Cell-specific miRNA causal regulation network inference method, device, electronic device and storage medium

Abstract

The present invention provides a cell-specific miRNA causal regulation network reasoning method, device, electronic device and storage medium, and relates to the field of gene identification technology. The cell-specific miRNA causal regulation network reasoning method includes: obtaining dual expression spectrum data of miRNA and target gene of target single cell. According to the dual expression spectrum data of miRNA and target gene of target single cell, a CPDAG graph is obtained, and the CPDAG graph includes multiple edges. The differential causal effect between background data and perturbation data of each edge in the CPDAG graph is calculated. According to the edges whose differential causal effects meet preset conditions, the cell-specific miRNA causal regulation network of the target single cell is obtained. An inference method for a cell-specific miRNA causal regulation network is provided, which realizes the construction of a cell-specific miRNA causal regulation network for each single cell, and the obtained cell-specific miRNA causal regulation network can reflect the real miRNA causal regulation mechanism in tissues or cells.

Description

Technical Field

[0001] The present invention relates to the field of gene identification technology, and in particular to a method, device, electronic device and storage medium for inferring a cell-specific miRNA causal regulation network. Background Art

[0002] Micro RNA (micro ribonucleic acid, miRNA) is a regulatory non-coding RNA (ribonucleic acid, RNA), which can regulate the expression level of thousands of protein-coding messenger RNA (messenger RNA, mRNA). Gene regulatory network is one of the important biological networks. It is considered to be an important feature of biological tissues or cells and leads to the uniqueness of each biological tissue or cell. The miRNA regulatory network is a type of gene regulatory network. In order to understand the biological cell-specific miRNA regulatory mechanism, it is necessary to study the cell-specific miRNA regulatory network (i.e. one cell corresponds to one miRNA regulatory network).

[0003] Existing technologies can only infer miRNA causal regulatory networks at the tissue or multi-cell level, but cannot infer miRNA causal regulatory networks at the single-cell level, and cannot reflect the true miRNA causal regulatory mechanism in tissues or cells. Summary of the invention

[0004] The purpose of the present invention is to provide a cell-specific miRNA causal regulatory network inference method, device, electronic device and storage medium to address the deficiencies in the above-mentioned prior art, based on single-cell transcriptome data and miRNA prior regulatory information, taking into account the heterogeneity of each single cell, using statistical perturbation strategy and differential causal effect analysis, to construct a miRNA causal regulatory network for each single cell, reflecting the true miRNA causal regulatory mechanism in tissues or cells.

[0005] To achieve the above purpose, the technical solution adopted by the embodiment of the present invention is as follows:

[0006] In a first aspect, an embodiment of the present invention provides a method for inferring a cell-specific miRNA causal regulatory network, comprising: obtaining dual expression spectrum data of the miRNA and target gene of a target single cell, wherein the dual expression spectrum data of the miRNA and target gene of the target single cell includes background data and perturbation data. According to the dual expression spectrum data of the miRNA and target gene of the target single cell, a CPDAG graph is obtained, wherein the CPDAG graph includes a plurality of edges. The differential causal effect of each edge in the CPDAG graph between the background data and the perturbation data is calculated. According to the edges whose differential causal effects meet preset conditions, a cell-specific miRNA causal regulatory network of the target single cell is obtained.

[0007] In some possible implementations, obtaining dual expression profile data of miRNA and target gene of target single cell includes: obtaining transcriptome data of target single cell, and obtaining dual expression profile data of miRNA and target gene of target single cell according to preset gene annotation information. Obtaining perturbation data according to dual expression profile data of miRNA and target gene of target single cell by statistical perturbation strategy, wherein the statistical perturbation strategy includes knocking out a single cell each time statistics are performed. The obtained dual expression profile data of miRNA and target gene of target single cell is used as background data.

[0008] In some possible implementations, a CPDAG graph is obtained based on the dual expression profile data of the miRNA and the target gene of the target single cell, including: obtaining a prior network graph of the miRNA target gene based on the dual expression profile data of the miRNA and the target gene of the target single cell. According to the causal relationship of each edge in the prior network graph, multiple DAG equivalence classes encoding the same conditional independence relationship are obtained. Multiple DAG equivalence classes encoding the same conditional independence relationship are represented by a CPDAG graph.

[0009] In some possible implementations, the differential causal effect of each edge in the CPDAG graph between background data and perturbation data is calculated, including: calculating the first causal effect of each edge in the CPDAG graph in the background data by correcting the potential confounding factor. Calculating the second causal effect of each edge in the CPDAG graph in the perturbation data by correcting the potential confounding factor. Obtaining the differential causal effect based on the first causal effect and the second causal effect.

[0010] In some possible implementations, after obtaining the cell-specific miRNA causal regulatory network of the target single cell, the method further includes: verifying the cell-specific miRNA causal regulatory network of the target single cell through at least one analysis method of network similarity analysis, clustering analysis, and verification analysis.

[0011] In a second aspect, an embodiment of the present invention further provides a cell-specific miRNA causal regulation network inference device, comprising:

[0012] The acquisition module is used to obtain the dual expression spectrum data of the miRNA and target gene of the target single cell, and the dual expression spectrum data of the miRNA and target gene of the target single cell includes background data and perturbation data. The acquisition module is also used to obtain a CPDAG graph based on the dual expression spectrum data of the miRNA and target gene of the target single cell, and the CPDAG graph includes multiple edges. The calculation module is used to calculate the differential causal effect of each edge in the CPDAG graph between the background data and the perturbation data. The acquisition module is also used to obtain the cell-specific miRNA causal regulatory network of the target single cell based on the edges whose differential causal effects meet the preset conditions.

[0013] In some possible implementations, the acquisition module is specifically used to acquire the transcriptome data of the target single cell, and acquire the dual expression profile data of the miRNA and the target gene of the target single cell according to the preset gene annotation information. The perturbation data is acquired according to the dual expression profile data of the miRNA and the target gene of the target single cell through a statistical perturbation strategy, and the statistical perturbation strategy includes knocking out a single cell each time the statistics are performed. The acquired dual expression profile data of the miRNA and the target gene of the target single cell is used as background data.

[0014] In some possible implementations, the acquisition module is specifically used to obtain a prior network diagram of miRNA target genes based on the dual expression profile data of miRNA and target genes of the target single cell. According to the causal relationship of each edge in the prior network diagram, multiple DAG equivalence classes encoding the same conditional independence relationship are obtained. Multiple DAG equivalence classes encoding the same conditional independence relationship are represented by a CPDAG diagram.

[0015] In some possible implementations, the calculation module is specifically used to calculate the first causal effect of each edge in the CPDAG graph in the background data by correcting the potential confounding factor. The second causal effect of each edge in the CPDAG graph in the perturbation data is calculated by correcting the potential confounding factor. According to the first causal effect and the second causal effect, the differential causal effect is obtained.

[0016] In some possible implementations, the device also includes a verification module for verifying the cell-specific miRNA causal regulatory network of the target single cell through at least one analysis method of network similarity analysis, cluster analysis, and verification analysis.

[0017] In a third aspect, an embodiment of the present invention provides an electronic device, comprising: a processor, a storage medium and a bus, wherein the storage medium stores machine-readable instructions executable by the processor. When the electronic device is running, the processor and the storage medium communicate through the bus, and the processor executes the machine-readable instructions to perform the steps of any method of the first aspect above.

[0018] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of any method of the first aspect are executed.

[0019] The beneficial effects of the present invention are: a CPDAG graph is obtained by obtaining the dual expression profile data of the miRNA and target gene of the target single cell, and the differential causal effect between the background data and the perturbation data of each edge in the CPDAG graph is calculated. The cell-specific miRNA causal regulatory network of the target single cell is obtained according to the differential causal effect. An inference method for a cell-specific miRNA causal regulatory network is provided, which realizes the construction of a cell-specific miRNA causal regulatory network for each single cell, and the obtained cell-specific miRNA causal regulatory network can reflect the real miRNA causal regulatory mechanism in tissues or cells. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0021] Figure 1 A schematic diagram of a process flow of a cell-specific miRNA causal regulatory network inference method provided in one embodiment of the present application;

[0022] Figure 2 A schematic diagram of the process of S110 in the cell-specific miRNA causal regulatory network inference method provided in one embodiment of the present application;

[0023] Figure 3 A schematic diagram of the process of S120 in the cell-specific miRNA causal regulatory network inference method provided in one embodiment of the present application;

[0024] Figure 4 A schematic diagram of the process of S130 in the cell-specific miRNA causal regulatory network inference method provided in one embodiment of the present application;

[0025] Figure 5 A similarity graph of a miRNA-mRNA causal regulatory network obtained in the cell-specific miRNA causal regulatory network inference method provided in an embodiment of the present application;

[0026] Figure 6 A schematic diagram of the structure of a cell-specific miRNA causal regulatory network inference device provided in one embodiment of the present application;

[0027] Figure 7 A schematic diagram of the structure of an electronic device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0028] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments.

[0029] Figure 1 A flow chart of a method for inference of a cell-specific miRNA causal regulatory network is provided for one embodiment of the present application, wherein the method may be executed by a desktop computer, a laptop computer, a server, a cloud server, a smart terminal, a tablet computer, or other device with data processing capabilities, without limitation.

[0030] like Figure 1 As shown, the method includes:

[0031] S110. Obtain dual expression profile data of miRNA and target gene of target single cell.

[0032] In some possible implementations, the transcriptome data of the target single cell may be given first, and then the dual expression profile data of the miRNA and target gene of the target single cell may be obtained using preset gene annotation information.

[0033] As an example, the dual expression profile data of miRNA and target gene of a target single cell may include:

[0034]

[0035]

[0036] There are n single cells, and the number of miRNAs and target genes in each single cell is r and t respectively. R represents the expression profile data of miRNA in each single cell, and T represents the expression profile data of target genes in each single cell.

[0037] As an example, the target gene of miRNA can be messenger RNA (mRNA), long non-coding RNA (lncRNA), pseudogene, circular RNA (circRNA), etc.

[0038] In some possible implementations, the dual expression profile data of miRNA and target gene of the target single cell includes background data and perturbation data. The perturbation data is a plurality of dual expression profile data of miRNA and target gene obtained by using a statistical perturbation strategy (such as knocking out one single cell at a time) based on the background data.

[0039] S120. Obtain a CPDAG map based on the dual expression profile data of miRNA and target gene of the target single cell.

[0040] In some possible implementations, a CPDAG graph refers to a Complete Partially Directed Acyclic Graph (CPDAG). A CPDAG graph can represent multiple directed acyclic graphs (DAGs) equivalence classes encoding the same conditional independence relationship, and the CPDAG graph contains both directed edges and undirected edges. An edge in a CPDAG graph includes two vertices R j (j∈{1,2,...,r}) and T k (k∈{1,2,...,t}). j →T k Indicates a regulatory relationship.

[0041] S130. Calculate the differential causal effect of each edge in the CPDAG graph between the background data and the perturbation data.

[0042] In some possible implementations, the first causal effect in the CPDAG diagram may be calculated based on the background data, and then the second causal effect in the CPDAG diagram may be calculated based on the disturbance data. Finally, the differential causal effect may be calculated based on the difference between the first causal effect and the second causal effect.

[0043] S140. Based on the edges whose differential causal effects meet the preset conditions, the cell-specific miRNA causal regulatory network of the target single cell is obtained.

[0044] In some possible implementations, an edge in the CPDAG graph (i.e., R j →T k ) is less than the preset p threshold, it can be determined that there is a significant difference in causal effect between the two nodes of the edge. j →T k , the cell-specific miRNA causal regulatory network of the target single cell can be obtained.

[0045] In this embodiment, a CPDAG graph is obtained by obtaining dual expression profile data of miRNA and target gene of a target single cell, and the differential causal effect between background data and perturbation data of each edge in the CPDAG graph is calculated. According to the differential causal effect, a cell-specific miRNA causal regulatory network of the target single cell is obtained. An inference method for a cell-specific miRNA causal regulatory network is provided, which realizes the construction of a cell-specific miRNA causal regulatory network for each single cell, and the obtained cell-specific miRNA causal regulatory network can reflect the real miRNA causal regulatory mechanism in tissues or cells.

[0046] Figure 2 This is a schematic diagram of the process of S110 in the cell-specific miRNA causal regulatory network inference method provided in one embodiment of the present application.

[0047] Some possible implementations include Figure 2 , obtain dual expression profile data of miRNA and target gene of target single cell, including:

[0048] S1101. Acquire transcriptome data of a target single cell, and acquire dual expression profile data of miRNA and target gene of the target single cell according to preset gene annotation information.

[0049] In some possible implementations, taking the target single cell as a leukemia cell and the target gene as mRNA as an example, the transcriptome data of the target single cell can be obtained by collecting the single cell transcriptome data of K562 leukemia cells from the Gene Expression Omnibus (GEO) (data number GSE114071). The collected single cell transcriptome data is preprocessed, such as logarithmic processing, removal of duplicate genes, and removal of genes that are constantly expressed in all single cells. After preprocessing, 212 miRNAs and 15,361 mRNA expression profile data of 19 K562 leukemia cells can be obtained.

[0050] The preset gene annotation information can be the prior information of miRNA target genes, for example, it can be the prior network of miRNA target genes from the TargetScan v8.0 database, which can include 235,109 miRNA-mRNA prior regulatory relationship pairs. Alternatively, it can be experimentally verified miRNA target gene experimentally verified regulatory relationship pairs from the miRTarBase v9.0 and TarBase v8.0 databases, from which 762,540 miRNA target gene experimentally verified regulatory relationship pairs can be obtained. By retaining the same miRNA and mRNA in the single-cell transcriptome data and the miRNA target gene prior information, 108 miRNAs and 7,450 mRNA expression profile data of 19 K562 leukemia cells were obtained.

[0051] Therefore, in this embodiment, the dual expression profile data (R and T) of the target single cell miRNA and target gene are:

[0052]

[0053]

[0054] In addition, 59,457 miRNA-mRNA prior regulatory relationship pairs and 110,914 miRNA-mRNA experimentally verified regulatory relationship pairs related to single-cell transcriptome data were obtained.

[0055] S1102. Obtain perturbation data based on the dual expression profile data of miRNA and target gene of the target single cell through a statistical perturbation strategy.

[0056] S1103. Use the acquired dual expression profile data of miRNA and target gene of the target single cell as background data.

[0057] In some possible implementations, the statistical perturbation strategy includes knocking out a single cell each time statistics are performed. The target single cell miRNA and target gene dual expression profile data (R and T) include n single cells, so n miRNA and target gene dual expression profile data can be obtained as perturbation data.

[0058] Figure 3 This is a schematic diagram of the process of S120 in the cell-specific miRNA causal regulatory network inference method provided in one embodiment of the present application.

[0059] For some possible implementations, refer to Figure 3 , based on the dual expression profile data of miRNA and target gene of the target single cell, obtaining the CPDAG map may include:

[0060] S1201. Obtain a priori network diagram of miRNA target genes based on dual expression profile data of miRNA and target genes of target single cells.

[0061] In some possible implementations, assuming that the miRNA target gene prior network graph G contains V nodes and E edges, then:

[0062] G=(V,E)

[0063] Among them, V represents a set of random variables consisting of r miRNAs and t target genes, and E represents the regulatory relationship between r miRNAs and t target genes.

[0064] V={R1,R2,…,R r ,T1,T2,…,T t}

[0065] Since there is no regulatory relationship between r miRNAs or t target genes, the miRNA target gene prior network graph G is a bipartite graph (BG). At the same time, the miRNA target gene prior network graph G only contains directed acyclic edges, so the miRNA target gene prior network graph G is also a directed acyclic graph (DAG).

[0066] S1202. According to the causal relationship of each edge in the prior network graph, multiple DAG equivalence classes encoding the same conditional independence relationship are obtained.

[0067] In some possible implementations, if there is an edge R in a DAG j →T k , then R j (j∈{1,2,...,r}) is T k The parent node of (k∈{1,2,...,t}), T k Then R j If and only if R j T k The DAG is called a causal DAG. The Markov hypothesis states that, given the parent node, a node in the DAG is conditionally independent from other non-child nodes.

[0068] S1203. Represent multiple DAG equivalence classes encoding the same conditional independence relations through a CPDAG graph.

[0069] In some possible implementations, given an observation data, multiple DAG equivalence classes encoding the same conditional independence relationship can be represented using CPDAG, which contains both directed edges and undirected edges.

[0070] As an example, the Peter Spirtes and Clark Glymour (PC) algorithm can be used to perform causal structure learning from the target single-cell miRNA and target gene dual expression profile data to identify the CPDAG causal structure graph.

[0071] Figure 4 This is a schematic diagram of the process of S130 in the cell-specific miRNA causal regulatory network inference method provided in one embodiment of the present application.

[0072] For some possible implementations, refer to Figure 4 , calculating the differential causal effect of each edge in the CPDAG graph between background data and perturbation data, can include:

[0073] S1301. Calculate the first causal effect of each edge in the CPDAG graph in the background data by correcting potential confounding factors.

[0074] In some possible implementations, the R in the CPDAG causal structure diagram can be estimated by the Do-calculus method. j →T k The do-calculus method calculates the causal effect of R in each DAG graph by simulating biological control experiments. j →T k causal effect.

[0075] For example, since a CPDAG may have multiple j →T k DAG diagram of the relationship, the do-calculus method needs to calculate R j →T k In order to obtain the causal effects in multiple DAG graphs, j →T k The causal effect is usually expressed as R j →T k The lower limit of the absolute value of the causal effect is taken as R j →T k For example, suppose R j →T k There are three DAG graphs, and the causal effects calculated by the do-calculus method are 0.5, -0.75, and 0.8 respectively. Then R j →T k The final causal effect is 0.5.

[0076] Given the dual expression profile data of single-cell miRNA and target genes, the R calculated by the do-calculus method j →T k The causal effect CE is defined as follows:

[0077]

[0078] To correct for potential confounding factors, consider the correction set R j →T k The causal effect CE corresponds to the linear regression coefficient β of the following formula:

[0079]

[0080] Among them, β0 represents the intercept, and η represents random noise with mean 0. Satisfies the backdoor criteria. j→T k The first causal effect in the background data can correspond to the linear regression coefficient β B .

[0081] S1302. Calculate the second causal effect of each edge in the CPDAG graph in the perturbed data by correcting potential confounding factors.

[0082] In some possible implementations, the algorithm for the second causal effect in S1302 is similar to that in S1301, R j →T k The second causal effect in the perturbed data can correspond to the linear regression coefficient β P .

[0083] S1303. Obtain a differential causal effect based on the first causal effect and the second causal effect.

[0084] In some possible implementations, the difference causal effect δ between background data and perturbation data can be defined as: δ = β P -β B

[0085] Calculate R by fitting a joint model of background and perturbation data j →T k The differential causal effect δ can also be calculated, and the significance of δ can also be calculated. Assume that I is an indicator random variable. If the observation comes from the background data, then I is equal to 1, and if the observation comes from the perturbed data, then I is equal to 0. The differential causal effect δ can be calculated jointly from all single cells by fitting the following linear model:

[0086]

[0087] In some possible implementations, the significance of the differential causal effect δ now corresponds to the significance of the coefficient estimate in the linear model. Using heteroscedasticity-consistent standard errors, asymptotically valid confidence intervals and significance p-values ​​for δ can be produced without taking into account the dependence of the noise level on the predicted values.

[0088] Given a p threshold of differential causal effect δ, we can determine R j →T k Is there a significant difference in causal effects between the background data and the perturbation data? In single cell i, merge all regulatory relationships R with significant differences in causal effects j →T k , and then obtain the cell i-specific miRNA causal regulatory network.

[0089] In some possible implementations, after obtaining the cell-specific miRNA causal regulatory network of the target single cell, the method further includes: verifying the cell-specific miRNA causal regulatory network of the target single cell through at least one analysis method of network similarity analysis, clustering analysis, and verification analysis.

[0090] The following is explained with a specific example:

[0091] First, assuming that the target single-cell transcriptome data and miRNA target gene prior information are obtained in a similar way as in S1101, the target single-cell miRNA and target gene dual expression profile data (R and T) can be obtained as follows:

[0092]

[0093]

[0094] At the same time, 59,457 miRNA-mRNA prior regulatory relationship pairs and 110,914 miRNA-mRNA experimentally verified regulatory relationship pairs related to the single-cell transcriptome data were obtained.

[0095] Then, the cell-specific miRNA causal regulatory network of leukemia cell i is calculated and inferred by the method shown in S130.

[0096] For example, given the miRNA and mRNA dual expression profile data (R and T) of K562 leukemia cells and the prior information of miRNA target genes, the cell-specific miRNA-mRNA causal regulatory network is inferred. In the PC algorithm, the significance level α of the conditional independent relationship test is set to 0.01. The p threshold of the differential causal effect of the causal regulatory relationship between miRNA and mRNA is set to 0.05. In each leukemia cell i, all miRNA-mRNA regulatory relationships with significant differential causal effects are fused to obtain the leukemia cell i-specific miRNA causal regulatory network.

[0097] Finally, the cell-specific miRNA causal regulatory network of the target single cell can be verified by at least one of network similarity analysis, cluster analysis, and verification analysis.

[0098] As an example, suppose the miRNA-mRNA causal regulatory network CN of given cell i and cell j is obtained i and CN j , network similarity value Sim(CN i ,CN j ) is calculated as follows:

[0099]

[0100] Among them, |CN i ∩CN j | is the logarithm of the same miRNA-mRNA causal regulatory relationship in the two networks, min(|CN i |,|CN j |) is the logarithm of the miRNA-mRNA causal regulatory relationship of the smallest network in the two networks. Sim(CN i ,CN j ) ranges from [0 1], and the larger the value, the more similar the two networks are.

[0101] As an example, the network similarity matrix Sim (i.e., the single-cell similarity matrix) can be obtained according to the above network similarity analysis:

[0102]

[0103] According to the network similarity matrix Sim, the single cell distance matrix Dis can be obtained:

[0104]

[0105] Then, based on the single-cell distance matrix Dis, the hierarchical clustering method was used to perform cluster analysis on the 19 single cells.

[0106] As an example, based on the experimentally verified miRNA-mRNA regulatory relationships collected from a third-party database, the overlap percentage between the miRNA-mRNA regulatory relationships in the cell-specific miRNA causal regulatory network and the experimentally verified miRNA-mRNA regulatory relationships can be obtained. A larger overlap percentage indicates that more predicted miRNA-mRNA regulatory relationships in the single cell are experimentally confirmed.

[0107] Figure 5 This is a miRNA-mRNA causal regulatory network similarity diagram obtained in the cell-specific miRNA causal regulatory network inference method provided in one embodiment of the present application.

[0108] In this embodiment, reference Figure 5 The similarity values ​​of the causal regulatory networks of miRNAs in 19 leukemia cells range from [0, 0.71]. This result shows that the causal regulatory network of miRNAs in each leukemia cell is very different.

[0109] Table 1 shows the predicted relationship and experimental verification relationship of leukemia miRNA-mRNA causal regulation in this embodiment. Referring to Table 1, the hierarchical clustering method can divide 19 leukemia cells into 2 clear categories. Table 2 shows the clustering analysis results in this embodiment. As shown in Table 2, the predicted miRNA-mRNA causal regulation relationship and the experimental verification miRNA-mRNA regulation relationship in each leukemia cell also have great differences.

[0110] Table 1

[0111]

[0112]

[0113] Table 2

[0114]

[0115] The above results show that the miRNA regulatory network is cell-specific and that miRNA participates in cellular biological processes in a causal regulatory manner. Therefore, the miRNA causal regulatory network must be inferred from the single-cell precision level. In addition, a cell-cell correlation network can be constructed through the cell-specific miRNA causal regulatory network, providing a new method for cell type annotation. In summary, the cell-specific miRNA causal regulatory network inference method proposed in the present invention can identify cancer cell-specific miRNA causal regulatory networks, provide technical support and understanding methods for personalized diagnosis and targeted treatment of human cancer in clinical practice, and has important biological significance.

[0116] Figure 6 A schematic diagram of the structure of a cell-specific miRNA causal regulation network inference device provided in one embodiment of the present application is shown in FIG. Figure 6 As shown,

[0117] A cell-specific miRNA causal regulation network inference device, comprising:

[0118] The acquisition module 21 is used to acquire the dual expression profile data of the miRNA and the target gene of the target single cell, wherein the dual expression profile data of the miRNA and the target gene of the target single cell includes background data and disturbance data.

[0119] The acquisition module 21 is also used to obtain a CPDAG graph based on the dual expression profile data of the miRNA and the target gene of the target single cell, and the CPDAG graph includes multiple edges.

[0120] The calculation module 22 is used to calculate the difference causal effect between the background data and the perturbation data for each edge in the CPDAG graph.

[0121] The acquisition module 21 is also used to obtain the cell-specific miRNA causal regulatory network of the target single cell based on the edges whose differential causal effects meet the preset conditions.

[0122] In some possible implementations, the acquisition module 21 is specifically used to acquire the transcriptome data of the target single cell, and acquire the dual expression profile data of the miRNA and the target gene of the target single cell according to the preset gene annotation information. The perturbation data is acquired according to the dual expression profile data of the miRNA and the target gene of the target single cell through a statistical perturbation strategy, and the statistical perturbation strategy includes knocking out a single cell each time the statistics are performed. The acquired dual expression profile data of the miRNA and the target gene of the target single cell is used as background data.

[0123] In some possible implementations, the acquisition module 21 is specifically used to obtain a prior network diagram of the miRNA target gene based on the dual expression profile data of the miRNA and the target gene of the target single cell. According to the causal relationship of each edge in the prior network diagram, multiple DAG equivalence classes encoding the same conditional independence relationship are obtained. Multiple DAG equivalence classes encoding the same conditional independence relationship are represented by a CPDAG diagram.

[0124] In some possible implementations, the calculation module 22 is specifically used to calculate the first causal effect of each edge in the CPDAG graph in the background data by correcting the potential confounding factor. The second causal effect of each edge in the CPDAG graph in the perturbation data is calculated by correcting the potential confounding factor. According to the first causal effect and the second causal effect, the differential causal effect is obtained.

[0125] In some possible implementations, the device further includes a verification module 23 for verifying the cell-specific miRNA causal regulation network of the target single cell by at least one of network similarity analysis, cluster analysis, and verification analysis.

[0126] The above-mentioned device is used to execute the method provided by the aforementioned embodiment, and its implementation principle and technical effect are similar, which will not be repeated here.

[0127] The above modules may be one or more integrated circuits configured to implement the above methods, such as one or more application specific integrated circuits (ASIC), or one or more digital singnal processors (DSP), or one or more field programmable gate arrays (FPGA). For another example, when a module is implemented in the form of a processing element scheduling program code, the processing element may be a general-purpose processor, such as a central processing unit (CPU) or other processor that can call program code. For another example, these modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).

[0128] Figure 7 A schematic diagram of the structure of an electronic device provided in one embodiment of the present application.

[0129] like Figure 7 As shown, the electronic device includes: a processor 31, a computer readable storage medium 32 and a bus 33, wherein:

[0130] The electronic device may include one or more processors 31, a bus 33 and a storage medium 32, wherein the storage medium 32 is used to store machine-readable instructions, the processor 31 is communicatively connected to the storage medium 32 via the bus 33, and the processor 31 executes the machine-readable instructions stored in the storage medium 32 to execute the above method embodiments.

[0131] The electronic device may be a general-purpose computer, a server, or a mobile terminal, etc., which is not limited here. The electronic device is used to implement the above method embodiment of the present application.

[0132] It should be noted that the processor 31 may include one or more processing cores (e.g., a single-core processor or a multi-core processor). By way of example only, the processor may include a central processing unit (CPU), an application specific integrated circuit (ASIC), an application specific instruction set processor (ASIP), a graphics processing unit (GPU), a physical processing unit (PPU), a digital signal processor (DSP), a field programmable gate array (FPGA), a programmable logic device (PLD), a controller, a microcontroller unit, a reduced instruction set computer (RISC), or a microprocessor, or any combination thereof.

[0133] The storage medium 32 may include: a large capacity memory, a removable memory, a volatile read-write memory, or a read-only memory (ROM), etc., or any combination thereof. As an example, the large capacity memory may include a disk, an optical disk, a solid-state drive, etc.; the removable memory may include a flash drive, a floppy disk, an optical disk, a memory card, a zip disk, a magnetic tape, etc.; the volatile read-write memory may include a random access memory (RAM); the RAM may include a dynamic RAM (DRAM), a double data rate synchronous dynamic RAM (DDR SDRAM); a static RAM (SRAM), a thyristor RAM (T-RAM), a zero capacitor RAM (Zero-RAM), etc. As an example, ROM may include mask ROM (Mask Read-Only Memory, MROM), programmable ROM (Programmable Read-Only Memory, PROM), erasable programmable ROM (ProgrammableErasable Read-only Memory, PEROM), electrically erasable programmable ROM (Electrically ErasableProgrammable read only memory, EEPROM), compact disc ROM (CD-ROM), and digital versatile disk ROM, etc.

[0134] For ease of explanation, only one processor 31 is described in the electronic device. However, it should be noted that the electronic device in the present application may also include multiple processors 31, so the steps performed by one processor described in the present application may also be performed jointly or individually by multiple processors. For example, if the processor 31 of the electronic device performs step A and step B, it should be understood that step A and step B may also be performed jointly by two different processors or individually in one processor. For example, the first processor performs step A, the second processor performs step B, or the first processor and the second processor perform steps A and B together.

[0135] Optionally, the present invention further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above method are executed.

[0136] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0137] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0138] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of modules or units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0139] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0140] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A cell-specific miRNA causal regulatory network inference method, characterized in that: include: Acquiring dual expression profile data of miRNA and target gene of a target single cell, wherein the dual expression profile data of miRNA and target gene of the target single cell includes background data and perturbation data; According to the dual expression profile data of the miRNA and the target gene of the target single cell, a CPDAG graph is obtained, wherein the CPDAG graph includes a plurality of edges; Calculating the differential causal effect of each edge in the CPDAG graph between background data and perturbation data; According to the edges whose differential causal effects meet preset conditions, a cell-specific miRNA causal regulatory network of the target single cell is obtained; The step of obtaining a CPDAG graph based on the dual expression profile data of the miRNA and the target gene of the target single cell comprises: obtaining a priori network graph of the miRNA target gene based on the dual expression profile data of the miRNA and the target gene of the target single cell; obtaining a plurality of DAG equivalence classes encoding the same conditional independence relationship based on the causal relationship of each edge in the priori network graph; and representing the plurality of DAG equivalence classes encoding the same conditional independence relationship through the CPDAG graph; The calculation of the differential causal effect of each edge in the CPDAG graph between the background data and the perturbation data includes: calculating the first causal effect of each edge in the CPDAG graph in the background data by correcting the potential confounding factor; calculating the second causal effect of each edge in the CPDAG graph in the perturbation data by correcting the potential confounding factor; and obtaining the differential causal effect based on the first causal effect and the second causal effect.

2. The method according to claim 1, characterized in that The step of obtaining dual expression profile data of miRNA and target gene of a target single cell includes: Acquire the transcriptome data of the target single cell, and acquire the dual expression profile data of the miRNA and the target gene of the target single cell according to the preset gene annotation information; The perturbation data is obtained according to the dual expression profile data of the miRNA and the target gene of the target single cell by using a statistical perturbation strategy, wherein the statistical perturbation strategy includes knocking out a single cell each time statistics are performed; The obtained dual expression profile data of miRNA and target gene of the target single cell is used as the background data.

3. The method according to claim 1 or 2, characterized in that: After obtaining the cell-specific miRNA causal regulatory network of the target single cell, the method further comprises: The cell-specific miRNA causal regulatory network of the target single cell is verified by at least one of network similarity analysis, cluster analysis and verification analysis.

4. A cell-specific miRNA causal regulation network inference device, characterized in that: include: An acquisition module, used to acquire dual expression profile data of miRNA and target gene of a target single cell, wherein the dual expression profile data of miRNA and target gene of the target single cell includes background data and perturbation data; The acquisition module is further used to acquire a CPDAG graph according to the dual expression profile data of the miRNA and the target gene of the target single cell, wherein the CPDAG graph includes a plurality of edges; A calculation module, used to calculate the differential causal effect of each edge in the CPDAG graph between background data and perturbation data; The acquisition module is further used to obtain the cell-specific miRNA causal regulatory network of the target single cell according to the edges whose differential causal effects meet the preset conditions; The acquisition module is specifically used to acquire the prior network diagram of the miRNA target gene according to the dual expression profile data of the miRNA and the target gene of the target single cell; obtain multiple DAG equivalent classes encoding the same conditional independence relationship according to the causal relationship of each edge in the prior network diagram; and represent the multiple DAG equivalent classes encoding the same conditional independence relationship through the CPDAG diagram; The calculation module is specifically used to calculate the first causal effect of each edge in the CPDAG graph in the background data by correcting the potential confounding factor; calculate the second causal effect of each edge in the CPDAG graph in the disturbance data by correcting the potential confounding factor; and obtain the differential causal effect based on the first causal effect and the second causal effect.

5. The device according to claim 4, characterized in that The acquisition module is specifically used to acquire the transcriptome data of the target single cell, and acquire the dual expression profile data of the miRNA and the target gene of the target single cell according to the preset gene annotation information; The perturbation data is obtained according to the dual expression profile data of the miRNA and the target gene of the target single cell by using a statistical perturbation strategy, wherein the statistical perturbation strategy includes knocking out a single cell each time statistics are performed; The obtained dual expression profile data of miRNA and target gene of the target single cell is used as the background data.

6. An electronic device, characterized in that: The electronic device comprises a processor, a storage medium and a bus, wherein the storage medium stores machine-readable instructions executable by the processor. When the electronic device is running, the processor communicates with the storage medium via the bus, and the processor executes the machine-readable instructions to execute the method described in any one of claims 1 to 3.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 3 is executed.

Images