene for mesodermal development, as an example. The results show that this technology not only successfully identified known CREs with regulatory functions reported in the literature (PMID (PubMed Unique Identifier): 11578861), but also predicted a novel candidate CRE named Neo_ME (e.g., Figure 3 (As shown).

[0155] ② The activity of Neo_ME in E6.5 and E7.5 mouse embryos was detected using an enhancer activity reporter system. The results showed that Neo_ME exhibited significant enhancer activity in mesodermal tissues. Simultaneously, qPCR analysis further confirmed that the spatial distribution of Mesp1 gene expression was highly consistent with the active region of Neo_ME (e.g., Figure 3 (As shown).

[0156] ③ In an in vitro pluripotent stem cell model, knockout of known CREs and Neo_ME, respectively, resulted in a significant downregulation of Mesp1 expression in both knockout groups compared to the control group. This functional experiment demonstrates that Neo_ME does indeed have a functional role in regulating Mesp1 expression (e.g., Figure 3 (As shown). The above multi-angle experimental evidence fully verifies the high accuracy and predictive ability of this application in identifying functional CREs.

[0157] In Experiment 2, the inventors validated the prediction results of this application using PCHiC data and multi-omics data from the GM12878 cell line. This specifically included the following steps:

[0158] ① For ease of description and comparison, the regulatory element identification method provided in this application is referred to as BioCRE in Experiments 2, 3, and 5. Based on BioCRE, CREs were predicted using two independent multi-omics replicates of the GM12878 cell line, and three replicates of PCHiC experimental data were used as a high-confidence physical interaction reference set to evaluate the accuracy of the predicted CREs by each technique. Compared with the current mainstream techniques Signac and ArchR, BioCRE achieved higher Macro-F1 scores on all prediction results (e.g., ...). Figure 4 (As shown).

[0159] ② To further refine the evaluation of the performance of different technologies on common target genes, the inventors analyzed the same set of genes predicted by each technology separately. The results showed that BioCRE still outperformed Signac and ArchR (e.g., in Macro-F1 scoring) in the Macro-F1 test. Figure 4 (As shown).

[0160] ③ Independent evaluation of the genes and their associated CREs identified by each technology revealed that BioCRE-specific predictions had a significantly higher True Positive Rate (TPR) (e.g., Figure 4 (As shown). In summary, the systematic validation based on high-throughput chromatin interaction data shows that the BioCRE of this application not only outperforms current mainstream technologies in overall prediction performance on cell lines, but also demonstrates significant advantages in identifying high-confidence, specific regulatory associations.

[0161] In Experiment 3, the inventors validated the prediction results of this application using PCHiC data and peripheral blood mononuclear cell (PBMC) multi-omics data. This specifically included the following steps:

[0162] ① Based on BioCRE, CRE predictions were performed on multi-omics repetitions of CD14 Monocytes, CD16 Monocytes, and CD4 T cells in PBMCs. PCHiC experimental data were used as a high-confidence chromatin spatial interaction reference set to evaluate the accuracy of the predicted regulatory associations. Compared to Signac and ArchR, BioCRE showed higher Macro-F1 scores in all prediction results (e.g., ...). Figure 5 (As shown).

[0163] ② To further refine the evaluation of the performance of different technologies on common target genes, the inventors analyzed the same set of genes predicted by each technology separately. The results showed that BioCRE still outperformed Signac and ArchR (e.g., in Macro-F1 scoring) in the Macro-F1 test. Figure 5 (As shown).

[0164] ③ Independent evaluation of the genes and their associated CREs identified by each technology-specific approach revealed that BioCRE-specific prediction results showed significantly higher TPRs (e.g., Figure 5 (As shown). In summary, the systematic validation based on high-throughput chromatin interaction data shows that the BioCRE of this application also has stability in the prediction results of PBMC.

[0165] In Experiment 4, the inventors validated the long-range CREs predicted in this application using multi-omics data from mouse gastrulation. The specific steps included:

[0166] ① Based on multi-omics data from the critical stages of mouse gastrulation (E6.5–E7.5) in this application, CRE prediction was performed, identifying a distal candidate CRE located more than 10 Mb upstream of the core mesoderm regulatory gene Mesp1 (chr7: 126, 784, 768–126, 785, 759). The chromatin opening dynamics in this region showed a high positive correlation with the gene expression level of Mesp1 (e.g., ...). Figure 6 (As shown).

[0167] ② Further integration of spatial transcriptome data (GEO-seq) from the E7.5 period revealed that the activity distribution pattern of these distant CREs exhibited highly consistent spatial co-localization characteristics with Mesp1 expression in the embryonic mesoderm region (e.g., Figure 6 (As shown).

[0168] ③ Integrated analysis using published ChIP-seq data of the key mesodermal transcription factor T (Brachyury) revealed a significantly enriched T protein binding peak in this distal CRE region, indicating that it possesses typical regulatory element characteristics and may be directly regulated by core developmental pathways (such as...). Figure 6 (As shown). This application can effectively identify long-distance functional regulatory elements that are easily missed by traditional techniques. The predicted distant CREs have multiple functional characteristics such as active regulatory state, spatial co-expression mode and binding to key transcription factors. They are very likely to be functional enhancers that are actually involved in the long-distance regulation of the Mesp1 gene. This fully demonstrates the unique advantages and biological discovery potential of this application in resolving complex gene regulatory topologies.

[0169] In Experiment 5, the inventors evaluated the stability and robustness of this application in predicting CREs at different sequencing depths through downsampling experiments. Specifically, the following steps were included:

[0170] ① Based on mouse gastrulation multi-omics data, the sequencing depth of single-cell transcriptome (scRNA-seq) and single-cell chromatin accessibility (scATAC-seq) data were simultaneously reduced at different ratios (20%, 40%, 60%, 80%, 100%) to evaluate the predictive consistency of each technology under sparse data conditions. The results showed that BioCRE exhibited significantly better stability than Signac and ArchR under all downsampling conditions, with a prediction overlap rate as high as 40.9% across different sequencing depths, while Signac and ArchR had overlap rates of 8.2% and 5.7%, respectively (e.g., ...). Figure 7 (As shown).

[0171] ② In experiments where scATAC-seq data was kept intact while scRNA-seq sequencing depth was gradually reduced, the prediction results from the three software programs all showed strong stability: BioCRE was 44.6%, Signac and ArchR were 42.9% and 61.9%, respectively (e.g., ...). Figure 7 (As shown).

[0172] ③ In tests where scRNA-seq data was fixed and scATAC-seq sequencing depth was gradually reduced, BioCRE showed the highest regulatory pair consistency rate (52.1%) among different technologies, significantly better than Signac (24.1%) and ArchR (46.2%). Figure 7 (As shown). In summary, multiple downsampling experiments have confirmed that the BioCRE in this application exhibits excellent predictive stability and robustness under different sequencing depths and data imbalance conditions. In particular, it can reliably identify functional gene-CRE associations even when chromatin accessibility data is limited, which is significantly better than existing mainstream technologies.

[0173] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0174] Based on the same inventive concept, this application also provides a control element identification device for implementing the control element identification method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more control element identification device embodiments provided below can be found in the limitations of the control element identification method described above, and will not be repeated here.

[0175] In one exemplary embodiment, such as Figure 8 As shown, a control element identification device is provided, comprising:

[0176] The first acquisition module 801 is used to acquire single-cell chromatin accessibility sequencing data and single-cell expression level sequencing data, and to acquire a set of regulatory relationships to be adjusted; the single-cell chromatin accessibility sequencing data includes sequencing data reflecting the chromatin accessibility of multiple regulatory elements, the single-cell expression level sequencing data includes sequencing data reflecting the expression of multiple genes on chromosomes, and the parameter value corresponding to each regulatory relationship in the set of regulatory relationships characterizes the regulatory strength of the regulatory element on the expression of the corresponding gene.

[0177] The second acquisition module 802 is used to acquire the expression level prediction results of multiple genes based on the regulatory relationship set and single-cell chromatin accessibility sequencing data, and to acquire the accessibility prediction results of multiple regulatory elements based on the regulatory relationship set and single-cell expression level sequencing data.

[0178] The adjustment module 803 is used to adjust the set of regulatory relationships based on the expression level prediction results, single-cell expression level sequencing data, accessibility prediction results, and single-cell chromatin accessibility sequencing data, so as to obtain the adjusted set of regulatory relationships.

[0179] The determination module 804 is used to determine the identification result of the control element based on the adjusted control relationship set.

[0180] In an exemplary embodiment, the adjustment module is further configured to: determine a first loss value based on the gene expression level difference between the expression level prediction result and the single-cell expression level sequencing data; determine a second loss value based on the accessibility difference between the accessibility prediction result and the single-cell chromatin accessibility sequencing data; determine a weighted regularization loss value based on the parameter value corresponding to each regulatory relationship in the regulatory relationship set; determine a target loss value based on the first loss value, the second loss value, and the weighted regularization loss value; and adjust the regulatory relationship set based on the target loss value.

[0181] In an exemplary embodiment, the regulatory element includes a cis-regulatory element, and the single-cell expression sequencing data includes sequencing data reflecting the expression of multiple genes on the same chromosome. The adjustment module 803 is further configured to determine, based on the single-cell expression sequencing data, a first expression level of each of the multiple genes on the same chromosome, and, based on expression level prediction results, a second expression level of each of the multiple genes on the same chromosome; for each gene on the same chromosome in the same cell, obtain a first difference between the first and second expression levels; and determine gene expression level differences based on each first difference.

[0182] In one exemplary embodiment, the regulatory element includes a cis-regulatory element, and the single-cell chromatin accessibility sequencing data includes sequencing data reflecting the chromatin accessibility of each of the plurality of cis-regulatory elements on the same chromosome. The adjustment module 803 is further configured to determine, based on the single-cell chromatin accessibility sequencing data, a first chromatin accessibility of each of the plurality of cis-regulatory elements on the same chromosome, and, based on accessibility prediction results, a second chromatin accessibility of each of the plurality of cis-regulatory elements on the same chromosome; determine a second difference between the first chromatin accessibility and the second chromatin accessibility for each cis-regulatory element on the same chromosome for the same cell; and determine an accessibility difference based on each second difference.

[0183] In an exemplary embodiment, the second acquisition module 802 is further configured to, for each gene on the same chromosome in the same cell, determine the regulatory strength of multiple regulatory elements on gene expression on the same chromosome based on a set of regulatory relationships, and determine the first chromatin accessibility of multiple regulatory elements on the same chromosome in the same cell based on single-cell chromatin accessibility sequencing data; determine the predicted regulatory contribution value of each regulatory element to the gene based on the regulatory strength of multiple regulatory elements on the gene expression and the first chromatin accessibility of multiple regulatory elements on the same chromosome in the same cell; the predicted regulatory contribution value is used to indicate the likelihood of the regulatory element regulating gene expression; and determine the predicted gene expression level based on the predicted regulatory contribution value of each regulatory element to the gene.

[0184] In an exemplary embodiment, the second acquisition module 802 is further configured to, for each regulatory element on the same chromosome of the same cell, determine the regulatory strength of the regulatory element on the expression of multiple genes on the same chromosome according to a set of regulatory relationships, and determine the first expression level of each of the multiple genes on the same chromosome according to single-cell expression sequencing data; determine the regulatory association prediction value of the regulatory element on each gene according to the regulatory strength of the regulatory element on the expression of multiple genes on the same chromosome and the first expression level of each of the multiple genes on the same chromosome; the regulatory association prediction value is used to indicate the association between the regulation of the regulatory element and the regulation of gene expression; and determine the accessibility prediction result of the regulatory element according to the regulatory association prediction value of the regulatory element on each gene.

[0185] In an exemplary embodiment, the determining module 803 is further configured to determine multiple regulatory relationship pairs based on the adjusted set of regulatory relationships; the regulatory relationship pairs are used to associate genes and regulatory elements that have a regulatory relationship with the genes; the significance evaluation result of the regulatory relationship pairs is determined based on a first significance probability and a second significance probability; the first significance probability reflects the significance of the gene among the various regulatory elements that have a regulatory relationship with the gene, and the second significance probability reflects the significance of the regulatory element among the various genes that have a regulatory relationship with the regulatory element; for each gene in the adjusted set of regulatory relationships, a target regulatory element is selected from the various regulatory elements that have a regulatory relationship with the gene based on the significance evaluation result of each regulatory relationship pair, thereby obtaining the gene's regulatory element identification result.

[0186] Each module in the aforementioned control element identification device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0187] In one exemplary embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 9 As shown, the computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When executed by the processor, the computer program implements a control element identification method. The display unit is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0188] Those skilled in the art will understand that Figure 9 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0189] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0190] In one exemplary embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above-described method embodiments.

[0191] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0192] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0193] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0194] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for identifying control elements, characterized in that, The method includes: Single-cell chromatin accessibility sequencing data and single-cell expression level sequencing data are obtained to obtain a set of regulatory relationships to be adjusted; the single-cell chromatin accessibility sequencing data includes sequencing data reflecting the chromatin accessibility of multiple regulatory elements, the single-cell expression level sequencing data includes sequencing data reflecting the expression of multiple genes on chromosomes, and the parameter value corresponding to each regulatory relationship in the set of regulatory relationships characterizes the regulatory strength of the regulatory element on the expression of the corresponding gene. Based on the set of regulatory relationships and the single-cell chromatin accessibility sequencing data, the expression level prediction results of multiple genes are obtained, and based on the set of regulatory relationships and the single-cell expression level sequencing data, the accessibility prediction results of multiple regulatory elements are obtained. Based on the expression level prediction results, the single-cell expression level sequencing data, the accessibility prediction results, and the single-cell chromatin accessibility sequencing data, the set of regulatory relationships is adjusted to obtain an adjusted set of regulatory relationships. Based on the adjusted set of control relationships, the identification result of the control element is determined.

2. The method according to claim 1, characterized in that, The adjustment of the regulatory relationship set based on the expression level prediction results, the single-cell expression level sequencing data, the accessibility prediction results, and the single-cell chromatin accessibility sequencing data includes: The first loss value is determined based on the gene expression level difference between the predicted expression level and the single-cell expression level sequencing data; A second loss value is determined based on the accessibility difference between the accessibility prediction results and the single-cell chromatin accessibility sequencing data; The weighted regularization loss value is determined based on the parameter value corresponding to each regulation relationship in the set of regulation relationships; The target loss value is determined based on the first loss value, the second loss value, and the weighted regularization loss value; The set of control relationships is adjusted based on the target loss value.

3. The method according to claim 2, characterized in that, The regulatory element includes a cis-regulatory element, and the single-cell expression sequencing data includes sequencing data reflecting the expression of multiple genes on the same chromosome; the differences in gene expression levels are determined through the following steps: Based on the single-cell expression sequencing data, the first expression level of each of the multiple genes on the same chromosome is determined, and based on the expression level prediction results, the second expression level of each of the multiple genes on the same chromosome is determined; For each of the genes on the same chromosome in the same cell, a first difference between the first expression level and the second expression level is obtained; The differences in gene expression levels are determined based on each of the first differences.

4. The method according to claim 2, characterized in that, The regulatory elements include cis-regulatory elements, and the single-cell chromatin accessibility sequencing data includes sequencing data reflecting the chromatin accessibility of each of the multiple cis-regulatory elements on the same chromosome; the accessibility differences are determined through the following steps: Based on the single-cell chromatin accessibility sequencing data, the first chromatin accessibility of each of the multiple cis-regulatory elements on the same chromosome is determined, and based on the accessibility prediction results, the second chromatin accessibility of each of the multiple cis-regulatory elements on the same chromosome is determined. For each of the cis-regulatory elements on the same chromosome of the same cell, a second difference between the first chromatin accessibility and the second chromatin accessibility is determined; The accessibility difference is determined based on each of the second differences.

5. The method according to claim 1, characterized in that, The step of obtaining expression level prediction results for multiple genes based on the set of regulatory relationships and the single-cell chromatin accessibility sequencing data includes: For each of the genes on the same chromosome in the same cell, the regulatory strength of the multiple regulatory elements on the same chromosome on the expression of the gene is determined according to the set of regulatory relationships, and the first chromatin accessibility of the multiple regulatory elements on the same chromosome in the same cell is determined according to the single-cell chromatin accessibility sequencing data. Based on the regulatory strength of multiple regulatory elements on the same chromosome on the expression of the gene and the accessibility of multiple regulatory elements on the same chromosome in the first chromatin of the same cell, the predicted regulatory contribution value of each regulatory element to the gene is determined; the predicted regulatory contribution value is used to indicate the likelihood that the regulatory element will regulate the expression of the gene. The predicted expression level of the gene is determined based on the predicted contribution of each regulatory element to the regulation of the gene.

6. The method according to claim 1, characterized in that, The step of obtaining accessibility prediction results for multiple regulatory elements based on the set of regulatory relationships and the single-cell expression level sequencing data includes: For each of the regulatory elements on the same chromosome in the same cell, the regulatory strength of the regulatory elements on the expression of multiple genes on the same chromosome is determined according to the set of regulatory relationships, and the first expression level of each of the multiple genes on the same chromosome is determined according to the single-cell expression level sequencing data. Based on the regulatory strength of the regulatory elements on the expression of multiple genes on the same chromosome and the first expression level of each of the multiple genes on the same chromosome, a regulatory association prediction value for each of the genes is determined; the regulatory association prediction value is used to indicate the association between the regulatory element and the regulation of gene expression. Based on the predicted regulatory association values ​​of the regulatory elements with each of the genes, the accessibility prediction results of the regulatory elements are determined.

7. The method according to any one of claims 1 to 6, characterized in that, The step of determining the identification result of the control element based on the adjusted control relationship set includes: Based on the adjusted set of regulatory relationships, multiple regulatory relationship pairs are determined; the regulatory relationship pairs are used to associate genes and regulatory elements that have a regulatory relationship with the genes. The significance assessment result of the regulatory relationship pair is determined based on the first significance probability and the second significance probability of the regulatory relationship pair; the first significance probability reflects the significance of the gene in each regulatory element that has a regulatory relationship with the gene, and the second significance probability reflects the significance of the regulatory element in each gene that has a regulatory relationship with the regulatory element. For each gene in the adjusted set of regulatory relationships, based on the significance evaluation results of each pair of regulatory relationships, a target regulatory element is selected from each regulatory element that has a regulatory relationship with the gene, and the regulatory element identification result of the gene is obtained.

8. A control element identification device, characterized in that, The device includes: The first acquisition module is used to acquire single-cell chromatin accessibility sequencing data and single-cell expression level sequencing data, and to acquire a set of regulatory relationships to be adjusted; the single-cell chromatin accessibility sequencing data includes sequencing data reflecting the chromatin accessibility of multiple regulatory elements, the single-cell expression level sequencing data includes sequencing data reflecting the expression of multiple genes on chromosomes, and the parameter value corresponding to each regulatory relationship in the set of regulatory relationships characterizes the regulatory strength of the regulatory element on the expression of the corresponding gene; The second acquisition module is used to acquire the expression level prediction results of multiple genes based on the set of regulatory relationships and the single-cell chromatin accessibility sequencing data, and to acquire the accessibility prediction results of multiple regulatory elements based on the set of regulatory relationships and the single-cell expression level sequencing data. An adjustment module is used to adjust the set of regulatory relationships based on the expression level prediction results, the single-cell expression level sequencing data, the accessibility prediction results, and the single-cell chromatin accessibility sequencing data, to obtain an adjusted set of regulatory relationships. The determination module is used to determine the identification result of the control element based on the adjusted control relationship set.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

11. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images