Breast cancer typing prediction method and system based on LncRNA and immune cell characteristics

By constructing a breast cancer classification model based on LncRNA and immune cell characteristics, combining metabolic pathway analysis and convolutional neural network, the shortcomings of existing breast cancer classification methods in predicting immunotherapy effects are solved, and high accuracy and convenience are achieved Predictions of sexual breast cancer typing lower the threshold for typing.

CN118942542AActive Publication Date: 2024-11-12SUN YAT SEN MEMORIAL HOSPITAL SUN YAT SEN UNIV
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202410918645.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-10
Publication Date
2024-11-12
Estimated Expiration
2044-07-10

AI Technical Summary

Technical Problem

The existing breast cancer typing methods are insufficient in predicting the effect of immunotherapy, and the traditional typing methods rely on gene sequencing, with high thresholds, and there are problems with inconsistent results of immunohistochemistry methods.

Method used

Using a typing prediction method based on LncRNA and immune cell characteristics, the clustering model of LncRNA and immune cell characteristics was constructed, combined with metabolic pathway analysis, breast cancer typing was generated, and the convolutional neural network model was used to perform typing prediction based on pathological slice images. , avoid gene sequencing and improve convenience and accuracy.

Benefits of technology

It improves the accuracy and sensitivity of the prediction of the efficacy of breast cancer immunotherapy, lowers the threshold for clinical application of typing, provides a more comprehensive basis for decision-making of treatment plans, and avoids the dependence of gene sequencing.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118942542A_ABST
    Figure CN118942542A_ABST
Patent Text Reader

Abstract

The invention provides a breast cancer typing prediction method and system based on LncRNA and immune cell characteristics. The method comprises the following steps: acquiring a first RNA sequencing data set and a second RNA sequencing data set; the first RNA sequencing data set is input into a preset LncRNA clustering model, so that all the RNA sequencing data in the first RNA sequencing data set are clustered into different LncRNA subtypes through the LncRNA clustering model; inputting the first RNA sequencing data set into a preset immune cell characteristic clustering model, so that the immune cell characteristic clustering model clusters each piece of RNA sequencing data in the first RNA sequencing data set into different immune subtypes; combining each LncRNA subtype and each immune subtype to obtain a plurality of different breast cancer types; and inputting the pathological section image of the to-be-predicted patient into the preset typing prediction model, so that the typing prediction model generates the corresponding breast cancer typing prediction result, and the accuracy of curative effect prediction and prognosis prediction of immunotherapy is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a breast cancer typing prediction method and system based on LncRNA and immune cell characteristics. Background Art

[0002] Breast cancer is a common malignant tumor, and its clinical classification is crucial for developing effective treatment plans and predicting patient prognosis. The rapid development of immunotherapy has brought new possibilities for breast cancer treatment, especially the application of immune checkpoint inhibitors (ICI). Immunotherapy activates the patient's own immune system to resist tumor cells, thereby causing a lasting anti-tumor response. Currently, breast cancer can be divided into multiple subtypes based on the different expression of biomarkers, and these subtypes may respond differently to immunotherapy. The following are some common clinical classifications of breast cancer and the corresponding classification methods or models:

[0003] (1) Hormone receptor status (ER, PR, HER2): Breast cancer is often classified according to the expression of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2). This classification is not only instructive for determining treatment plans, but can also be related to the efficacy of immunotherapy.

[0004] (2) Immune cell infiltration: The effectiveness of immunotherapy is closely related to the infiltration of immune cells around the tumor. Some typing methods focus on the type and number of immune cells, such as lymphocytes and plasma cells, to help predict the patient's response to immunotherapy.

[0005] (3) PD-L1 expression: Programmed cell death ligand 1 (PD-L1) is an immune checkpoint molecule whose overexpression may affect the effect of immunotherapy. Some classification methods focus on evaluating the expression level of PD-L1 on the surface of tumor cells or immune cells as a biomarker for predicting the response to immunotherapy.

[0006] (4) Neoantigen load: Neoantigens are antigens expressed on the surface of tumor cells due to mutations. Some classification models focus on the neoantigen load of tumors because a high neoantigen load may be associated with a stronger immune response and better prognosis.

[0007] However, none of the existing biomarkers have achieved satisfactory accuracy. PD-L1 expression is a stratification factor or patient inclusion criterion for clinical trials, and is also an indication for the use of ICIs for certain cancer types in treatment guidelines. However, clinical trials have also found the efficacy of ICIs in patients without PD-L1 expression. In addition, only a small number of patients respond to these therapies. This shows that the current classification methods are not accurate enough for predicting patient prognosis. At the same time, most of the existing breast cancer classification diagnoses are based on immunohistochemistry test results. The advantages of immunohistochemistry are good specificity, high sensitivity, simplicity and speed, relatively low cost, and good repeatability. The disadvantages are that they are not comprehensive enough, and the results between different laboratories may be inconsistent. Therefore, a classification prediction model is needed to combine artificial intelligence technology to predict breast cancer classification for patients, assist medical workers in breast cancer classification diagnosis and treatment plan formulation, and overcome the limitations of immunohistochemistry methods to a certain extent. Summary of the invention

[0008] In response to the above technical problems, the present invention provides a breast cancer typing prediction method and system based on LncRNA and immune cell characteristics. Since the obtained breast cancer typing prediction results indicate the patient's suitability for immunotherapy, the present invention improves the accuracy of immunotherapy efficacy prediction and prognosis prediction for patients, and provides data support for the formulation of tumor treatment plans.

[0009] In a first aspect, the present invention provides a method for predicting breast cancer typing based on LncRNA and immune cell characteristics, comprising:

[0010] Obtaining a first RNA sequencing data set of several historical breast cancer patients and a second RNA sequencing data set of several patients receiving immunotherapy;

[0011] Inputting the first RNA sequencing data set into a preset LncRNA clustering model, so that the LncRNA clustering model clusters each RNA sequencing data in the first RNA sequencing data set into different LncRNA subtypes, wherein the LncRNA clustering model is constructed based on the feature set of the second RNA sequencing data set;

[0012] Inputting the first RNA sequencing data set into a preset immune cell feature clustering model, so that the immune cell feature clustering model clusters each RNA sequencing data in the first RNA sequencing data set into different immune subtypes, wherein the immune cell feature clustering model is constructed based on each immune cell in the second RNA sequencing data set and the corresponding ssGSEA score;

[0013] Combining the various LncRNA subtypes and the various immune subtypes, several different breast cancer classifications are obtained;

[0014] The pathological section image of the patient to be predicted is input into a preset typing prediction model so that the typing prediction model generates a corresponding breast cancer typing prediction result, wherein the breast cancer typing prediction result is one of the several different breast cancer typings, and the typing prediction model is constructed based on the pathological section data of historical breast cancer patients and a pre-trained convolutional neural network model.

[0015] The embodiment of the present invention provides a method for predicting breast cancer typing based on LncRNA and immune cell characteristics. The embodiment first constructs a LncRNA clustering model and an immune cell characteristic clustering model according to the second RNA sequencing data set, which is used to mine the LncRNA subtypes and immune subtypes in the first RNA sequencing data set. Then, in combination with each LncRNA subtype and each immune subtype, several breast cancer typings are constructed. The breast cancer typing constructed in this embodiment fully considers the effects of long non-coding RNA and immune cells on the efficacy of tumor immunotherapy, and reconstructs a breast cancer typing that is more sensitive to immunotherapy. The breast cancer typing has better sensitivity and accuracy than traditional indicators for predicting the efficacy of immunotherapy. Further, in accordance with the breast cancer typing, the embodiment also provides a typing prediction model based on pathological images, which avoids the gene sequencing required for traditional tumor typing, and can perform typing prediction only with the help of clinical pathological sections, thereby improving the convenience and accuracy of breast cancer typing prediction, and lowering the threshold for clinical application of the breast cancer typing. The obtained breast cancer typing prediction results can provide a decision-making basis for medical workers to formulate cancer treatment plans.

[0016] In a possible implementation, constructing the LncRNA clustering model according to the feature set of the second RNA sequencing data set includes:

[0017] Transcribe the corresponding LncRNA expression dataset from the second RNA sequencing dataset;

[0018] According to the influence factor of each feature in the LncRNA expression data set on immunotherapy, a number of LncRNA features are extracted from the LncRNA expression data set, and the LncRNA feature set is formed by combining them;

[0019] A lncRNA clustering model is constructed according to the lncRNA feature set.

[0020] The embodiment of the present invention provides a method for constructing a LncRNA clustering model. Since the patients in the second RNA sequencing data set are patients receiving immunotherapy, features related to immunotherapy can be extracted from the second RNA sequencing data set, and then these features can be used to construct a clustering model. Specifically, the second RNA sequencing data set is first transcribed into a LncRNA expression data set, and several LncRNAs that affect the difference in efficacy are screened from the LncRNA expression data set, that is, the several LncRNA features, which are combined to form a LncRNA feature set, and finally a LncRNA clustering model is constructed based on these feature sets. This embodiment constructs a LncRNA clustering model through the second RNA sequencing data set, which improves the accuracy of subsequent clustering of the model. At the same time, the constructed LncRNA clustering model belongs to an unsupervised machine learning model, which does not require human intervention and adjustment, and can automatically cluster the input data set to generate multiple different LncRNA subtypes, thereby improving the efficiency of LncRNA subtype generation.

[0021] In a possible implementation, constructing the immune cell feature clustering model according to each immune cell in the second RNA sequencing data set and the corresponding ssGSEA score includes:

[0022] Inputting the second RNA sequencing data set into a preset bioinformatics analysis software, so that the bioinformatics analysis software determines several types of immune cells and ssGSEA scores corresponding to various immune cells through ssGSEA analysis;

[0023] An immune cell feature clustering model was constructed based on the various immune cells and the corresponding ssGSEA scores.

[0024] The embodiment of the present invention provides a method for constructing an immune cell feature clustering model, and the model is also constructed based on a second RNA sequencing data set of patients who have received immunotherapy. First, the second RNA sequencing data set is subjected to ssGSEA analysis through a preset bioinformatics analysis software, and several types of immune cells in the second RNA sequencing data set and the ssGSEA scores corresponding to various immune cells are analyzed and identified. The ssGSEA scores of immune cells also reveal the effects of the immune cells on immunotherapy, and therefore, an immune cell feature clustering model is constructed based on the various immune cells and the corresponding ssGSEA scores. This embodiment constructs an immune cell feature clustering model from the perspective of immune cells based on the second RNA sequencing data set, and combines the ssGSEA scores of each immune cell to improve the accuracy and rationality of subsequent clustering of the model.

[0025] Furthermore, the combination of the various LncRNA subtypes and the various immune subtypes can obtain several different breast cancer typings, including:

[0026] Combining the LncRNA subtypes and the immune subtypes in pairs to generate a first number of LncRNA-immune subtypes;

[0027] GSEA metabolic pathway analysis was used to perform pathway enrichment scoring on each of the LncRNA-immune subtypes to obtain the corresponding pathway enrichment score;

[0028] According to each of the pathway enrichment scores, the metabolic pathway enrichment characteristics of each of the LncRNA-immune subtypes are determined, and the LncRNA-immune subtypes with the same metabolic pathway enrichment characteristics are combined to obtain a second number of breast cancer typing, wherein the second number is less than or equal to the first number.

[0029] In an embodiment of the present invention, after obtaining each LncRNA subtype and each immune subtype, they are further merged into a two-dimensional index to generate a first number of LncRNA-immune subtypes. And each LncRNA-immune subtype may have a LncRNA-immune subtype with a similar therapeutic effect on immunotherapy. In order to improve the accuracy and efficiency of subsequent classification predictions, it is necessary to merge the LncRNA-immune subtypes with similar therapeutic effects on immunotherapy into the same type. Therefore, this embodiment uses the metabolic pathway enrichment feature as an evaluation index for each LncRNA-immune subtype, and merges the LncRNA-immune subtypes with the same metabolic pathway enrichment feature into the same breast cancer typing, and finally obtains the second number of breast cancer typing. The breast cancer typing obtained in this embodiment significantly distinguishes different immunotherapy efficacy and prognosis survival time, and reveals the metabolic connotation of each breast cancer typing, providing data support for the formulation of subsequent tumor treatment plans.

[0030] In a possible implementation, the classification prediction model generates a corresponding breast cancer classification prediction result based on the pathological section image of the patient to be predicted, including:

[0031] Performing image segmentation on the pathological slice image according to tissue regions in the pathological slice image to obtain a plurality of image blocks;

[0032] Extracting a number of feature vectors from each of the image blocks using a convolutional neural network model with pre-trained parameters;

[0033] According to the feature vector, the image blocks are sorted using a multi-instance learning algorithm based on an attention mechanism, and a corresponding attention score is assigned to each image block;

[0034] Aggregating image information of each image block according to the attention score to generate aggregate information;

[0035] According to the aggregated information, through a multi-branch network architecture and multi-task objectives, the convolutional neural network model is used to predict the breast cancer typing of the pathological slice image to generate a breast cancer typing prediction result for the patient to be predicted.

[0036] The embodiment of the present invention provides a method for predicting breast cancer typing based on pathological slice images, using digital high-resolution histological slices as input to a typing prediction model. For each pathological slice image, the tissue content is automatically segmented and evenly divided into thousands to tens of thousands of regions as small image blocks. Then, the corresponding feature vectors are extracted from each image block through a convolutional neural network model with pre-trained parameters. According to the feature vectors, each image block is sorted, and according to the relative importance of the image block, that is, the attention score, the image information of each image block is aggregated to obtain the aggregated information of the entire pathological slice image. Finally, breast cancer typing is predicted based on the aggregated information, and a breast cancer typing prediction result for the patient to be predicted is generated, which realizes breast cancer typing diagnosis based on digital pathological slice features, avoids the gene sequencing required for traditional tumor typing, and can perform typing prediction with the help of clinical pathological slices only, thereby improving the convenience and accuracy of breast cancer typing prediction.

[0037] In a second aspect, accordingly, the present invention provides a breast cancer typing prediction system based on LncRNA and immune cell characteristics, comprising an acquisition module, a LncRNA subtype generation module, an immune subtype generation module, a breast cancer typing generation module, and a prediction module;

[0038] Wherein, the acquisition module is used to acquire a first RNA sequencing data set of several historical breast cancer patients and a second RNA sequencing data set of several patients receiving immunotherapy;

[0039] The LncRNA subtype generation module is used to input the first RNA sequencing data set into a preset LncRNA clustering model, so that the LncRNA clustering model clusters each RNA sequencing data in the first RNA sequencing data set into different LncRNA subtypes, wherein the LncRNA clustering model is constructed based on the feature set of the second RNA sequencing data set;

[0040] The immune subtype generation module is used to input the first RNA sequencing data set into a preset immune cell feature clustering model, so that the immune cell feature clustering model clusters each RNA sequencing data in the first RNA sequencing data set into different immune subtypes, wherein the immune cell feature clustering model is constructed based on each immune cell in the second RNA sequencing data set and the corresponding ssGSEA score;

[0041] The breast cancer typing generation module is used to combine the various LncRNA subtypes and the various immune subtypes to obtain several different breast cancer typings;

[0042] The prediction module is used to input the pathological section image of the patient to be predicted into a preset typing prediction model so that the typing prediction model generates a corresponding breast cancer typing prediction result, wherein the breast cancer typing prediction result is one of the several different breast cancer typings, and the typing prediction model is constructed based on the pathological section data of historical breast cancer patients and a pre-trained convolutional neural network model.

[0043] In a possible implementation, constructing the LncRNA clustering model according to the feature set of the second RNA sequencing data set includes:

[0044] Transcribe the corresponding LncRNA expression dataset from the second RNA sequencing dataset;

[0045] According to the influence factor of each feature in the LncRNA expression data set on immunotherapy, a number of LncRNA features are extracted from the LncRNA expression data set, and the LncRNA feature set is formed by combining them;

[0046] A lncRNA clustering model is constructed according to the lncRNA feature set.

[0047] In a possible implementation, constructing the immune cell feature clustering model according to each immune cell in the second RNA sequencing data set and the corresponding ssGSEA score includes:

[0048] Inputting the second RNA sequencing data set into a preset bioinformatics analysis software, so that the bioinformatics analysis software determines several types of immune cells and ssGSEA scores corresponding to various immune cells through ssGSEA analysis;

[0049] An immune cell feature clustering model was constructed based on the various immune cells and the corresponding ssGSEA scores.

[0050] Furthermore, the breast cancer typing generation module combines the various LncRNA subtypes and the various immune subtypes to obtain several different breast cancer typings, including:

[0051] Combining the LncRNA subtypes and the immune subtypes in pairs to generate a first number of LncRNA-immune subtypes;

[0052] GSEA metabolic pathway analysis was used to perform pathway enrichment scoring on each of the LncRNA-immune subtypes to obtain the corresponding pathway enrichment score;

[0053] According to each of the pathway enrichment scores, the metabolic pathway enrichment characteristics of each of the LncRNA-immune subtypes are determined, and the LncRNA-immune subtypes with the same metabolic pathway enrichment characteristics are combined to obtain a second number of breast cancer typing, wherein the second number is less than or equal to the first number.

[0054] In a possible implementation, the prediction module includes an image segmentation unit, a feature extraction unit, a sorting unit, an information aggregation unit, and a prediction unit;

[0055] Wherein, the image segmentation unit is used to perform image segmentation on the pathological slice image according to the tissue area in the pathological slice image to obtain a plurality of image blocks;

[0056] The feature extraction unit is used to extract a number of feature vectors from each image block using a convolutional neural network model with pre-trained parameters;

[0057] The sorting unit is used to sort the image blocks according to the feature vectors using a multi-instance learning algorithm based on an attention mechanism, and assign a corresponding attention score to each image block;

[0058] The information aggregation unit is used to aggregate the image information of each image block according to the attention score to generate aggregated information;

[0059] The prediction unit is used to predict the breast cancer typing of the pathological slice image using the convolutional neural network model based on the aggregated information, through a multi-branch network architecture and multi-task objectives, to generate a breast cancer typing prediction result for the patient to be predicted. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 : A schematic diagram of an embodiment of a method for predicting breast cancer typing based on LncRNA and immune cell characteristics provided by the present invention.

[0061] Figure 2 : A schematic diagram of the LncRNA subtype clustering results of an embodiment of a breast cancer typing prediction method based on LncRNA and immune cell characteristics provided by the present invention.

[0062] Figure 3 : A schematic diagram of the prognosis survival curves of different LncRNA subtypes of an embodiment of a breast cancer typing prediction method based on LncRNA and immune cell characteristics provided by the present invention.

[0063] Figure 4: A schematic diagram of immune subtype clustering results of an embodiment of a breast cancer typing prediction method based on LncRNA and immune cell characteristics provided by the present invention.

[0064] Figure 5 : A schematic diagram of the prognosis survival curves of different immune subtypes of an embodiment of a breast cancer typing prediction method based on LncRNA and immune cell characteristics provided by the present invention.

[0065] Figure 6 : A schematic diagram of metabolic pathway enrichment characteristics of four breast cancer typings in an embodiment of a breast cancer typing prediction method based on LncRNA and immune cell characteristics provided by the present invention.

[0066] Figure 7 : A schematic diagram of a process for constructing breast cancer typing according to an embodiment of a method for predicting breast cancer typing based on LncRNA and immune cell characteristics provided by the present invention.

[0067] Figure 8 : A schematic diagram comparing the overall survival of patients with four breast cancer typing undergoing immunotherapy in an embodiment of a breast cancer typing prediction method based on LncRNA and immune cell characteristics provided by the present invention.

[0068] Fig. 9 : A schematic diagram of a process flow of breast cancer typing prediction according to an embodiment of a method for predicting breast cancer typing based on LncRNA and immune cell characteristics provided by the present invention.

[0069] Fig.10 : A schematic diagram of the construction process of a typing prediction model of an embodiment of a breast cancer typing prediction method based on LncRNA and immune cell characteristics provided by the present invention.

[0070] Fig.11 : A structural schematic diagram of an embodiment of a breast cancer typing prediction system based on LncRNA and immune cell characteristics provided by the present invention.

[0071] Fig.12 : A schematic diagram of the structure of a prediction module of an embodiment of a breast cancer typing prediction system based on LncRNA and immune cell characteristics provided by the present invention. DETAILED DESCRIPTION

[0072] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions 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. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0073] It should be noted that the step numbers in the text are only for the convenience of explanation of the specific embodiments and do not serve to limit the order in which the steps are executed. In the description of this application, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features.

[0074] Throughout this specification, the full name of LncRNA described in the specification is long non-coding RNA (Long non-coding RNA, lncRNA), which is a non-coding RNA with a length greater than 200 nucleotides. Studies have shown that lncRNA plays an important role in many life activities such as dosage compensation effect, epigenetic regulation, cell cycle regulation and cell differentiation regulation. The expression or functional abnormality of long-chain non-coding RNA is closely related to the occurrence of human diseases, including cancer, degenerative neurological diseases and many other major diseases that seriously endanger human health, which are specifically manifested in the abnormality of long-chain non-coding RNA in sequence and spatial structure, abnormal expression level, abnormal interaction with binding proteins, etc.

[0075] Embodiment 1:

[0076] like Figure 1 As shown, Example 1 provides a breast cancer typing prediction method based on LncRNA and immune cell characteristics, including steps S1-S5:

[0077] Step S1, obtaining a first RNA sequencing data set of several historical breast cancer patients and a second RNA sequencing data set of several patients receiving immunotherapy;

[0078] Step S2, inputting the first RNA sequencing data set into a preset LncRNA clustering model, so that the LncRNA clustering model clusters each RNA sequencing data in the first RNA sequencing data set into different LncRNA subtypes, wherein the LncRNA clustering model is constructed based on the feature set of the second RNA sequencing data set;

[0079] Step S3, inputting the first RNA sequencing data set into a preset immune cell feature clustering model, so that the immune cell feature clustering model clusters each RNA sequencing data in the first RNA sequencing data set into different immune subtypes, wherein the immune cell feature clustering model is constructed based on each immune cell in the second RNA sequencing data set and the corresponding ssGSEA score;

[0080] Step S4, combining the various LncRNA subtypes and the various immune subtypes to obtain several different breast cancer typings;

[0081] Step S5: input the pathological section image of the patient to be predicted into a preset typing prediction model, so that the typing prediction model generates a corresponding breast cancer typing prediction result, wherein the breast cancer typing prediction result is one of the several different breast cancer typings, and the typing prediction model is constructed based on the pathological section data of historical breast cancer patients and a pre-trained convolutional neural network model.

[0082] The embodiment of the present invention provides a method for predicting breast cancer typing based on LncRNA and immune cell characteristics. The embodiment first constructs a LncRNA clustering model and an immune cell characteristic clustering model according to the second RNA sequencing data set, which is used to mine the LncRNA subtypes and immune subtypes in the first RNA sequencing data set. Then, in combination with each LncRNA subtype and each immune subtype, several breast cancer typings are constructed. The breast cancer typing constructed in this embodiment fully considers the effects of long non-coding RNA and immune cells on the efficacy of tumor immunotherapy, and reconstructs a breast cancer typing that is more sensitive to immunotherapy. The breast cancer typing has better sensitivity and accuracy than traditional indicators for predicting the efficacy of immunotherapy. Further, in accordance with the breast cancer typing, the embodiment also provides a typing prediction model based on pathological images, which avoids the gene sequencing required for traditional tumor typing, and can perform typing prediction only with the help of clinical pathological sections, thereby improving the convenience and accuracy of breast cancer typing prediction, and lowering the threshold for clinical application of the breast cancer typing. The obtained breast cancer typing prediction results can provide a decision-making basis for medical workers to formulate cancer treatment plans.

[0083] In a preferred embodiment, the first RNA sequencing data set of the several historical breast cancer patients in step S1 comes from the RNA sequencing data of 925 breast cancer patients in The Cancer Genome Atlas (TCGA), and the second RNA sequencing data set of the several patients receiving immunotherapy comes from Sun Yat-sen Memorial Hospital (SYSMH), Sun Yat-sen Memory Hospital, specifically, the RNA sequencing data of 83 breast cancer patients (7 other patients received chemotherapy combined with immunotherapy) who visited Sun Yat-sen Memorial Hospital, Sun Yat-sen University from September 2019 to February 2022.

[0084] In a possible implementation, in step S2, constructing the LncRNA clustering model according to the feature set of the second RNA sequencing data set includes:

[0085] Transcribe the corresponding LncRNA expression dataset from the second RNA sequencing dataset;

[0086] According to the influence factor of each feature in the LncRNA expression data set on immunotherapy, a number of LncRNA features are extracted from the LncRNA expression data set, and the LncRNA feature set is formed by combining them;

[0087] A lncRNA clustering model is constructed according to the lncRNA feature set.

[0088] In a preferred embodiment, 198 LncRNAs that affect the difference in efficacy were screened out from the transcriptome sequencing LncRNA expression data of 7 breast cancer patients who received chemotherapy combined with immunotherapy at Sun Yat-sen Memorial Hospital of Sun Yat-sen University (responsive vs. non-responsive: 3 vs. 4), and a LncRNA clustering model based on deep learning was constructed based on the 198 LncRNAs. Furthermore, the LncRNA clustering model was used to perform unsupervised subtype clustering of 925 breast cancer samples of the first RNA sequencing data set to construct LncRNA-based breast cancer immune subtypes. In order to ensure the stability of the classification, we performed 1000 iterations and a resampling rate of 80%, and used the cumulative distribution function curve to determine the number of subtypes. The results showed that there were 369, 334, and 222 patients in 3 different subtypes (LncRNAcluster 1, LncRNA-cluster 2, and LncRNA-cluster 3), respectively, and the clustering results were as follows: Figure 2 shown.

[0089] Figure 3 Figure 2 is the prognostic survival curve of patients with three subtypes of immunotherapy-related LncRNA. Survival comparison showed significant differences among the three clusters (log-rank test, P<0.001). Patients in LncRNA-cluster1 had the best overall survival, while patients in LncRNA-cluster3 had the worst prognosis.

[0090] The embodiment of the present invention provides a method for constructing a LncRNA clustering model. Since the patients in the second RNA sequencing data set are patients receiving immunotherapy, features related to immunotherapy can be extracted from the second RNA sequencing data set, and then these features can be used to construct a clustering model. Specifically, the second RNA sequencing data set is first transcribed into a LncRNA expression data set, and several LncRNAs that affect the difference in efficacy are screened from the LncRNA expression data set, that is, the several LncRNA features, which are combined to form a LncRNA feature set, and finally a LncRNA clustering model is constructed based on these feature sets. This embodiment constructs a LncRNA clustering model through the second RNA sequencing data set, which improves the accuracy of subsequent clustering of the model. At the same time, the constructed LncRNA clustering model belongs to an unsupervised machine learning model, which does not require human intervention and adjustment, and can automatically cluster the input data set to generate multiple different LncRNA subtypes, thereby improving the efficiency of LncRNA subtype generation.

[0091] In a possible implementation, in step S3, constructing the immune cell feature clustering model according to each immune cell in the second RNA sequencing data set and the corresponding ssGSEA score includes:

[0092] Inputting the second RNA sequencing data set into a preset bioinformatics analysis software, so that the bioinformatics analysis software determines several types of immune cells and ssGSEA scores corresponding to various immune cells through ssGSEA analysis;

[0093] An immune cell feature clustering model was constructed based on the various immune cells and the corresponding ssGSEA scores.

[0094] In a preferred embodiment, the second RNA sequencing dataset is used to determine the infiltration levels of different immune cell populations by using ssGSEA in the Gene Set Variation Analysis of the RBioConductor software package. A total of 28 types of immune cells are identified through ssGSEA analysis, and then the immune cell feature clustering model is constructed based on the ssGSEA scores of each immune cell type. The 925 breast cancer samples in the first RNA sequencing dataset are subjected to unsupervised subtype clustering using the immune cell feature clustering model to construct breast cancer immune subtypes based on immune cell features. In order to ensure the stability of the classification, we performed 1000 iterations and a resampling rate of 80%, and used the cumulative distribution function curve to determine the number of subtypes. The results are as follows. Figure 4 As shown, the tumor samples were mainly divided into two subtypes.

[0095] There were 525 and 400 patients in two different subtypes (Immune cell cluster 1 and Immune cell cluster 2), respectively. Figure 5 The following are the prognostic survival curves of patients with two immune subtypes. Figure 5 It can be found that patients with immune cell subtype 1 have better survival than patients with immune cell subtype 2 (log-rank test, P<0.001). This suggests that immune cell subtype 1 and immune cell subtype 2 may be at the highest and lowest states of immunity.

[0096] The embodiment of the present invention provides a method for constructing an immune cell feature clustering model, and the model is also constructed based on a second RNA sequencing data set of patients who have received immunotherapy. First, the second RNA sequencing data set is subjected to ssGSEA analysis through a preset bioinformatics analysis software, and several types of immune cells in the second RNA sequencing data set and the ssGSEA scores corresponding to various immune cells are analyzed and identified. The ssGSEA scores of immune cells also reveal the effects of the immune cells on immunotherapy, and therefore, an immune cell feature clustering model is constructed based on the various immune cells and the corresponding ssGSEA scores. This embodiment constructs an immune cell feature clustering model from the perspective of immune cells based on the second RNA sequencing data set, and combines the ssGSEA scores of each immune cell to improve the accuracy and rationality of subsequent clustering of the model.

[0097] Furthermore, in step S4, the various LncRNA subtypes and the various immune subtypes are combined to obtain several different breast cancer typings, including:

[0098] Combining the LncRNA subtypes and the immune subtypes in pairs to generate a first number of LncRNA-immune subtypes;

[0099] GSEA metabolic pathway analysis was used to perform pathway enrichment scoring on each of the LncRNA-immune subtypes to obtain the corresponding pathway enrichment score;

[0100] According to each of the pathway enrichment scores, the metabolic pathway enrichment characteristics of each of the LncRNA-immune subtypes are determined, and the LncRNA-immune subtypes with the same metabolic pathway enrichment characteristics are combined to obtain a second number of breast cancer typing, wherein the second number is less than or equal to the first number.

[0101] In a preferred embodiment, the three LncRNA subtypes obtained in step S2 and the two immune subtypes obtained in step S3 are combined to obtain six combinations. By performing GSEA metabolic pathway analysis on these six combinations, the metabolic pathway enrichment characteristics are determined and named according to their characteristics, and finally four breast cancer typings are obtained, namely: high fatty acid metabolism type, high amino acid metabolism type, high glucose metabolism type, and high folate metabolism type. Among them, ‘LncRNA-cluster 1&Immune cellcluster 1’ was determined as high fatty acid metabolism type (High-FA), ‘LncRNA-cluster 2&Immune cellcluster 1’ or ‘LncRNA-cluster 1&Immune cell cluster 2’ was determined as high amino acid metabolism type (High-AA), ‘LncRNA-cluster 2&Immune cell cluster 2’ or ‘LncRNA-cluster3&Immunecell cluster 1’ was determined as high glucose metabolism type (High-Glu), and ‘LncRNA-cluster 3&Immune cellcluster 2’ was determined as high folate metabolism type (High-Folate). The metabolic pathway enrichment features of the four breast cancer types are as follows: Figure 6 shown. Figure 7 This is a schematic diagram of the process of constructing a new breast cancer classification based on the first RNA sequencing data set and the second RNA sequencing data set in this embodiment.

[0102] In addition, the overall survival (OS) of immunotherapy was significantly different among the four groups. Among them, the OS benefited the most from immunotherapy in the high fatty acid metabolism group (log-rank test, P=0.00063). More importantly, higher immunotherapy-related LncRNA expression was significantly associated with longer OS, and tumor-infiltrating immune cells were also significantly associated with patient prognosis. The comparison of the overall survival of patients with four types of breast cancer treated with immunotherapy is shown in the figure below. Figure 8 shown.

[0103] In an embodiment of the present invention, after obtaining each LncRNA subtype and each immune subtype, they are further merged into a two-dimensional index to generate a first number of LncRNA-immune subtypes. And each LncRNA-immune subtype may have a LncRNA-immune subtype with a similar therapeutic effect on immunotherapy. In order to improve the accuracy and efficiency of subsequent classification predictions, it is necessary to merge the LncRNA-immune subtypes with similar therapeutic effects on immunotherapy into the same type. Therefore, this embodiment uses the metabolic pathway enrichment feature as an evaluation index for each LncRNA-immune subtype, and merges the LncRNA-immune subtypes with the same metabolic pathway enrichment feature into the same breast cancer typing, and finally obtains the second number of breast cancer typing. The breast cancer typing obtained in this embodiment significantly distinguishes different immunotherapy efficacy and prognosis survival time, and reveals the metabolic connotation of each breast cancer typing, providing data support for the formulation of subsequent tumor treatment plans.

[0104] In a possible implementation, in step S5, the classification prediction model generates a corresponding breast cancer classification prediction result according to the pathological section image of the patient to be predicted, such as Fig. 9 As shown, steps S501-S505 are included:

[0105] Step S501, performing image segmentation on the pathological slice image according to the tissue regions in the pathological slice image to obtain a plurality of image blocks;

[0106] Step S502: extracting a number of feature vectors from each image block using a convolutional neural network model with pre-trained parameters;

[0107] Step S503: sorting the image blocks according to the feature vectors using a multi-instance learning algorithm based on an attention mechanism, and assigning a corresponding attention score to each image block;

[0108] Step S504: Aggregate the image information of each image block according to the attention score to generate aggregate information;

[0109] Step S505: According to the aggregated information, through a multi-branch network architecture and multi-task objectives, the convolutional neural network model is used to predict the breast cancer typing of the pathological slice image, and a breast cancer typing prediction result of the patient to be predicted is generated.

[0110] The embodiment of the present invention provides a method for predicting breast cancer typing based on pathological slice images, using digital high-resolution histological slices as input to a typing prediction model. For each pathological slice image, the tissue content is automatically segmented and evenly divided into thousands to tens of thousands of regions as small image blocks. Then, the corresponding feature vectors are extracted from each image block through a convolutional neural network model with pre-trained parameters. According to the feature vectors, each image block is sorted, and according to the relative importance of the image block, that is, the attention score, the image information of each image block is aggregated to obtain the aggregated information of the entire pathological slice image. Finally, breast cancer typing is predicted based on the aggregated information, and a breast cancer typing prediction result for the patient to be predicted is generated, which realizes breast cancer typing diagnosis based on digital pathological slice features, avoids the gene sequencing required for traditional tumor typing, and can perform typing prediction with the help of clinical pathological slices only, thereby improving the convenience and accuracy of breast cancer typing prediction.

[0111] In a preferred embodiment, the model is constructed based on the pathological omics data of breast cancer patients in the TCGA database and the SYSMH database, and the patient data is in the form of digital high-resolution histological sections as the main input. First, an artificial intelligence algorithm is used to automatically segment the tissue content for each whole slice image, and it is evenly divided into thousands to tens of thousands of regions as small image blocks. These images are processed by a convolutional neural network (CNN) with fixed pre-trained parameters, which extracts compact and descriptive feature vectors from each image block as an encoder. Secondly, through a multi-instance learning algorithm based on an attention mechanism, CNN learns to use feature vectors to sort all tissue regions in the slide, and aggregates their information in the entire image according to their relative importance, giving greater weight to regions that are considered to have high diagnostic relevance. Again, by using a multi-branch network architecture and multi-task objectives, CNN can predict different metabolic-immune subtypes. In addition, the attention score assigned to each region by the network can be used to explain the prediction of the model. The specific steps are as follows:

[0112] The specific steps are as follows:

[0113] ①Data collection and preparation:

[0114] Patient data collection: Pathological omics data of breast cancer patients were collected from the TCGA database and the SYSMH database, with digital high-resolution histological sections as the main input.

[0115] Image preprocessing: Each whole slice image is automatically segmented to generate thousands to tens of thousands of small image blocks as processing units.

[0116] ②Convolutional Neural Network (CNN) Processing:

[0117] Feature extraction: A CNN with pre-trained parameters is used as an encoder to extract a compact and descriptive feature vector from each image patch.

[0118] Attention-based Ranking: Using an attention-based multi-instance learning algorithm, the CNN learns to rank the tissue regions in the slides and aggregate the information according to relative importance.

[0119] ③Multi-task target prediction:

[0120] Breast cancer classification prediction: Through a multi-branch network architecture and multi-task objectives, CNN predicts different breast cancer classifications.

[0121] Attention score explanation: The attention score assigned by the network to each region is used to explain the model's prediction results.

[0122] ④Result output and explanation:

[0123] Generation of diagnostic results: Generate pathological classification diagnostic results for breast cancer patients based on the model’s prediction results.

[0124] Attention score explanation: The attention score output by the model can be used to explain the importance assessment of each region, thereby providing diagnostic interpretability and transparency. The construction process of the classification prediction model in this embodiment is as follows: Fig.10 shown.

[0125] Compared with the existing technology, this model has the following advantages:

[0126] (1) Comprehensive multi-omics data: This model fully integrates multi-omics data, including long noncoding RNA (LncRNA) and immune cell characteristics, and constructs a breast cancer clinical classification model through multi-faceted information, making the prediction more comprehensive.

[0127] (2) Application of deep learning technology: By adopting deep learning technology, especially convolutional neural network (CNN) and models with attention mechanism, this model realizes efficient analysis of pathological images and provides a new intelligent means for the clinical classification of breast cancer.

[0128] (3) Superior performance indicators: By using indicators such as the receiver operating characteristic (ROC) curve and its area under the curve (AUC) to evaluate, the model achieved excellent performance on the training set and validation set, with a micro-average AUC of up to 0.96 and an excellent macro-average ROC performance.

[0129] (4) Integration of classification and prognosis: This model can not only perform clinical classification of breast cancer, but also achieve prognosis prediction, providing doctors with more comprehensive information and helping to develop more individualized treatment plans.

[0130] (5) Lowering the threshold for clinical application: Through the classification diagnosis model based on pathological images, this model avoids dependence on gene sequencing and can perform classification diagnosis based solely on clinical pathological sections, thus lowering the threshold for clinical application of new classifications.

[0131] (6) Accurate prediction of immunotherapy effects: Compared with traditional indicators, this model has better sensitivity and accuracy, provides accurate prediction of the efficacy and prognosis of breast cancer immunotherapy, and provides patients with more effective treatment options.

[0132] Embodiment 2:

[0133] like Fig.11 As shown, accordingly, embodiment 2 provides a breast cancer typing prediction system based on LncRNA and immune cell characteristics, including an acquisition module 10, a LncRNA subtype generation module 20, an immune subtype generation module 30, a breast cancer typing generation module 40 and a prediction module 50;

[0134] Wherein, the acquisition module 10 is used to acquire a first RNA sequencing data set of a number of historical breast cancer patients and a second RNA sequencing data set of a number of patients receiving immunotherapy;

[0135] The LncRNA subtype generation module 20 is used to input the first RNA sequencing data set into a preset LncRNA clustering model, so that the LncRNA clustering model clusters each RNA sequencing data in the first RNA sequencing data set into different LncRNA subtypes, wherein the LncRNA clustering model is constructed based on the feature set of the second RNA sequencing data set;

[0136] The immune subtype generation module 30 is used to input the first RNA sequencing data set into a preset immune cell feature clustering model, so that the immune cell feature clustering model clusters each RNA sequencing data in the first RNA sequencing data set into different immune subtypes, wherein the immune cell feature clustering model is constructed based on each immune cell in the second RNA sequencing data set and the corresponding ssGSEA score;

[0137] The breast cancer typing generation module 40 is used to combine the various LncRNA subtypes and the various immune subtypes to obtain several different breast cancer typings;

[0138] The prediction module 50 is used to input the pathological section image of the patient to be predicted into a preset typing prediction model so that the typing prediction model generates a corresponding breast cancer typing prediction result, wherein the breast cancer typing prediction result is one of the several different breast cancer typings, and the typing prediction model is constructed based on the pathological section data of historical breast cancer patients and a pre-trained convolutional neural network model.

[0139] In a possible implementation, constructing the LncRNA clustering model according to the feature set of the second RNA sequencing data set includes:

[0140] Transcribe the corresponding LncRNA expression dataset from the second RNA sequencing dataset;

[0141] According to the influence factor of each feature in the LncRNA expression data set on immunotherapy, a number of LncRNA features are extracted from the LncRNA expression data set, and the LncRNA feature set is formed by combining them;

[0142] A lncRNA clustering model is constructed according to the lncRNA feature set.

[0143] In a possible implementation, constructing the immune cell feature clustering model according to each immune cell in the second RNA sequencing data set and the corresponding ssGSEA score includes:

[0144] Inputting the second RNA sequencing data set into a preset bioinformatics analysis software, so that the bioinformatics analysis software determines several types of immune cells and ssGSEA scores corresponding to various immune cells through ssGSEA analysis;

[0145] An immune cell feature clustering model was constructed based on the various immune cells and the corresponding ssGSEA scores.

[0146] Furthermore, the breast cancer typing generation module 40 combines the various LncRNA subtypes and the various immune subtypes to obtain several different breast cancer typings, including:

[0147] Combining the LncRNA subtypes and the immune subtypes in pairs to generate a first number of LncRNA-immune subtypes;

[0148] GSEA metabolic pathway analysis was used to perform pathway enrichment scoring on each of the LncRNA-immune subtypes to obtain the corresponding pathway enrichment score;

[0149] According to each of the pathway enrichment scores, the metabolic pathway enrichment characteristics of each of the LncRNA-immune subtypes are determined, and the LncRNA-immune subtypes with the same metabolic pathway enrichment characteristics are combined to obtain a second number of breast cancer typing, wherein the second number is less than or equal to the first number.

[0150] In one possible implementation, Fig.12 As shown, the prediction module 50 includes an image segmentation unit 501, a feature extraction unit 502, a sorting unit 503, an information aggregation unit 504 and a prediction unit 505;

[0151] The image segmentation unit 501 is used to segment the pathological slice image according to the tissue area in the pathological slice image to obtain a plurality of image blocks;

[0152] The feature extraction unit 502 is used to extract a number of feature vectors from each image block using a convolutional neural network model with pre-trained parameters;

[0153] The sorting unit 503 is used to sort the image blocks according to the feature vectors using a multi-instance learning algorithm based on an attention mechanism, and assign a corresponding attention score to each image block;

[0154] The information aggregation unit 504 is used to aggregate the image information of each image block according to the attention score to generate aggregated information;

[0155] The prediction unit 505 is used to predict the breast cancer typing of the pathological slice image using the convolutional neural network model according to the aggregated information, through a multi-branch network architecture and multi-task objectives, to generate a breast cancer typing prediction result for the patient to be predicted.

[0156] The embodiment of the present invention provides a breast cancer typing prediction system based on LncRNA and immune cell characteristics. The embodiment first constructs a LncRNA clustering model and an immune cell characteristic clustering model according to the second RNA sequencing data set, which is used to mine the LncRNA subtypes and immune subtypes in the first RNA sequencing data set. Then, in combination with each LncRNA subtype and each immune subtype, several breast cancer typings are constructed. The breast cancer typing constructed in this embodiment fully considers the effects of long non-coding RNA and immune cells on the efficacy of tumor immunotherapy, and reconstructs a breast cancer typing that is more sensitive to immunotherapy. The breast cancer typing has better sensitivity and accuracy than traditional indicators for predicting the efficacy of immunotherapy. Further, in accordance with the breast cancer typing, the embodiment also provides a typing prediction model based on pathological images, which avoids the gene sequencing required for traditional tumor typing, and can perform typing prediction only with the help of clinical pathological sections, thereby improving the convenience and accuracy of breast cancer typing prediction, and lowering the threshold for clinical application of the breast cancer typing. The obtained breast cancer typing prediction results can provide a decision-making basis for medical workers to formulate cancer treatment plans.

[0157] The more detailed working principle and step flow of this embodiment can refer to, but are not limited to, the relevant records of Embodiment 1.

[0158] The specific embodiments described above further illustrate the purpose, technical solutions and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A breast cancer typing prediction method based on LncRNA and immune cell characteristics, characterized in that: include: Obtaining a first RNA sequencing data set of several historical breast cancer patients and a second RNA sequencing data set of several patients receiving immunotherapy; Inputting the first RNA sequencing data set into a preset LncRNA clustering model, so that the LncRNA clustering model clusters each RNA sequencing data in the first RNA sequencing data set into different LncRNA subtypes, wherein the LncRNA clustering model is constructed based on the feature set of the second RNA sequencing data set; Inputting the first RNA sequencing data set into a preset immune cell feature clustering model, so that the immune cell feature clustering model clusters each RNA sequencing data in the first RNA sequencing data set into different immune subtypes, wherein the immune cell feature clustering model is constructed based on each immune cell in the second RNA sequencing data set and the corresponding ssGSEA score; Combining the various LncRNA subtypes and the various immune subtypes, several different breast cancer classifications are obtained; The pathological section image of the patient to be predicted is input into a preset typing prediction model so that the typing prediction model generates a corresponding breast cancer typing prediction result, wherein the breast cancer typing prediction result is one of the several different breast cancer typings, and the typing prediction model is constructed based on the pathological section data of historical breast cancer patients and a pre-trained convolutional neural network model.

2. A breast cancer typing prediction method based on LncRNA and immune cell characteristics according to claim 1, characterized in that: The step of constructing the LncRNA clustering model according to the feature set of the second RNA sequencing data set comprises: Transcribe the corresponding LncRNA expression dataset from the second RNA sequencing dataset; According to the influence factor of each feature in the LncRNA expression data set on immunotherapy, a number of LncRNA features are extracted from the LncRNA expression data set, and the LncRNA feature set is formed by combining them; A lncRNA clustering model is constructed according to the lncRNA feature set.

3. The method for predicting breast cancer typing based on LncRNA and immune cell characteristics according to claim 1, characterized in that: The step of constructing the immune cell feature clustering model according to each immune cell in the second RNA sequencing data set and the corresponding ssGSEA score comprises: Inputting the second RNA sequencing data set into a preset bioinformatics analysis software, so that the bioinformatics analysis software determines several types of immune cells and ssGSEA scores corresponding to various immune cells through ssGSEA analysis; An immune cell feature clustering model was constructed based on the various immune cells and the corresponding ssGSEA scores.

4. The method for predicting breast cancer typing based on LncRNA and immune cell characteristics according to claim 1, characterized in that: The combination of the various LncRNA subtypes and the various immune subtypes obtains several different breast cancer classifications, including: Combining the LncRNA subtypes and the immune subtypes in pairs to generate a first number of LncRNA-immune subtypes; GSEA metabolic pathway analysis was used to perform pathway enrichment scoring on each of the LncRNA-immune subtypes to obtain the corresponding pathway enrichment score; According to each of the pathway enrichment scores, the metabolic pathway enrichment characteristics of each of the LncRNA-immune subtypes are determined, and the LncRNA-immune subtypes with the same metabolic pathway enrichment characteristics are combined to obtain a second number of breast cancer typing, wherein the second number is less than or equal to the first number.

5. The method for predicting breast cancer typing based on LncRNA and immune cell characteristics according to claim 1, characterized in that: The classification prediction model generates corresponding breast cancer classification prediction results according to the pathological section images of the patient to be predicted, including: Performing image segmentation on the pathological slice image according to tissue regions in the pathological slice image to obtain a plurality of image blocks; Extracting a number of feature vectors from each of the image blocks using a convolutional neural network model with pre-trained parameters; According to the feature vector, the image blocks are sorted using a multi-instance learning algorithm based on an attention mechanism, and a corresponding attention score is assigned to each image block; Aggregating image information of each image block according to the attention score to generate aggregate information; According to the aggregated information, through a multi-branch network architecture and multi-task objectives, the convolutional neural network model is used to predict the breast cancer typing of the pathological slice image to generate a breast cancer typing prediction result for the patient to be predicted.

6. A breast cancer typing prediction system based on LncRNA and immune cell characteristics, characterized in that: It includes acquisition module, LncRNA subtype generation module, immune subtype generation module, breast cancer classification generation module and prediction module; Wherein, the acquisition module is used to acquire a first RNA sequencing data set of several historical breast cancer patients and a second RNA sequencing data set of several patients receiving immunotherapy; The LncRNA subtype generation module is used to input the first RNA sequencing data set into a preset LncRNA clustering model, so that the LncRNA clustering model clusters each RNA sequencing data in the first RNA sequencing data set into different LncRNA subtypes, wherein the LncRNA clustering model is constructed based on the feature set of the second RNA sequencing data set; The immune subtype generation module is used to input the first RNA sequencing data set into a preset immune cell feature clustering model, so that the immune cell feature clustering model clusters each RNA sequencing data in the first RNA sequencing data set into different immune subtypes, wherein the immune cell feature clustering model is constructed based on each immune cell in the second RNA sequencing data set and the corresponding ssGSEA score; The breast cancer typing generation module is used to combine the various LncRNA subtypes and the various immune subtypes to obtain several different breast cancer typings; The prediction module is used to input the pathological section image of the patient to be predicted into a preset typing prediction model so that the typing prediction model generates a corresponding breast cancer typing prediction result, wherein the breast cancer typing prediction result is one of the several different breast cancer typings, and the typing prediction model is constructed based on the pathological section data of historical breast cancer patients and a pre-trained convolutional neural network model.

7. A breast cancer typing prediction system based on LncRNA and immune cell characteristics as claimed in claim 6, characterized in that: The step of constructing the LncRNA clustering model according to the feature set of the second RNA sequencing data set comprises: Transcribe the corresponding LncRNA expression dataset from the second RNA sequencing dataset; According to the influence factor of each feature in the LncRNA expression data set on immunotherapy, a number of LncRNA features are extracted from the LncRNA expression data set, and the LncRNA feature set is formed by combining them; A lncRNA clustering model is constructed according to the lncRNA feature set.

8. A breast cancer typing prediction system based on LncRNA and immune cell characteristics as claimed in claim 6, characterized in that: The step of constructing the immune cell feature clustering model according to each immune cell in the second RNA sequencing data set and the corresponding ssGSEA score comprises: Inputting the second RNA sequencing data set into a preset bioinformatics analysis software, so that the bioinformatics analysis software determines several types of immune cells and ssGSEA scores corresponding to various immune cells through ssGSEA analysis; An immune cell feature clustering model was constructed based on the various immune cells and the corresponding ssGSEA scores.

9. A breast cancer typing prediction system based on LncRNA and immune cell characteristics as claimed in claim 6, characterized in that: The breast cancer typing generation module combines the various LncRNA subtypes and the various immune subtypes to obtain several different breast cancer typings, including: Combining the LncRNA subtypes and the immune subtypes in pairs to generate a first number of LncRNA-immune subtypes; GSEA metabolic pathway analysis was used to perform pathway enrichment scoring on each of the LncRNA-immune subtypes to obtain the corresponding pathway enrichment score; According to each of the pathway enrichment scores, the metabolic pathway enrichment characteristics of each of the LncRNA-immune subtypes are determined, and the LncRNA-immune subtypes with the same metabolic pathway enrichment characteristics are combined to obtain a second number of breast cancer typing, wherein the second number is less than or equal to the first number.

10. A breast cancer typing prediction system based on LncRNA and immune cell characteristics as claimed in claim 6, characterized in that: The prediction module includes an image segmentation unit, a feature extraction unit, a sorting unit, an information aggregation unit and a prediction unit; Wherein, the image segmentation unit is used to perform image segmentation on the pathological slice image according to the tissue area in the pathological slice image to obtain a plurality of image blocks; The feature extraction unit is used to extract a number of feature vectors from each image block using a convolutional neural network model with pre-trained parameters; The sorting unit is used to sort the image blocks according to the feature vectors using a multi-instance learning algorithm based on an attention mechanism, and assign a corresponding attention score to each image block; The information aggregation unit is used to aggregate the image information of each image block according to the attention score to generate aggregated information; The prediction unit is used to predict the breast cancer typing of the pathological slice image using the convolutional neural network model based on the aggregated information, through a multi-branch network architecture and multi-task objectives, to generate a breast cancer typing prediction result for the patient to be predicted.

Citation Information

Patent Citations

  • Tumor immune typing method based on lncRNA markers

    CN113355411A

  • Method for predicting survival prognosis of bladder cancer patient based on lncRNA optimization model

    CN115762792A

  • UTUC molecular typing and single sample classifier and construction method thereof

    CN116987789A

  • Prognosis risk prediction model for patient suffering from lung adenocarcinoma spread through air cavity and application of prognosis risk prediction model

    CN117106916A

  • Systems and methods for predicting response of triple-negative breast cancer to neoadjuvant chemotherapy

    US20240161276A1