Prediction method and device for breast cancer recurrence risk, equipment and storage medium

Abstract

The invention is suitable for the technical field of breast cancer recurrence risks, and provides a breast cancer recurrence risk prediction method and device, equipment and a storage medium. The breast cancer recurrence risk prediction model is obtained by training a pre-constructed initial breast cancer recurrence risk prediction model based on a plurality of breast cancer recurrence risk training samples; each breast cancer recurrence risk training sample carries genotype states of a plurality of related single nucleotide polymorphism variation sites, information of a plurality of pathological characteristics and actual recurrence risk probability; and predicting the recurrence risk of the breast cancer recurrence risk sample to be predicted through the breast cancer recurrence risk prediction model to obtain a predicted recurrence risk probability corresponding to the breast cancer recurrence risk sample to be predicted. According to the invention, the problem of low prediction accuracy caused by predicting the prognosis recurrence risk of the breast cancer patient through a single molecular marker in the prior art can be solved.

Description

Technical Field

[0001] The present application belongs to the technical field of breast cancer recurrence risk, and in particular, relates to a method, device, equipment and storage medium for predicting breast cancer recurrence risk. Background Art

[0002] According to tumor registration data collected by the tumor registry center, breast cancer is one of the most common malignant tumors in women worldwide. Although breast cancer patients have a high overall survival rate after surgical treatment, local recurrence or distant metastasis of breast cancer after surgery still exists, especially in the first 2 years. The risk of recurrence is high, and the level of recurrence risk determines the further treatment plan for breast cancer patients. Therefore, how to effectively evaluate and predict the risk of recurrence of breast cancer patients after surgical treatment and determine further treatment plans are difficult problems that need to be solved urgently.

[0003] Molecular markers related to breast cancer prognosis include genotyping, RNA (ribonucleic acid) expression levels, and protein expression levels. The risk of postoperative recurrence of breast cancer is usually affected by the patient's genetic background (for example, different races, different regional populations) and tumor pathological characteristics. However, the current prediction of the risk of recurrence of breast cancer patients after surgical treatment mostly relies on a single molecular marker. This method of predicting the prognosis and recurrence risk of breast cancer patients using a single molecular marker ignores the impact of the patient's tumor pathological characteristics on postoperative recurrence, so the prediction accuracy of the recurrence risk is low, which will also affect the further treatment plan for breast cancer patients. Summary of the invention

[0004] The embodiments of the present application provide a method, apparatus, device and storage medium for predicting the recurrence risk of breast cancer, which can solve the problem of low prediction accuracy caused by the existing prediction of the prognosis recurrence risk of breast cancer patients by a single molecular marker.

[0005] In a first aspect, the present application provides a method for predicting the risk of breast cancer recurrence, comprising: Obtaining a breast cancer recurrence risk prediction model, wherein the breast cancer recurrence risk prediction model is obtained by training a pre-constructed initial breast cancer recurrence risk prediction model based on multiple breast cancer recurrence risk training samples, each of the breast cancer recurrence risk training samples carries the genotype status of multiple related single nucleotide polymorphism variation sites, information on multiple pathological characteristics, and an actual recurrence risk probability, and the related single nucleotide polymorphism variation sites represent single nucleotide polymorphism variation sites in the peripheral blood of the breast cancer recurrence risk training samples that are related to breast cancer prognosis; The breast cancer recurrence risk prediction model is used to predict the recurrence risk of the breast cancer recurrence risk sample to be predicted, and the predicted recurrence risk probability corresponding to the breast cancer recurrence risk sample to be predicted is obtained.

[0006] In a possible implementation manner of the first aspect, before obtaining the breast cancer recurrence risk prediction model, the method includes: Acquire a plurality of breast cancer recurrence risk training samples; Performing weighted calculations on the genotype states of the multiple related single nucleotide polymorphism variation sites and the information of the multiple pathological characteristics carried by each of the breast cancer recurrence risk training samples, respectively, to obtain a predicted recurrence risk probability corresponding to each of the breast cancer recurrence risk training samples; The pre-constructed initial breast cancer recurrence risk prediction model is trained according to the predicted recurrence risk probabilities and the actual recurrence risk probabilities corresponding to each of the plurality of breast cancer recurrence risk training samples to obtain the breast cancer recurrence risk prediction model.

[0007] In a possible implementation of the first aspect, the pre-constructed initial breast cancer recurrence risk prediction model is trained according to the predicted recurrence risk probabilities and the actual recurrence risk probabilities corresponding to each of the plurality of breast cancer recurrence risk training samples to obtain the breast cancer recurrence risk prediction model, including: According to the predicted recurrence risk probabilities and the actual recurrence risk probabilities corresponding to each of the plurality of breast cancer recurrence risk training samples, a prediction error of the initial breast cancer recurrence risk prediction model is calculated by a cross-validation method; According to the prediction error of the initial breast cancer recurrence risk prediction model, optimizing and adjusting the adjustment parameters of the initial breast cancer recurrence risk prediction model; Determine the adjustment parameter when the prediction error of the initial breast cancer recurrence risk prediction model is the smallest as the optimal adjustment parameter; According to the optimal adjustment parameters, respectively determining the optimal weight of the genotype status of each of the relevant single nucleotide polymorphism variation sites and the information of each of the pathological characteristics; The breast cancer recurrence risk prediction model is determined according to the optimal weights of the genotype status of all the relevant single nucleotide polymorphism variation sites and the information of all the pathological characteristics.

[0008] In a possible implementation of the first aspect, the related single nucleotide polymorphism variation sites include: related single nucleotide polymorphism variation site rs2285892, related single nucleotide polymorphism variation site rs554098, related single nucleotide polymorphism variation site rs372637371, related single nucleotide polymorphism variation site rs113756220, related single nucleotide polymorphism variation site rs17489363, related single nucleotide polymorphism variation site rs4140390, related single nucleotide polymorphism variation site rs5901000 and related single nucleotide polymorphism variation site rs111033566; The pathological characteristics include: histological grade characteristics, tumor category characteristics, lymph node metastasis characteristics, TNM staging characteristics, ER status characteristics, PR status characteristics, HER-2 status characteristics, Ki-67 proliferation index characteristics and molecular subtype characteristics.

[0009] In a possible implementation of the first aspect, predicting the recurrence risk of a sample to be predicted for breast cancer recurrence risk by using the breast cancer recurrence risk prediction model to obtain a predicted recurrence risk probability corresponding to the sample to be predicted for breast cancer recurrence risk includes: Inputting the breast cancer recurrence risk sample to be predicted into the breast cancer recurrence risk prediction model, wherein the breast cancer recurrence risk sample to be predicted carries the genotype status of multiple related single nucleotide polymorphism variation sites and information of multiple pathological characteristics; Normalizing the genotype states of the multiple related single nucleotide polymorphism variation sites and the information of the multiple pathological characteristics carried by the sample to be predicted for the risk of breast cancer recurrence, respectively, to obtain the genotype states of the multiple normalized related single nucleotide polymorphism variation sites and the information of the multiple normalized pathological characteristics; According to the optimal weight of the genotype status of each relevant single nucleotide polymorphism variation site and each pathological feature information, the genotype status of all the normalized relevant single nucleotide polymorphism variation sites and all the normalized pathological feature information are weightedly calculated to obtain the corresponding predicted recurrence risk probability of the breast cancer recurrence risk sample to be predicted.

[0010] In a possible implementation of the first aspect, the genotype states of the multiple relevant single nucleotide polymorphism variation sites carried by each of the breast cancer recurrence risk training samples are obtained by the following method: Obtaining a plurality of the related single nucleotide polymorphism variation sites associated with breast cancer prognosis, and specific detection primers corresponding to each of the related single nucleotide polymorphism variation sites; Acquire genomic DNA of a plurality of the breast cancer recurrence risk training samples, wherein the genomic DNA of the breast cancer recurrence risk training samples is obtained by extracting peripheral blood of the breast cancer recurrence risk training samples; Using specific detection primers corresponding to each of the relevant single nucleotide polymorphism variation sites, the genotype status of multiple relevant single nucleotide polymorphism variation sites in the genomic DNA of each of the breast cancer recurrence risk training samples is detected through molecular biology methods to obtain the genotype status corresponding to each of the multiple relevant single nucleotide polymorphism variation sites of the breast cancer recurrence risk training sample.

[0011] In a possible implementation of the first aspect, the obtaining of the plurality of relevant single nucleotide polymorphism variation sites associated with breast cancer prognosis includes: Obtain multiple candidate single nucleotide polymorphism variation sites; The risk correlation between each candidate single nucleotide polymorphism variation site and breast cancer prognosis was calculated by COX univariate and multivariate regression analysis; According to the risk association between each candidate single nucleotide polymorphism variation site and the prognosis of breast cancer, relevant single nucleotide polymorphism variation sites associated with the prognosis of breast cancer are selected from multiple candidate single nucleotide polymorphism variation sites.

[0012] In a second aspect, an embodiment of the present application provides a device for predicting the risk of recurrence of breast cancer, comprising: A model acquisition module, used for acquiring a breast cancer recurrence risk prediction model, wherein the breast cancer recurrence risk prediction model is obtained by training a pre-constructed initial breast cancer recurrence risk prediction model based on a plurality of breast cancer recurrence risk training samples, each of the breast cancer recurrence risk training samples carries the genotype status of a plurality of relevant single nucleotide polymorphism variation sites, information of a plurality of pathological characteristics and an actual recurrence risk probability, and the relevant single nucleotide polymorphism variation sites represent single nucleotide polymorphism variation sites in the peripheral blood of the breast cancer recurrence risk training samples that are related to breast cancer prognosis; The risk prediction module is used to predict the recurrence risk of the breast cancer recurrence risk sample to be predicted by using the breast cancer recurrence risk prediction model, and obtain the predicted recurrence risk probability corresponding to the breast cancer recurrence risk sample to be predicted.

[0013] In a third aspect, an embodiment of the present application provides a device for predicting the risk of recurrence of breast cancer, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements any of the methods described above when executing the computer program.

[0014] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it implements any of the methods described above.

[0015] In a fifth aspect, an embodiment of the present application provides a computer program product, which, when executed on a terminal device, enables the terminal device to execute any of the methods described above.

[0016] Compared with the prior art, the embodiments of the present application have the following beneficial effects: by obtaining a breast cancer recurrence risk prediction model, since the breast cancer recurrence risk prediction model is obtained by training a pre-constructed initial breast cancer recurrence risk prediction model based on multiple breast cancer recurrence risk training samples, the trained breast cancer recurrence risk prediction model contains information of multiple breast cancer recurrence risk training samples, and each breast cancer recurrence risk training sample carries the genotype status of multiple related single nucleotide polymorphism variation sites, information of multiple pathological characteristics and actual recurrence risk probability, so that the breast cancer recurrence risk prediction model contains richer information, and improves the prediction accuracy of the breast cancer recurrence risk prediction model for breast cancer prognosis recurrence risk; through the breast cancer recurrence risk prediction model, the recurrence risk of the breast cancer recurrence risk sample to be predicted is predicted, so that the predicted recurrence risk probability corresponding to the obtained breast cancer recurrence risk sample to be predicted is more accurate, and the accuracy of the predicted recurrence risk probability corresponding to the breast cancer recurrence risk sample to be predicted is improved, thereby solving the problem of low prediction accuracy when predicting the prognosis recurrence risk of breast cancer patients by a single molecular marker. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 This is a flow chart of a method for predicting breast cancer recurrence risk provided in one embodiment of the present application; Figure 2 It is a schematic diagram of the analysis results of a related single nucleotide polymorphism variation site provided in one embodiment of the present application; Figure 3 is a schematic diagram of an analysis result of an optimal adjustment parameter provided by an embodiment of the present application; Figure 4 This is a schematic diagram of the prediction accuracy of a breast cancer recurrence risk prediction model in a training group provided by an embodiment of the present application; Figure 5 This is a schematic diagram of the prediction accuracy in a validation group of a breast cancer recurrence risk prediction model provided in one embodiment of the present application; Figure 6 This is a schematic diagram of comparative analysis of the prediction performance of a breast cancer recurrence risk prediction model provided by an embodiment of the present application and two other breast cancer recurrence risk prediction models; Figure 7 It is a schematic diagram of the structure of a device for predicting the risk of recurrence of breast cancer provided in one embodiment of the present application. DETAILED DESCRIPTION

[0019] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0020] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0021] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0022] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0023] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0024] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0025] Single-nucleotide polymorphisms (SNPs) refer to DNA sequence polymorphisms caused by the variation of a single nucleotide at the genome level. SNPs are also related to the occurrence, development and prognosis of cancer.

[0026] Currently, the prognostic risk of breast cancer is mostly predicted based on the results of 21 genes (Oncotype Dx) and 70 genes (Mamma Print).

[0027] The 21-gene test uses a high-throughput, real-time reverse transcription polymerase chain reaction (RT-PCR) method to detect the expression of 16 cancer-related genes and 5 internal reference genes (a total of 21 genes) in paraffin-embedded breast cancer tissues. Based on the expression of the 21 genes, the prognosis and recurrence risk of breast cancer patients is scored, and breast cancer patients are determined to have a low recurrence risk, a medium recurrence risk, or a high recurrence risk based on the score. At the same time, the score can also be used to determine whether breast cancer patients need adjuvant chemotherapy.

[0028] 70-gene detection refers to the use of cDNA chip technology to measure the expression levels of 70 genes related to cancer biological characteristics such as cell cycle regulation, proliferation, angiogenesis, invasion, metastasis, and signal transduction. The prognosis and recurrence risk of breast cancer patients are scored based on the expression levels of these 70 genes, and breast cancer patients are determined to have a low or high recurrence risk based on the score. At the same time, the score can also be used to determine whether breast cancer patients need adjuvant chemotherapy.

[0029] However, human genetics is diverse, and there are significant genetic differences between different races and populations in different regions. It is very necessary to establish a breast cancer surgical recurrence risk model based on the molecular characteristics (SNP genotypes) and pathological characteristics of different populations. In addition, the prognosis and recurrence risk of breast cancer is also affected by the information of the pathological characteristics of breast cancer patients. The genotype status of a single single nucleotide polymorphism has defects and deficiencies in predicting the risk of postoperative recurrence and metastasis of early and locally advanced breast cancer. The prognosis and recurrence risk of breast cancer determined solely by the genotype status of a single single nucleotide polymorphism is not accurate.

[0030] The present application example provides a method for predicting the risk of breast cancer recurrence, see Figure 1 , Figure 1 This is a flow chart of a method for predicting breast cancer recurrence risk provided by an embodiment of the present application, comprising: Step S11, obtaining a breast cancer recurrence risk prediction model, wherein the breast cancer recurrence risk prediction model is obtained by training a pre-constructed initial breast cancer recurrence risk prediction model based on multiple breast cancer recurrence risk training samples, each breast cancer recurrence risk training sample carries the genotype status of multiple related single nucleotide polymorphism variation sites, information of multiple pathological characteristics and actual recurrence risk probability, and the related single nucleotide polymorphism variation sites represent single nucleotide polymorphism variation sites in the peripheral blood of the breast cancer recurrence risk training samples that are related to breast cancer prognosis; Step S12, predicting the recurrence risk of the breast cancer recurrence risk sample to be predicted by using the breast cancer recurrence risk prediction model, and obtaining the predicted recurrence risk probability corresponding to the breast cancer recurrence risk sample to be predicted.

[0031] It should be noted that for breast cancer patients, after treatment or surgery, further treatment plans need to be adjusted according to their prognosis. Predictive assessment of the risk of recurrence in breast cancer patients can help determine further treatment plans for breast cancer patients. For example, patients with a low risk of recurrence can be exempted from chemotherapy or weakened treatment, while patients with a high risk of recurrence are recommended to receive chemotherapy or intensive treatment.

[0032] Specifically, an initial breast cancer recurrence risk model is pre-built, which may be a LASSO (least absolute shrinkage and selection operator) regression model. The initial breast cancer recurrence risk model corresponds to an initial adjustment parameter, which determines the selection of variable features that affect the prediction accuracy of the initial breast cancer recurrence risk model, as well as the weight coefficient of each selected variable feature.

[0033] Multiple breast cancer recurrence risk training samples are obtained in advance, and each breast cancer recurrence risk training sample can be a patient for predicting the prognosis of early or locally advanced breast cancer. Each breast cancer recurrence risk training sample carries the values ​​of multiple variable features, such as: the genotype status of multiple related single nucleotide polymorphism variation sites and information on multiple pathological features.

[0034] Among them, the relevant single nucleotide polymorphism variation sites are used as molecular markers to represent the single nucleotide polymorphism variation sites related to breast cancer prognosis in the peripheral blood of the breast cancer recurrence risk training samples. Single nucleotide polymorphism (SNP) refers to the variation of a single nucleotide (A, T, C or G) at a specific position in the genome to form different alleles.

[0035] In an optional example, the related single nucleotide polymorphism variation sites include the related single nucleotide polymorphism variation site rs2285892, the related single nucleotide polymorphism variation site rs554098, the related single nucleotide polymorphism variation site rs372637371, the related single nucleotide polymorphism variation site rs113756220, the related single nucleotide polymorphism variation site rs17489363, the related single nucleotide polymorphism variation site rs4140390, the related single nucleotide polymorphism variation site rs5901000 and the related single nucleotide polymorphism variation site rs111033566.

[0036] Specifically, the genotype state of a relevant single nucleotide polymorphism variation site of a breast cancer recurrence risk training sample represents the allele type of the breast cancer recurrence risk training sample at the relevant single nucleotide polymorphism variation site.

[0037] In an optional example, a single nucleotide polymorphism variation site of a breast cancer recurrence risk training sample has two alleles. At this site, the genetic variation of the breast cancer recurrence risk training sample usually only involves two different nucleotides, for example: it is a variation between A and G, then the allele types at this site are A and G.

[0038] Specifically, the information of multiple pathological features carried by each breast cancer recurrence risk training sample can be obtained through channels such as hospital information systems or online public resources. It indicates the abnormal cell or tissue state appearing in the lesion site (breast), which is a change in breast cancer at the tissue and cellular levels and can reflect the severity of breast cancer.

[0039] In an optional example, the pathological features include histological grade features, tumor category features, lymph node metastasis features, TNM staging features, ER (estrogen receptor) status features, PR (progesterone receptor) status features, HER-2 status features, Ki-67 proliferation index features, and molecular subtype features. For example, the information of the histological grade feature of a breast cancer recurrence risk training sample can be low malignancy (grade I), moderate malignancy (grade II), or high malignancy (grade III).

[0040] Specifically, the actual recurrence risk probability of a breast cancer recurrence risk training sample represents the probability of the actual recurrence risk of the breast cancer recurrence risk training sample. The higher the probability of the recurrence risk, the more likely the breast cancer recurrence risk training sample is to have breast cancer recurrence.

[0041] By using multiple breast cancer recurrence risk training samples that carry the genotype status of multiple relevant single nucleotide polymorphism variation sites, information on multiple pathological characteristics, and actual recurrence risk probabilities, the initial breast cancer recurrence risk model can be trained, the optimal adjustment parameters can be selected, and the variable features most relevant to the prediction of breast cancer prognosis recurrence risk, as well as the weight coefficient of each most relevant variable feature, can be determined to obtain a breast cancer recurrence risk model. It should be noted that the training rounds and number of times for training the initial breast cancer recurrence risk model can be set according to specific circumstances, and this embodiment does not specifically limit this.

[0042] A breast cancer recurrence risk sample to be predicted is obtained. When predicting the recurrence risk of the breast cancer recurrence risk sample to be predicted, the trained breast cancer recurrence risk model is obtained, the genotype status of multiple related single nucleotide polymorphism variation sites and information on multiple pathological characteristics carried by the breast cancer recurrence risk sample to be predicted are input into the breast cancer recurrence risk model, the recurrence risk of the breast cancer recurrence risk sample to be predicted is predicted, and the predicted recurrence risk probability corresponding to the breast cancer recurrence risk sample to be predicted is obtained.

[0043] When the predicted recurrence risk probability corresponding to the sample of breast cancer recurrence risk to be predicted is lower than the first threshold, the sample of breast cancer recurrence risk to be predicted is considered to be at low risk of recurrence and can be exempted from chemotherapy or weakened treatment; when the predicted recurrence risk probability corresponding to the sample of breast cancer recurrence risk to be predicted is higher than or equal to the first threshold, the sample of breast cancer recurrence risk to be predicted is considered to be at high risk of recurrence and chemotherapy or enhanced treatment can be recommended.

[0044] It is understandable that, based on the above technical solution, the breast cancer recurrence risk prediction model established based on multiple related single nucleotide polymorphism variant sites and multiple pathological characteristics can be effectively used to predict the risk of recurrence and metastasis of breast cancer patients. By detecting the genotype status of rs2285892, rs554098, rs372637371, rs113756220, rs17489363, rs4140390, rs5901000 and rs111033566 of breast cancer prognosis patients, as well as the information of histological grading characteristics, tumor category characteristics, lymph node metastasis characteristics, TNM staging characteristics, ER status characteristics, PR status characteristics, HER-2 status characteristics, Ki-67 proliferation index characteristics and molecular subtype characteristics, the risk of recurrence and metastasis of breast cancer prognosis patients is predicted, which is of great significance to the further diagnosis and treatment of patients and prolonging the survival of patients.

[0045] It can be understood that the technical solution provided in this embodiment obtains a breast cancer recurrence risk prediction model. Since the breast cancer recurrence risk prediction model is obtained by training a pre-constructed initial breast cancer recurrence risk prediction model based on multiple breast cancer recurrence risk training samples, the trained breast cancer recurrence risk prediction model contains information of multiple breast cancer recurrence risk training samples, and each breast cancer recurrence risk training sample carries the genotype status of multiple related single nucleotide polymorphism variation sites, information of multiple pathological characteristics and actual recurrence risk probability, so that the breast cancer recurrence risk prediction model contains richer information, and improves the prediction accuracy of the breast cancer recurrence risk prediction model for breast cancer prognosis recurrence risk; through the breast cancer recurrence risk prediction model, the recurrence risk of the breast cancer recurrence risk sample to be predicted is predicted, so that the predicted recurrence risk probability corresponding to the obtained breast cancer recurrence risk sample to be predicted is more accurate, and the accuracy of the predicted recurrence risk probability corresponding to the breast cancer recurrence risk sample to be predicted is improved, thereby solving the existing problem of low prediction accuracy when predicting the prognosis recurrence risk of breast cancer patients by a single molecular marker.

[0046] In a possible implementation, before obtaining the breast cancer recurrence risk prediction model in step S11, the method further includes: Step S101, obtaining a plurality of breast cancer recurrence risk training samples; Step S102, performing weighted calculations on the genotype states of multiple relevant single nucleotide polymorphism variation sites and information on multiple pathological characteristics carried by each breast cancer recurrence risk training sample, to obtain a predicted recurrence risk probability corresponding to each breast cancer recurrence risk training sample; Step S103, training the pre-constructed initial breast cancer recurrence risk prediction model according to the predicted recurrence risk probabilities and actual recurrence risk probabilities corresponding to each of the plurality of breast cancer recurrence risk training samples to obtain a breast cancer recurrence risk prediction model.

[0047] Specifically, before obtaining the breast cancer recurrence risk prediction model, the breast cancer recurrence risk prediction model has been trained, and the above method also includes: A plurality of breast cancer recurrence risk training samples are obtained, each of which carries information on the genotype status of a plurality of relevant single nucleotide polymorphism variation sites and a plurality of pathological characteristics.

[0048] According to the adjustment parameters of the initial breast cancer recurrence risk prediction model and the coefficients of the corresponding variable features, the values ​​of the variable features carried by each breast cancer recurrence risk training sample are weighted and calculated. Among them, each variable feature includes the genotype status of multiple related single nucleotide polymorphism mutation sites and information on multiple pathological characteristics, and the predicted recurrence risk probability corresponding to each breast cancer recurrence risk training sample is calculated.

[0049] According to the predicted recurrence risk probabilities and actual recurrence risk probabilities corresponding to the multiple breast cancer recurrence risk training samples calculated above, the prediction error of the pre-built breast cancer recurrence risk prediction model is calculated. According to the calculated prediction error, the initial breast cancer recurrence risk prediction model is trained and optimized. According to the size of the prediction error, when the prediction error is minimum, the training of the initial breast cancer recurrence risk prediction model is completed, and the breast cancer recurrence risk prediction model is obtained.

[0050] Multiple breast cancer recurrence risk training samples can also be grouped to obtain training group breast cancer recurrence risk training samples and validation group breast cancer recurrence risk training samples, and all breast cancer recurrence risk training samples in the training group breast cancer recurrence risk training samples are used to train the initial breast cancer recurrence risk prediction model. The predicted recurrence risk probability corresponding to all breast cancer recurrence risk training samples in the training group breast cancer recurrence risk training samples is obtained by weighted calculation of the genotype status of multiple related single nucleotide polymorphism variation sites and multiple pathological characteristics carried by all breast cancer recurrence risk training samples in the training group breast cancer recurrence risk training samples. According to the predicted recurrence risk probability and the actual recurrence risk probability corresponding to all breast cancer recurrence risk training samples in the training group breast cancer recurrence risk training samples, the pre-constructed initial breast cancer recurrence risk prediction model is trained to obtain a breast cancer recurrence risk prediction model.

[0051] In an optional example, 893 breast cancer recurrence risk training samples are grouped to obtain a training group of breast cancer recurrence risk training samples containing 446 breast cancer recurrence risk training samples and a validation group of breast cancer recurrence risk training samples containing 447 breast cancer recurrence risk training samples. The 446 breast cancer recurrence risk training samples in the training group of breast cancer recurrence risk training samples are used to train the initial breast cancer recurrence risk prediction model, and the predicted recurrence risk probability corresponding to the breast cancer recurrence risk training samples is calculated.

[0052] In a possible implementation, step S103, training a pre-constructed initial breast cancer recurrence risk prediction model according to the predicted recurrence risk probabilities and actual recurrence risk probabilities corresponding to each of a plurality of breast cancer recurrence risk training samples to obtain a breast cancer recurrence risk prediction model, includes: According to the predicted recurrence risk probability and the actual recurrence risk probability corresponding to each of the multiple breast cancer recurrence risk training samples, the prediction error of the initial breast cancer recurrence risk prediction model is calculated by cross-validation method; According to the prediction error of the initial breast cancer recurrence risk prediction model, the adjustment parameters of the initial breast cancer recurrence risk prediction model are optimized and adjusted; Determine the adjustment parameter when the prediction error of the initial breast cancer recurrence risk prediction model is the smallest as the optimal adjustment parameter; According to the optimal adjustment parameters, the optimal weights of the genotype status of each relevant single nucleotide polymorphism variation site and the information of each pathological feature are determined respectively; A breast cancer recurrence risk prediction model is determined based on the optimal weight of the genotype status of all relevant single nucleotide polymorphism variation sites and the information of all pathological characteristics.

[0053] Specifically, according to the predicted recurrence risk probabilities and actual recurrence risk probabilities corresponding to each of the multiple breast cancer recurrence risk training samples, the pre-constructed initial breast cancer recurrence risk prediction model is trained to obtain a breast cancer recurrence risk prediction model, including: The LASSO (least absolute shrinkage and selection operator) regression model was used to construct an initial breast cancer recurrence risk prediction model. By adjusting the adjustment parameter λ, the absolute value of the weight of the variable feature was penalized, so that the weights of some unimportant variable features became zero. The variable features were automatically selected, overfitting was avoided, and the calculation of the initial breast cancer recurrence risk prediction model was simplified.

[0054] A plurality of adjustment parameters λ are preset, and for each adjustment parameter λ: a plurality of breast cancer recurrence risk training samples are grouped, and the grouped plurality of breast cancer recurrence risk training samples are input into an initial breast cancer recurrence risk prediction model in batches to obtain predicted recurrence risk probabilities of the plurality of breast cancer recurrence risk training samples, and the predicted recurrence risk probability corresponding to the adjustment parameter λ is calculated according to the predicted recurrence risk probabilities of the plurality of breast cancer recurrence risk training samples.

[0055] According to the predicted recurrence risk probability and the actual recurrence risk probability corresponding to the adjustment parameter λ, the prediction error of the initial breast cancer recurrence risk prediction model corresponding to the adjustment parameter λ is calculated.

[0056] Through the cross-validation method, the adjustment parameter λ of the initial breast cancer recurrence risk prediction model is optimized and adjusted according to the prediction error of the initial breast cancer recurrence risk prediction model corresponding to each adjustment parameter λ.

[0057] The adjustment parameter λ of the initial breast cancer recurrence risk prediction model with the smallest prediction error is selected, and the adjustment parameter λ of the initial breast cancer recurrence risk prediction model with the smallest prediction error is taken as the optimal adjustment parameter.

[0058] According to the optimal adjustment parameters, the optimal weights of the genotype status of each relevant single nucleotide polymorphism variation site and the information of each pathological feature are determined respectively. When the optimal weight of a variable feature is 0, it means that the variable feature is an unimportant variable feature. When the optimal weight of a variable feature is not 0, it means that the variable feature is the best variable feature.

[0059] According to the optimal weight of the genotype status of all relevant single nucleotide polymorphism variation sites and the information of all pathological characteristics, a calculation formula for predicting the probability of recurrence risk is determined, and a breast cancer recurrence risk prediction model is determined.

[0060] In an optional example, if Figure 3 As shown, Figure 3This is a schematic diagram of the analysis results of an optimal adjustment parameter provided by an embodiment of the present application. Through the cross-validation method, the optimal variable feature selection is performed on all relevant single nucleotide polymorphism variation sites and all pathological features carried in 446 breast cancer recurrence risk training samples, and the adjustment parameter λ (logarithm of the penalty coefficient) that minimizes the cross-validation prediction error is determined to be 0.001739049, and 0.001739049 is the optimal adjustment parameter. For example: the optimal variable feature includes 8 relevant single nucleotide polymorphism sites and 8 pathological features, which are used to predict the probability of recurrence risk. The eight related SNPs included: rs2285892, rs554098, rs372637371, rs113756220, rs17489363, rs4140390, rs5901000 and rs111033566; the eight pathological features included: histological grade, tumor category, lymph node metastasis, TNM stage, PR status, HER-2 status, Ki-67 proliferation index and molecular subtype. The weight (coefficient) of the ER status feature was set to 0.

[0061] Suppose that the genotype status of all relevant single nucleotide polymorphism variation sites and all pathological characteristics of a breast cancer recurrence risk training sample are obtained. According to the optimal weights of the genotype status of the relevant single nucleotide polymorphism variation sites and the information of pathological characteristics, the predicted recurrence risk probability of the breast cancer recurrence risk training sample is calculated according to the following formula: Predicted recurrence risk probability = (0.5361*histological grade III) + (0.9367* T2) + (1.8775*T3 / 4) + (0.6012*lymph node positive) - (1.3432*TNM.II) - (1.1226*TNM.III) + (0.5726*PR positive) - (0.7857*Her-2 positive) + (0.0956*Ki-67 proliferation index high) - (3.1425* Luminal A type) + (0.8960* TNBC type) - (0.5332*rs2285892) - (0.9467*rs554098) + (0.5231*rs372637371) - (0.2451*rs113756220) + (0.5345*rs17489363) + (0.7679*rs4140390) + (0.7968*rs5901000) + (0.9801*rs111033566).

[0062] In a possible implementation, the related single nucleotide polymorphism variation sites in step S11 include: related single nucleotide polymorphism variation site rs2285892, related single nucleotide polymorphism variation site rs554098, related single nucleotide polymorphism variation site rs372637371, related single nucleotide polymorphism variation site rs113756220, related single nucleotide polymorphism variation site rs17489363, related single nucleotide polymorphism variation site rs4140390, related single nucleotide polymorphism variation site rs5901000 and related single nucleotide polymorphism variation site rs111033566; Pathological characteristics include: histological grade characteristics, tumor category characteristics, lymph node metastasis characteristics, TNM staging characteristics, ER status characteristics, PR status characteristics, HER-2 status characteristics, Ki-67 proliferation index characteristics and molecular subtype characteristics.

[0063] Specifically, see Figure 2 , Figure 21 is a schematic diagram of the analysis results of a related single nucleotide polymorphism variation site provided in an embodiment of the present application, the genotype status of the related single nucleotide polymorphism variation site rs2285892 includes GG, GA and AA, the genotype status of the related single nucleotide polymorphism variation site rs554098 includes CC, TC and TT, the genotype status of the related single nucleotide polymorphism variation site rs372637371 includes CC and CT, the genotype status of the related single nucleotide polymorphism variation site rs113756220 includes CC and CT, and the genotype status of the related single nucleotide polymorphism variation site rs113756221 includes CC and CT. The genotypes of the relevant single nucleotide polymorphism variation site rs17489363 include GG, AG and AA, the genotypes of the relevant single nucleotide polymorphism variation site rs4140390 include CC, CG and GG, the genotypes of the relevant single nucleotide polymorphism variation site rs5901000 include GG and GGAA, and the genotypes of the relevant single nucleotide polymorphism variation site rs111033566 include AA and AT.

[0064] Pathological characteristics include histological grade characteristics, tumor category characteristics, lymph node metastasis characteristics, TNM staging characteristics, ER status characteristics, PR status characteristics, HER-2 status characteristics, Ki-67 proliferation index characteristics and molecular subtype characteristics.

[0065] Among them, the degree of differentiation of histological grading characteristics is directly related to the prognosis and recurrence risk of breast cancer. The information of histological grading characteristics includes grade I, grade II, and grade III. The higher the grade, the higher the risk of recurrence and the worse the prognosis.

[0066] The tumor category feature indicates the size of the tumor. The information of the tumor category feature includes less than or equal to 2 cm, greater than 2 cm and not greater than 5 cm, and greater than 5 cm. The larger the tumor, the greater the risk of recurrence.

[0067] The lymph node metastasis feature indicates the number of lymph node metastases. The more the number, the worse the prognosis of breast cancer. Taking the ipsilateral axillary lymph node metastasis as an example, the information of the lymph node metastasis feature includes low risk, medium risk and high risk. In addition, the location of the metastatic lymph node is also related to the prognosis. The higher the metastasis location, the worse the prognosis.

[0068] The TNM staging system provides information on the staging of breast cancer, which comprehensively evaluates the status of the tumor (T), lymph nodes (N), and distant metastasis (M).

[0069] The ER status feature indicates the level of estrogen, and the information of the ER status feature is expressed in percentages. The PR status feature indicates the level of progesterone, and the information of the PR status feature is also expressed in percentages.

[0070] Information on HER-2 status characteristics includes - or 0 (negative), 1+ (weakly positive), 2+ (needs FISH test confirmation) and 3+ (definitely positive).

[0071] The higher the Ki-67 proliferation index characteristic, the more active the tumor proliferation and the worse the prognosis of breast cancer.

[0072] Information on molecular subtype characteristics includes Luminal A, Luminal B, HER2-enriched, and Basal-like subtypes.

[0073] In a possible implementation, in step S12, the recurrence risk of the sample to be predicted for breast cancer recurrence risk is predicted by using a breast cancer recurrence risk prediction model to obtain a predicted recurrence risk probability corresponding to the sample to be predicted for breast cancer recurrence risk, including: Inputting the breast cancer recurrence risk sample to be predicted into the breast cancer recurrence risk prediction model, wherein the breast cancer recurrence risk sample to be predicted carries the genotype status of multiple related single nucleotide polymorphism variation sites and information of multiple pathological characteristics; Normalizing the genotype states of multiple relevant single nucleotide polymorphism variation sites and the information of multiple pathological characteristics carried by the samples to be predicted for the risk of breast cancer recurrence, respectively, to obtain the genotype states of multiple normalized relevant single nucleotide polymorphism variation sites and the information of multiple normalized pathological characteristics; According to the optimal weight of the genotype status of each relevant single nucleotide polymorphism variation site and each pathological feature information, the genotype status of all normalized relevant single nucleotide polymorphism variation sites and all normalized pathological feature information are weightedly calculated to obtain the corresponding predicted recurrence risk probability of the breast cancer recurrence risk sample to be predicted.

[0074] Specifically, the breast cancer recurrence risk sample to be predicted is input into the breast cancer recurrence risk prediction model. The breast cancer recurrence risk sample to be predicted carries the genotype status of multiple related single nucleotide polymorphism variation sites and information on multiple pathological characteristics.

[0075] The genotype status of multiple relevant single nucleotide polymorphism variation sites and multiple pathological characteristics of the samples to be predicted for breast cancer recurrence risk are normalized to obtain multiple normalized genotype status of relevant single nucleotide polymorphism variation sites and multiple normalized pathological characteristics. The purpose of normalization is to facilitate the processing of variable feature information and provide convenience for subsequent prediction of recurrence risk probability.

[0076] According to the optimal weights of the genotype status of each relevant single nucleotide polymorphism variation site and each pathological feature information calculated above, the genotype status of all normalized relevant single nucleotide polymorphism variation sites and all normalized pathological feature information are weightedly calculated to obtain the corresponding predicted recurrence risk probability of the breast cancer recurrence risk sample to be predicted.

[0077] In a possible implementation, the genotype states of multiple relevant single nucleotide polymorphism variation sites carried by each breast cancer recurrence risk training sample are obtained by the following method: Obtaining multiple relevant single nucleotide polymorphism variation sites associated with breast cancer prognosis, and specific detection primers corresponding to each relevant single nucleotide polymorphism variation site; Acquire genomic DNA of multiple breast cancer recurrence risk training samples, wherein the genomic DNA of the breast cancer recurrence risk training samples is obtained by extracting peripheral blood of the breast cancer recurrence risk training samples; Using specific detection primers corresponding to each relevant single nucleotide polymorphism variation site, molecular biological methods are used to detect the genotype status of multiple relevant single nucleotide polymorphism variation sites in the genomic DNA of each breast cancer recurrence risk training sample, and the genotype status corresponding to each of the multiple relevant single nucleotide polymorphism variation sites of the breast cancer recurrence risk training sample is obtained.

[0078] Specifically, the genotype status of multiple relevant single nucleotide polymorphism variation sites carried by each breast cancer recurrence risk training sample is obtained by the following method: Based on the risk association with breast cancer prognosis, multiple relevant single nucleotide polymorphism variation sites associated with breast cancer prognosis were identified.

[0079] Each single nucleotide polymorphism variation site corresponds to a specific detection primer, and the specific detection primer corresponding to each of the above-mentioned related single nucleotide polymorphism variation sites is determined.

[0080] By extracting peripheral blood of breast cancer recurrence risk training samples, genomic DNA of multiple breast cancer recurrence risk training samples can be obtained.

[0081] For each breast cancer recurrence risk training sample, each relevant single nucleotide polymorphism variation site, a specific detection primer corresponding to the relevant single nucleotide polymorphism variation site is used to detect the genotype of the relevant single nucleotide polymorphism variation site in the genomic DNA of the breast cancer recurrence risk training sample by molecular biology methods to obtain the genotype state corresponding to the relevant single nucleotide polymorphism variation site of the breast cancer recurrence risk training sample. Similarly, the genotype state corresponding to each relevant single nucleotide polymorphism variation site of the breast cancer recurrence risk training sample is obtained by molecular biology methods.

[0082] According to the same method as above, the genotype status corresponding to each relevant single nucleotide polymorphism variation site of each breast cancer recurrence risk training sample is obtained.

[0083] The molecular biology method may include any one of Sanger sequencing, TaqMan probe method, ARMS-PCR method, gene chip method, mass spectrometry and the like.

[0084] In a possible implementation, multiple relevant single nucleotide polymorphism variation sites associated with breast cancer prognosis are obtained, including: Obtain multiple candidate single nucleotide polymorphism variation sites; The risk association between each candidate single nucleotide polymorphism variation site and breast cancer prognosis was calculated through COX univariate and multivariate regression analysis; According to the risk association between each candidate single nucleotide polymorphism variation site and breast cancer prognosis, relevant single nucleotide polymorphism variation sites associated with breast cancer prognosis are selected from multiple candidate single nucleotide polymorphism variation sites.

[0085] Specifically, refer to Figure 2 For hundreds of candidate single nucleotide polymorphism variant sites, Cox univariate analysis (survival analysis method) and multivariate regression analysis were performed to analyze the genotype hazard ratio ( Figure 2 The risk ratio in the above table is used to calculate the risk association between each candidate single nucleotide polymorphism variation site and breast cancer prognosis, such as Figure 2 The p-value in . Select all candidate single nucleotide polymorphism variant sites with p-value less than 0.05 as the relevant single nucleotide polymorphism variant sites related to breast cancer prognosis. For example: select Figure 2 8 single nucleotide polymorphism sites in .

[0086] Among them, when calculating the genotype hazard ratio of the genotype state of each candidate single nucleotide polymorphism variation site, the selected confidence interval ( Figure 2 The confidence interval (CI) in the HR estimate is 95%. A wider confidence interval indicates a greater uncertainty in the estimate, while a narrower confidence interval indicates a more precise estimate.

[0087] In a possible implementation, the method further includes: According to the genotype status of 8 related single nucleotide polymorphism sites and the information of 8 pathological characteristics, the above breast cancer recurrence risk prediction model was used to calculate the predicted recurrence risk probability of the above 446 training group breast cancer recurrence risk training samples, and the predicted recurrence risk probability ranged from 0.009 to 35.689. The median of 0.992 was selected as the cut-off value, and the 446 training group breast cancer recurrence risk training samples were divided into a low-risk group and a high-risk group. Figure 4 As shown, Figure 4 : is a schematic diagram of the prediction accuracy of a breast cancer recurrence risk prediction model in a training group provided by an embodiment of the present application. In the training group, the survival of the low-risk group is significantly better than that of the high-risk group ( Figure 4 A in Figure 1) The time-dependent ROC curve was used to evaluate the accuracy of the breast cancer recurrence risk prediction model. The AUC of the 5-year survival breast cancer recurrence risk prediction model was 0.82 ( Figure 4 B in the figure).

[0088] In a possible implementation, the method further includes: validating the breast cancer recurrence risk prediction model.

[0089] 447 validation group breast cancer recurrence risk training samples were used, and the breast cancer recurrence risk prediction model was used to verify the above multiple samples. The predicted recurrence risk probability ranged from 0.009 to 53.486. The median of 2.46 was selected as the cut-off value, and the 447 validation group breast cancer recurrence risk training samples were divided into a low-risk group and a high-risk group. Figure 5 As shown, Figure 5 This is a schematic diagram of the prediction accuracy of a breast cancer recurrence risk prediction model validation group provided in an embodiment of the present application. The survival of patients in the low-risk group is significantly better than that in the high-risk group ( Figure 5 A in the figure), the time-dependent ROC curve was used to evaluate the accuracy of the breast cancer recurrence risk prediction model, and the AUC value of the breast cancer recurrence risk prediction model for 5-year survival was 0.74 ( Figure 5 B in the figure).

[0090] In a possible implementation, the method further includes: comparatively evaluating the performance of multiple models.

[0091] like Figure 6 As shown, Figure 6: This is a schematic diagram of the comparative analysis of the prediction performance of a breast cancer recurrence risk prediction model provided by an embodiment of the present application and two other breast cancer recurrence risk prediction models. A LASSO Cox regression model is established using 8 SNPs, and another LASSO Cox regression model is established using 8 clinical pathological features. The time-dependent ROC curves are used to obtain the 5-year survival AUCs of the two LASSO Cox regression models in the training group, which are 0.72 and 0.75 respectively. Comparing the ROC curves of the breast cancer recurrence risk prediction model in this embodiment and the two LASSO Cox regression models, the P values ​​are 0.001 and 0.008 respectively ( Figure 6 A in the figure). The time-dependent ROC curve was used to obtain the AUC of 5-year survival in the validation set, which were 0.69 and 0.68, respectively. The ROC curves of the breast cancer recurrence risk prediction model in this embodiment and the two LASSO Cox regression models were compared, and the P values ​​were 0.221 and 0.014, respectively. Figure 6 It is obvious that the breast cancer recurrence risk prediction model in this embodiment is significantly better than the model using SNP or pathological features alone.

[0092] It can be understood that the technical solution provided in this embodiment obtains a breast cancer recurrence risk prediction model. Since the breast cancer recurrence risk prediction model is obtained by training a pre-constructed initial breast cancer recurrence risk prediction model based on multiple breast cancer recurrence risk training samples, the trained breast cancer recurrence risk prediction model contains information of multiple breast cancer recurrence risk training samples, and each breast cancer recurrence risk training sample carries the genotype status of multiple related single nucleotide polymorphism variation sites, information of multiple pathological characteristics and actual recurrence risk probability, so that the breast cancer recurrence risk prediction model contains richer information, and improves the prediction accuracy of the breast cancer recurrence risk prediction model for breast cancer prognosis recurrence risk; through the breast cancer recurrence risk prediction model, the recurrence risk of the breast cancer recurrence risk sample to be predicted is predicted, so that the predicted recurrence risk probability corresponding to the obtained breast cancer recurrence risk sample to be predicted is more accurate, and the accuracy of the predicted recurrence risk probability corresponding to the breast cancer recurrence risk sample to be predicted is improved, thereby solving the existing problem of low prediction accuracy when predicting the prognosis recurrence risk of breast cancer patients by a single molecular marker.

[0093] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0094] Corresponding to the method described in the above embodiment, Figure 7A schematic structural diagram of a device for predicting the risk of recurrence of breast cancer provided in an embodiment of the present application is shown. For ease of explanation, only the parts related to the embodiment of the present application are shown.

[0095] Reference Figure 7 The device comprises: a model acquisition module 71, which is used to acquire a breast cancer recurrence risk prediction model, wherein the breast cancer recurrence risk prediction model is obtained by training a pre-constructed initial breast cancer recurrence risk prediction model based on a plurality of breast cancer recurrence risk training samples, each breast cancer recurrence risk training sample carries the genotype status of a plurality of related single nucleotide polymorphism variation sites, information of a plurality of pathological characteristics and an actual recurrence risk probability, and the related single nucleotide polymorphism variation sites represent single nucleotide polymorphism variation sites in the peripheral blood of the breast cancer recurrence risk training samples that are related to breast cancer prognosis; The risk prediction module 72 is used to predict the recurrence risk of the breast cancer recurrence risk sample to be predicted by using the breast cancer recurrence risk prediction model, and obtain the predicted recurrence risk probability corresponding to the breast cancer recurrence risk sample to be predicted.

[0096] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0097] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0098] An embodiment of the present application also provides a device for predicting the risk of recurrence of breast cancer. The terminal device includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor. When the processor executes the computer program, the steps in any of the above method embodiments are implemented.

[0099] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0100] An embodiment of the present application provides a computer program product. When the computer program product runs on a mobile terminal, the mobile terminal can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0101] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can at least include: any entity or device that can carry the computer program code to the camera / terminal device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, RandomAccess Memory), electric carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electric carrier signals and telecommunication signals.

[0102] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0103] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0104] In the embodiments provided in the present application, it should be understood that the disclosed devices / network equipment and methods can be implemented in other ways. For example, the device / network equipment embodiments described above are merely schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0105] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0106] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for predicting the risk of breast cancer recurrence, characterized in that: include: Obtaining a breast cancer recurrence risk prediction model, wherein the breast cancer recurrence risk prediction model is obtained by training a pre-constructed initial breast cancer recurrence risk prediction model based on multiple breast cancer recurrence risk training samples, each of the breast cancer recurrence risk training samples carries the genotype status of multiple related single nucleotide polymorphism variation sites, information on multiple pathological characteristics, and an actual recurrence risk probability, and the related single nucleotide polymorphism variation sites represent single nucleotide polymorphism variation sites in the peripheral blood of the breast cancer recurrence risk training samples that are related to breast cancer prognosis; The breast cancer recurrence risk prediction model is used to predict the recurrence risk of the breast cancer recurrence risk sample to be predicted, and the predicted recurrence risk probability corresponding to the breast cancer recurrence risk sample to be predicted is obtained.

2. The method for predicting the risk of breast cancer recurrence according to claim 1, wherein: Before obtaining the breast cancer recurrence risk prediction model, the method includes: Acquire a plurality of breast cancer recurrence risk training samples; Performing weighted calculations on the genotype states of the multiple related single nucleotide polymorphism variation sites and the information of the multiple pathological characteristics carried by each of the breast cancer recurrence risk training samples, respectively, to obtain a predicted recurrence risk probability corresponding to each of the breast cancer recurrence risk training samples; The pre-constructed initial breast cancer recurrence risk prediction model is trained according to the predicted recurrence risk probabilities and the actual recurrence risk probabilities corresponding to each of the plurality of breast cancer recurrence risk training samples to obtain the breast cancer recurrence risk prediction model.

3. The method for predicting the risk of breast cancer recurrence according to claim 2, wherein: The pre-constructed initial breast cancer recurrence risk prediction model is trained according to the predicted recurrence risk probabilities and the actual recurrence risk probabilities corresponding to each of the plurality of breast cancer recurrence risk training samples to obtain the breast cancer recurrence risk prediction model, including: According to the predicted recurrence risk probabilities and the actual recurrence risk probabilities corresponding to each of the plurality of breast cancer recurrence risk training samples, a prediction error of the initial breast cancer recurrence risk prediction model is calculated by a cross-validation method; According to the prediction error of the initial breast cancer recurrence risk prediction model, optimizing and adjusting the adjustment parameters of the initial breast cancer recurrence risk prediction model; Determine the adjustment parameter when the prediction error of the initial breast cancer recurrence risk prediction model is the smallest as the optimal adjustment parameter; According to the optimal adjustment parameters, respectively determining the optimal weight of the genotype status of each of the relevant single nucleotide polymorphism variation sites and the information of each of the pathological characteristics; The breast cancer recurrence risk prediction model is determined according to the optimal weights of the genotype status of all the relevant single nucleotide polymorphism variation sites and the information of all the pathological characteristics.

4. The method for predicting the risk of breast cancer recurrence according to any one of claims 1 to 3, characterized in that: The related single nucleotide polymorphism variation sites include: related single nucleotide polymorphism variation site rs2285892, related single nucleotide polymorphism variation site rs554098, related single nucleotide polymorphism variation site rs372637371, related single nucleotide polymorphism variation site rs113756220, related single nucleotide polymorphism variation site rs17489363, related single nucleotide polymorphism variation site rs4140390, related single nucleotide polymorphism variation site rs5901000 and related single nucleotide polymorphism variation site rs111033566; The pathological characteristics include: histological grade characteristics, tumor category characteristics, lymph node metastasis characteristics, TNM staging characteristics, ER status characteristics, PR status characteristics, HER-2 status characteristics, Ki-67 proliferation index characteristics and molecular subtype characteristics.

5. The method for predicting the risk of breast cancer recurrence according to claim 3, wherein: The method of predicting the recurrence risk of the breast cancer recurrence risk sample to be predicted by using the breast cancer recurrence risk prediction model to obtain the predicted recurrence risk probability corresponding to the breast cancer recurrence risk sample to be predicted includes: Inputting the breast cancer recurrence risk sample to be predicted into the breast cancer recurrence risk prediction model, wherein the breast cancer recurrence risk sample to be predicted carries the genotype status of multiple related single nucleotide polymorphism variation sites and information of multiple pathological characteristics; Normalizing the genotype states of the multiple related single nucleotide polymorphism variation sites and the information of the multiple pathological characteristics carried by the sample to be predicted for the risk of breast cancer recurrence, respectively, to obtain the genotype states of the multiple normalized related single nucleotide polymorphism variation sites and the information of the multiple normalized pathological characteristics; According to the optimal weight of the genotype status of each relevant single nucleotide polymorphism variation site and each pathological feature information, the genotype status of all the normalized relevant single nucleotide polymorphism variation sites and all the normalized pathological feature information are weightedly calculated to obtain the corresponding predicted recurrence risk probability of the breast cancer recurrence risk sample to be predicted.

6. The method for predicting the risk of breast cancer recurrence according to claim 3, wherein: The genotype states of the multiple relevant single nucleotide polymorphism variation sites carried by each of the breast cancer recurrence risk training samples are obtained by the following method: Obtaining a plurality of the related single nucleotide polymorphism variation sites associated with breast cancer prognosis, and specific detection primers corresponding to each of the related single nucleotide polymorphism variation sites; Acquire genomic DNA of a plurality of the breast cancer recurrence risk training samples, wherein the genomic DNA of the breast cancer recurrence risk training samples is obtained by extracting peripheral blood of the breast cancer recurrence risk training samples; Using specific detection primers corresponding to each of the relevant single nucleotide polymorphism variation sites, the genotype status of multiple relevant single nucleotide polymorphism variation sites in the genomic DNA of each of the breast cancer recurrence risk training samples is detected through molecular biology methods to obtain the genotype status corresponding to each of the multiple relevant single nucleotide polymorphism variation sites of the breast cancer recurrence risk training sample.

7. The method for predicting the risk of breast cancer recurrence according to claim 6, wherein: The step of obtaining a plurality of related single nucleotide polymorphism variation sites associated with breast cancer prognosis includes: Obtain multiple candidate single nucleotide polymorphism variation sites; The risk correlation between each candidate single nucleotide polymorphism variation site and breast cancer prognosis was calculated by COX univariate and multivariate regression analysis; According to the risk association between each candidate single nucleotide polymorphism variation site and the prognosis of breast cancer, relevant single nucleotide polymorphism variation sites related to the prognosis of breast cancer are selected from multiple candidate single nucleotide polymorphism variation sites.

8. A device for predicting the risk of breast cancer recurrence, characterized in that: include: A model acquisition module, used for acquiring a breast cancer recurrence risk prediction model, wherein the breast cancer recurrence risk prediction model is obtained by training a pre-constructed initial breast cancer recurrence risk prediction model based on a plurality of breast cancer recurrence risk training samples, each of the breast cancer recurrence risk training samples carries the genotype status of a plurality of relevant single nucleotide polymorphism variation sites, information of a plurality of pathological characteristics and an actual recurrence risk probability, and the relevant single nucleotide polymorphism variation sites represent single nucleotide polymorphism variation sites in the peripheral blood of the breast cancer recurrence risk training samples that are related to breast cancer prognosis; The risk prediction module is used to predict the recurrence risk of the breast cancer recurrence risk sample to be predicted by using the breast cancer recurrence risk prediction model, and obtain the predicted recurrence risk probability corresponding to the breast cancer recurrence risk sample to be predicted.

9. A device for predicting the recurrence risk of breast cancer, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images