A method for establishing a hypoxia-induced model using cobalt chloride

Abstract

The present application relates to the technical field of hypoxia induction model, and particularly relates to a method for establishing a hypoxia induction model by using cobalt chloride, comprising: sequentially exposing a plurality of types of experimental cells to cobalt chloride solutions with a fixed concentration step gradient decreasing for treatment; establishing a relationship network of the hypoxia induction model; inputting target cells and corresponding expected hypoxia types into the relationship network of the hypoxia induction model respectively and outputting predicted optimal treatment conditions and predicted apoptosis rates respectively, wherein, a recommended label of a cell induction parameter of the target cells of the relationship network of the hypoxia induction model is labeled according to a cosine similarity of a gene expression profile; obtaining a measured apoptosis rate; determining a correction instruction according to a deviation amount of the predicted apoptosis rate, including determining a gradient concentration of the concentration decreasing treatment, or determining a number of newly added cycle treatments; and adjusting the relationship network of the hypoxia induction model according to the correction instruction. The present application realizes the improvement of the accuracy of the model output result.

Description

Technical Field

[0001] This invention relates to the field of hypoxia-induced model technology, and in particular to a method for establishing a hypoxia-induced model using cobalt chloride. Background Technology

[0002] In existing technologies, cobalt chloride induces a hypoxic response in cells by mimicking the mechanism of hypoxia-inducible factor stabilization. Under normal oxygen conditions, HIF-1α protein is hydroxylated by proline hydroxylase and degraded by ubiquitin ligase. Cobalt chloride, as a competitive inhibitor of iron ions, can inhibit the activity of PHD, thereby preventing the degradation of HIF-1α, allowing it to accumulate under normoxic conditions and activate the expression of downstream hypoxia-related genes. The concentration of cobalt chloride is usually between 50 μM and 300 μM, and the treatment time varies from several hours to several days depending on the cell type and experimental purpose. The optimal concentration and treatment time need to be determined through preliminary experiments to avoid excessive toxicity or insufficient efficacy.

[0003] Chinese Patent Publication No. CN107182937B discloses a method for constructing an in vivo hypoxic-type prediabetes animal model, comprising: feeding rats with a high-fat, high-sugar diet for 24-30 weeks. The hypoxic-type prediabetes animal model is successfully constructed when the rats exhibit the following symptoms: decreased tissue oxygen concentration; deterioration of hemorheological parameters; and increased expression of hypoxia-inducible factor (HIF). The high-sugar, high-fat diet formulation, by weight, comprises 18% lard, 15% sucrose, 12% egg yolk powder, 5% casein, 1.2% cholesterol, 0.2% sodium cholate, and 48.6% rat maintenance diet. The decreased tissue oxygen concentration is defined as an oxygen saturation of less than 90% as measured by a pulse oximeter. The deterioration of hemorheological parameters is defined as significantly higher than normal values ​​for whole blood viscosity, plasma viscosity, and hematocrit as measured by a hemorheological analyzer. The increased expression of HIF is defined as an expression of HIF in plasma ranging from 0 to orders of magnitude greater than 10, as measured by enzyme-linked immunosorbent assay (ELISA). Therefore, it can be seen that the method for constructing the in vivo hypoxic prediabetic animal model has the problem of increased deviation in apoptosis rate prediction due to the different sensitivity of different types of cells to cobalt chloride. Summary of the Invention

[0004] Therefore, the present invention provides a hypoxia-induced model using cobalt chloride to overcome the problem in the prior art where differences in the sensitivity of different cell types to cobalt chloride lead to increased deviations in apoptosis rate prediction.

[0005] To achieve the above objectives, this invention provides a hypoxia-induced model using cobalt chloride, comprising: Several types of experimental cells were sequentially exposed to cobalt chloride solutions with a fixed concentration gradient and the cell induction parameters of each treatment group were detected. The cell induction parameters included induction treatment time, concentration of cobalt chloride solution, apoptosis rate of experimental cells, gene expression profile of experimental cells, and duration of HIF-1α peak. A relational network for the hypoxia-induced model was established based on the cell induction parameters of each treatment result. Obtain the cell induction parameters of the target cells and the corresponding expected hypoxia type; The target cells and the corresponding expected hypoxia type are respectively input into the relational network of the hypoxia-induced model, and the predicted optimal treatment conditions and predicted apoptosis rate are respectively output, wherein, Recommended labels for the cell induction parameters of the target cells in the hypoxia-induced model relationship network are labeled based on the cosine similarity of the gene expression profiles of the target cells and known cells. The target cells were treated under the optimal treatment conditions described above to obtain the measured apoptosis rate. The correction instructions are determined based on the deviation between the measured apoptosis rate of the target cells and the predicted apoptosis rate output by the hypoxia-induced model. These instructions may include determining the gradient concentration of the concentration-decreasing treatment or determining the number of additional cycles. The relational network of the hypoxia-induced model is adjusted according to the correction instructions.

[0006] Furthermore, the relational network of the hypoxia-induced model is established with cell induction parameters as nodes and the similarity of cell induction parameters of several groups of experimental cells as edge weights.

[0007] Furthermore, the anticipated hypoxia types include acute hypoxia with a peak duration of HIF-1α less than or equal to a preset duration and chronic hypoxia with a peak duration of HIF-1α greater than the preset duration.

[0008] Furthermore, the deviation is the absolute value of the difference between the measured apoptosis rate and the predicted apoptosis rate.

[0009] Furthermore, if the expected hypoxia type is acute and the deviation is greater than or equal to a preset deviation, then the gradient concentration is reduced.

[0010] Furthermore, the gradient concentration is determined based on the difference between the deviation and the preset deviation.

[0011] Furthermore, if the hypoxia type is expected to be chronic and the deviation is greater than or equal to the preset deviation, the duration of a single treatment is reduced, and the number of additional gradient treatments is determined based on the duration of the single treatment.

[0012] Furthermore, the processing time per cycle is determined based on the difference between the deviation and the preset deviation.

[0013] Furthermore, recommended labels for the cell induction parameters of the target cells in the hypoxia-induced model's relational network include: The cosine similarity is compared with a preset similarity. If the cosine similarity is greater than or equal to the preset similarity, then the expected optimal treatment conditions and the expected apoptosis rate are output. If the cosine similarity is less than the preset similarity, the cell induction parameters of the node with the maximum similarity are extracted as the recommended labels for the target cells of the hypoxia-induced model's relational network.

[0014] Furthermore, the gradient concentration is the absolute value of the concentration difference between two adjacent cobalt chloride solutions when the solutions are exposed to a cobalt chloride solution with a fixed concentration step size decreasing.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: The method of the present invention, through the system described herein, treats several types of experimental cells by sequentially exposing them to a cobalt chloride solution with a fixed concentration gradient decreasing step size, and detects the cell induction parameters of each treatment group. Due to the increased prediction bias of apoptosis rate caused by the difference in sensitivity of different cell types to cobalt chloride, a relational network is established with cell induction parameters as nodes and the similarity between parameter groups as edge weights. This integrates multiple types of data and multiple hypoxia types, thereby covering the characteristics of different cell types and reducing the prediction bias of apoptosis rate caused by differences in cell sensitivity. Furthermore, errors in experimental parameters can lead to different induction responses in the same cell type in different experiments, or differences in gene expression profiles at different passage stages or differentiation states—that is, aging cells or cells at different stages of the cell cycle—respond differently to hypoxic stimuli. Unlike other methods, this approach calculates the cosine similarity of gene expression profiles between target cells and known cells to identify the known cell type most similar to the target cell in gene expression patterns. This allows for the recommendation of appropriate cell induction parameters. This similarity-based recommendation mechanism enhances the model's ability to reasonably predict hypoxia responses even in the absence of complete experimental data on target cells. Furthermore, due to varying air exposure times before solution exposure during batch cell treatment, systematic errors arise between model predictions and measured values. By comparing the measured apoptosis rate of target cells with the model's predicted apoptosis rate, the deviation is calculated. Based on this deviation, dynamic correction instructions are generated, including adjusting the gradient concentration of the decreasing treatment or adding more cycles. This targeted compensation for variable perturbations introduced by differences in operational timing improves the model's accuracy in practical applications. Furthermore, the method of the present invention improves the accuracy of induction modeling by obtaining the cell induction parameters of the target cells and the corresponding expected hypoxia type, and inputting them into the relational network of the hypoxia induction model, and outputting the expected optimal treatment conditions and the expected apoptosis rate; it not only considers the differences in cell types themselves, but also combines different hypoxia types, enabling the model to more realistically simulate the hypoxic environment in vivo, and improving the accuracy of model prediction.

[0016] Furthermore, the method of the present invention reduces the prediction bias caused by differences in cell sensitivity and improves the prediction efficiency by labeling the relationship network with recommended tags based on the cosine similarity of the gene expression profiles of the target cell and known cells, retrieving the node with the highest similarity in the relationship network, assigning its cell induction parameters as recommended tags to the target cell, and inheriting the stable induction features of the highly similar nodes.

[0017] Furthermore, the method of the present invention achieves model updating and adaptive optimization by comparing the deviation between the measured apoptosis rate of target cells and the predicted apoptosis rate output by the model, and generating correction instructions based on the deviation, including adjusting the gradient concentration or adding more treatment times. When the difference in pretreatment time of target cell batches leads to pre-exposure of cells in the air, it will partially activate the oxidative stress pathway, thereby causing a shift in its sensitivity to subsequent cobalt chloride induction. In particular, the deviation between the measured apoptosis rate and the predicted value increases for acute hypoxia type. By adjusting the gradient concentration, the rapid cell death caused by excessive concentration is avoided, which affects the accurate transmission of hypoxia signals, and the model gradually approaches a more accurate hypoxia induction condition, thereby improving modeling efficiency and accuracy.

[0018] Furthermore, when the hypoxia type is expected to be chronic, the method of the present invention simulates long-term, gradual hypoxia stimulation by reducing the duration of a single treatment and correspondingly increasing the number of treatments. This more closely resembles the physiological and pathological state of chronic hypoxia, thereby offsetting the metabolic memory effect caused by previous exposure and enhancing the biological simulation capability of the model. Attached Figure Description

[0019] Figure 1 This is an overall flowchart of the method for establishing a hypoxia-induced model using cobalt chloride, as described in an embodiment of the present invention. Figure 2 This is a flowchart illustrating the process of determining the gradient concentration for a concentration-decreasing treatment in the method for establishing a hypoxia-induced model using cobalt chloride, as described in an embodiment of the present invention. Figure 3 This is a flowchart illustrating the method for determining the number of additional cyclic treatments in the method for establishing a hypoxia-induced model using cobalt chloride, as described in an embodiment of the present invention. Figure 4 This is a flowchart illustrating the labeling and recommendation process for target cells in a method for establishing a hypoxia-induced model using cobalt chloride, as described in an embodiment of the present invention. Detailed Implementation

[0020] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0021] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0022] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0023] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0024] Please see Figure 1 , Figure 2 , Figure 3 as well as Figure 4 The diagrams shown are, respectively, an overall flowchart, a flowchart for determining the gradient concentration of the concentration decrease treatment, a flowchart for determining the number of new cycle treatments, and a flowchart for adjusting the labeling and recommendation of target cells in the method for establishing a hypoxia-induced model using cobalt chloride according to an embodiment of the present invention. The present invention provides a method for establishing a hypoxia-induced model using cobalt chloride, comprising: Several types of experimental cells were sequentially exposed to cobalt chloride solutions with a fixed concentration gradient and the cell induction parameters of each treatment group were detected. The cell induction parameters included induction treatment time, concentration of cobalt chloride solution, apoptosis rate of experimental cells, gene expression profile of experimental cells, and duration of HIF-1α peak. A relational network for the hypoxia-induced model was established based on the cell induction parameters of each treatment result. Obtain the cell induction parameters of the target cells and the corresponding expected hypoxia type; The target cells and the corresponding expected hypoxia type are respectively input into the relational network of the hypoxia-induced model, and the predicted optimal treatment conditions and predicted apoptosis rate are respectively output, wherein, Recommended labels for the cell induction parameters of the target cells in the hypoxia-induced model relationship network are labeled based on the cosine similarity of the gene expression profiles of the target cells and known cells. The target cells were treated under the optimal treatment conditions described above to obtain the measured apoptosis rate. The correction instructions are determined based on the deviation between the measured apoptosis rate of the target cells and the predicted apoptosis rate output by the hypoxia-induced model. These instructions may include determining the gradient concentration of the concentration-decreasing treatment or determining the number of additional cycles. The relational network of the hypoxia-induced model is adjusted according to the correction instructions.

[0025] Specifically, the types of cells used in the experiments included human umbilical vein endothelial cells and mouse fibroblasts.

[0026] Specifically, gene expression profiles are obtained by transcribing the whole genome and detected by high-throughput RNA sequencing technology; the duration of HIF-1α peak is the period during which the HIF-1α protein level exceeds twice the baseline value and is detected by mass spectrometry or immunofluorescence.

[0027] Specifically, the relationship network of the hypoxia-induced model is adjusted according to the correction instruction to use new cell induction parameter data obtained from new experiments on target cells according to the correction instruction, and added as new nodes to the original relationship network. The edge weights are then recalculated and updated based on the similarity between the new nodes and all existing nodes.

[0028] In this embodiment, the cobalt chloride solution was prepared into a series of concentrations, with the difference between two adjacent concentrations being a fixed concentration step size of 100 μM, forming a concentration hierarchy from high to low, with concentration gradients of 500 μM, 400 μM, 300 μM, 200 μM, 100 μM, and 0 μM. Cells were exposed to these solutions sequentially in order from high to low concentration to simulate the gradual relief of hypoxia.

[0029] Specifically, cell induction parameters include induction treatment time, cobalt chloride solution concentration, apoptosis rate of experimental cells, gene expression profile of experimental cells, and duration of HIF-1α peak. The induction treatment time was 24 hours for each concentration; the apoptosis rate of the experimental cells was the percentage of apoptosis detected by flow cytometry.

[0030] In practice, the method of this invention uses the system described in this invention to treat several types of experimental cells sequentially with a cobalt chloride solution of fixed concentration step-decreasing gradient, and detects the cell induction parameters of each treatment group. Due to the increased prediction bias of apoptosis rate caused by the different sensitivity of different cell types to cobalt chloride, a relational network is established with cell induction parameters as nodes and the similarity between parameter groups as edge weights. This integrates multiple types of data and multiple hypoxia types to cover the characteristics of different cell types, reducing the prediction bias of apoptosis rate caused by differences in cell sensitivity. Since errors in experimental parameters can lead to different induction responses of the same cell type in different experiments, or because there are differences in gene expression profiles of cells at different passage stages or differentiation states (i.e., aging cells or cells at different cycle stages respond differently to hypoxic stimuli), the method calculates the target... The model uses cosine similarity of gene expression profiles between target cells and known cells to identify known cell types that are most similar to target cells in gene expression patterns. This allows for the recommendation of appropriate cell induction parameters. This similarity-based recommendation mechanism enhances the model's ability to reasonably predict hypoxia responses even in the absence of complete experimental data for target cells. Because the varying exposure time of cells to air before solution exposure during batch cell treatment introduces systematic errors between model predictions and measured values, the model calculates the deviation by comparing the measured apoptosis rate of target cells with the model's predicted apoptosis rate. Based on this deviation, it dynamically generates correction instructions, including adjusting the gradient concentration of the decreasing treatment or adding more cycles. This targeted compensation for variable perturbations introduced by differences in operational timing improves the model's accuracy in practical applications.

[0031] Specifically, the relational network of the hypoxia-induced model is established with cell induction parameters as nodes and the similarity of cell induction parameters of several groups of experimental cells as edge weights.

[0032] Specifically, each node represents a set of cell induction parameters, which are several of all cell induction parameters obtained after treating a single type of cell at a single concentration.

[0033] Specifically, the similarity of cell induction parameters for several groups of experimental cells is calculated using cosine similarity to determine the similarity between any two node parameter vectors, and this similarity value is used as the weight of the edge connecting these two nodes.

[0034] Specifically, the relationship network for establishing the hypoxia-induced model is a graph-based data model. Each node represents a complete set of cell induction parameters. For example, a node represents the parameter data generated by a certain type of cell under a specific concentration and duration of treatment. The edge between two nodes represents the similarity between the two sets of parameters. The similarity is obtained by calculating the cosine similarity between the basic gene expression profile of the target cell without cobalt chloride treatment and the baseline expression profile of all known cell types in the relationship network.

[0035] In practice, the method of the present invention obtains the cell induction parameters of the target cells and the corresponding expected hypoxia type, and inputs them into the relational network of the hypoxia induction model to output the expected optimal treatment conditions and the expected apoptosis rate, thereby improving the accuracy of induction modeling. It not only considers the differences in cell types themselves, but also combines different hypoxia types, enabling the model to more realistically simulate the hypoxic environment in vivo and improve the accuracy of model prediction.

[0036] Specifically, the anticipated hypoxia types include acute hypoxia with a peak duration of HIF-1α less than or equal to a preset duration and chronic hypoxia with a peak duration of HIF-1α greater than the preset duration.

[0037] Specifically, the deviation is the absolute value of the difference between the measured apoptosis rate and the predicted apoptosis rate.

[0038] Specifically, the general range of the preset duration is [14, 20h], and the preferred embodiment of the preset duration is 16h.

[0039] Specifically, if the expected hypoxia type is acute and the deviation is greater than or equal to the preset deviation, then the gradient concentration is reduced.

[0040] Specifically, the gradient concentration is determined based on the difference between the deviation and the preset deviation.

[0041] In practice, when the target cells are rat cardiomyocytes, if the difference between the deviation and the preset deviation exceeds 1%, the gradient concentration is reduced by 5 μM. For example, if the difference between the deviation and the preset deviation is 4%, and the treatment conditions for the target cells are an initial concentration of 500 μM and a gradient concentration of 100 μM, then the single-round gradient reduction treatments after reduction are 500 μM, 420 μM, 340 μM, 260 μM, 180 μM, and 100 μM respectively.

[0042] Specifically, if the hypoxia type is expected to be chronic and the deviation is greater than or equal to the preset deviation, the duration of a single treatment is reduced, and the number of additional gradient treatments is determined based on the duration of the single treatment.

[0043] Specifically, the processing time per cycle is determined based on the difference between the deviation and the preset deviation.

[0044] Specifically, the number of additional gradient treatments = original single induction treatment duration × initial concentration gradient group number ÷ adjusted single induction treatment duration - initial concentration gradient group number. For example, if the original single treatment duration is 24 hours, the initial concentration gradient group number is 6, and the adjusted single treatment duration is 18 hours, then the number of additional treatments = (24 × 6) / 18 - 6 = 2. The calculated number of additional treatments ensures that the total induction duration remains unchanged. At the same time, by shortening the single treatment duration and increasing the number of cycles, the gradual stress response of cells under chronic hypoxia is simulated more accurately, offsetting the metabolic memory effect caused by previous exposure, and further improving the model's accuracy in simulating chronic hypoxia.

[0045] In practice, when the target cells are human pancreatic cancer cells PANC-1, if the difference between the deviation and the preset deviation exceeds 1%, the processing time per session is reduced by 0.5 hours.

[0046] In practice, the method of the present invention uses recommended tags in the relationship network based on the cosine similarity of the gene expression profiles of the target cell and known cells. By retrieving the node with the highest similarity in the relationship network, the cell induction parameters of the node are used as recommended tags to give the target cell stable induction features inherited from the highly similar node. This reduces the prediction bias caused by differences in cell sensitivity and improves the prediction efficiency.

[0047] Specifically, recommended labels for annotating the cell induction parameters of the target cells in the hypoxia-induced model's relational network include: The cosine similarity is compared with a preset similarity. If the cosine similarity is greater than or equal to the preset similarity, then the expected optimal treatment conditions and the expected apoptosis rate are output. If the cosine similarity is less than the preset similarity, the cell induction parameters of the node with the maximum similarity are extracted as the recommended labels for the target cells of the hypoxia-induced model's relational network.

[0048] Specifically, the general range of preset similarity is [0.8, 0.88], and the preferred embodiment of preset similarity is 0.85.

[0049] In this embodiment, if the cosine similarity is greater than or equal to 0.85, the target cell is considered to be highly similar to such known cells, and the optimal treatment conditions and corresponding apoptosis rate corresponding to the known cell node are directly used as the recommended label for prediction. The apoptosis rate corresponding to the known cell node is the predicted apoptosis rate. If the cosine similarity is less than the preset similarity, the top 3 known cell nodes with the highest similarity are found, and the cell induction parameters of these nodes are weighted and averaged to generate a recommended label for the target cell.

[0050] In practice, the method described in this invention compares the deviation between the measured apoptosis rate of target cells and the predicted apoptosis rate output by the model, and generates correction instructions based on the deviation, including adjusting the gradient concentration or adding more treatment times, thereby achieving model updates and adaptive optimization. When the pretreatment time difference of target cell batches leads to pre-exposure of cells in the air, it will partially activate the oxidative stress pathway, which will cause a shift in its sensitivity to subsequent cobalt chloride induction. In particular, the deviation between the measured apoptosis rate and the predicted value will be significantly increased for acute hypoxia type. By adjusting the gradient concentration, the rapid cell death caused by excessive concentration is avoided, which affects the accurate transmission of hypoxia signals, and the model gradually approaches a more accurate hypoxia induction condition, thereby improving modeling efficiency and accuracy.

[0051] Specifically, the gradient concentration is the absolute value of the concentration difference between two adjacent cobalt chloride solutions when the solutions are exposed to a cobalt chloride solution with a fixed concentration step size decreasing.

[0052] In practice, when the hypoxia type is expected to be chronic, the method of the present invention simulates long-term, gradual hypoxia stimulation by reducing the duration of a single treatment and correspondingly increasing the number of treatments. This more closely resembles the physiological and pathological state of chronic hypoxia, thereby offsetting the metabolic memory effect caused by previous exposure and enhancing the biological simulation capability of the model.

[0053] Specifically, the relational network of the hypoxia-induced model is adjusted according to the correction instruction. For example, if the correction of acute hypoxia type requires reducing the gradient concentration, the correction instruction is to reduce the gradient concentration. The target cells are reprocessed according to the correction instruction, and the corrected experimental data is recorded as new nodes in the relational network, and the edge weights are updated.

[0054] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A method for establishing a hypoxia-induced model using cobalt chloride, characterized in that, include: Several types of experimental cells were sequentially exposed to cobalt chloride solutions with a fixed concentration gradient and the cell induction parameters of each treatment group were detected. The cell induction parameters included induction treatment time, concentration of cobalt chloride solution, apoptosis rate of experimental cells, gene expression profile of experimental cells, and duration of HIF-1α peak. A relational network for the hypoxia-induced model was established based on the cell induction parameters of each treatment result. Obtain the cell induction parameters of the target cells and the corresponding expected hypoxia type; The target cells and the corresponding expected hypoxia type are respectively input into the relational network of the hypoxia-induced model, and the predicted optimal treatment conditions and predicted apoptosis rate are respectively output, wherein, Recommended labels for the cell induction parameters of the target cells in the hypoxia-induced model relationship network are labeled based on the cosine similarity of the gene expression profiles of the target cells and known cells. The target cells were treated under the optimal treatment conditions described above to obtain the measured apoptosis rate. The correction instructions are determined based on the deviation between the measured apoptosis rate of the target cells and the predicted apoptosis rate output by the hypoxia-induced model. These instructions may include determining the gradient concentration of the concentration-decreasing treatment or determining the number of additional cycles. The relational network of the hypoxia-induced model is adjusted according to the correction instructions.

2. The method for establishing a hypoxia-induced model using cobalt chloride according to claim 1, characterized in that, The hypoxia-induced model's relational network is established using cell induction parameters as nodes and the similarity of cell induction parameters among several groups of experimental cells as edge weights.

3. The method for establishing a hypoxia-induced model using cobalt chloride according to claim 2, characterized in that, The anticipated hypoxia types include acute hypoxia with a peak duration of HIF-1α less than or equal to a preset duration and chronic hypoxia with a peak duration of HIF-1α greater than the preset duration.

4. The method for establishing a hypoxia-induced model using cobalt chloride according to claim 3, characterized in that, The deviation is the absolute value of the difference between the measured apoptosis rate and the predicted apoptosis rate.

5. The method for establishing a hypoxia-induced model using cobalt chloride according to claim 4, characterized in that, If the expected hypoxia type is acute and the deviation is greater than or equal to the preset deviation, then the gradient concentration is reduced.

6. The method for establishing a hypoxia-induced model using cobalt chloride according to claim 5, characterized in that, The gradient concentration is determined based on the difference between the deviation and the preset deviation.

7. The method for establishing a hypoxia-induced model using cobalt chloride according to claim 6, characterized in that, If the expected hypoxia type is chronic and the deviation is greater than or equal to the preset deviation, then the duration of a single treatment is reduced, and the number of additional gradient treatments is determined based on the duration of the single treatment.

8. The method for establishing a hypoxia-induced model using cobalt chloride according to claim 7, characterized in that, The processing time per cycle is determined based on the difference between the deviation and the preset deviation.

9. The method for establishing a hypoxia-induced model using cobalt chloride according to claim 7, characterized in that, Recommended labels for the cell induction parameters of the target cells in the hypoxia-induced model's relational network include: The cosine similarity is compared with a preset similarity. If the cosine similarity is greater than or equal to the preset similarity, then the expected optimal treatment conditions and the expected apoptosis rate are output. If the cosine similarity is less than the preset similarity, the cell induction parameters of the node with the maximum similarity are extracted as the recommended labels for the target cells of the hypoxia-induced model's relational network.

10. The method for establishing a hypoxia-induced model using cobalt chloride according to claim 8, characterized in that, The gradient concentration is the absolute value of the concentration difference between two adjacent cobalt chloride solutions when the solutions are exposed to a cobalt chloride solution with a fixed concentration step size decreasing.

Images