Liver malignant tumor microvascular invasion identification method and system

Abstract

The invention provides a method and a system for identifying liver malignant tumor microvascular invasion. The method comprises the following steps: firstly, acquiring a tissue sample, and performing transcriptome or proteome sequencing to generate initial sequencing data; performing gene expression value sorting on the initial data, and determining sample sequencing data; differential genes are obtained through difference analysis, and the genes are selected as a sample set. The sample set is randomly distributed into a training set and a verification set, the training set is evaluated by adopting a ten-fold cross validation method, and an out-of-bag error rate is obtained to select a proper model hyper-parameter. On the basis, target genes are selected through preset importance ranking, expression values of the target genes are ranked, a random forest model is constructed, a verification set is used for verification, and a recognition model is obtained; and inputting a to-be-detected sample into the recognition model to obtain a recognition result. The method is independent of a specific detection platform and means, various sequencing data can be integrated for capillary invasion judgment, and a basis for timely intervention treatment is provided for clinical patients with high metastasis risk.

Description

Technical Field

[0001] The present invention relates to the field of intelligent medical technology, and in particular to a method and system for identifying microvascular invasion of malignant liver tumors. Background Art

[0002] Regarding the diagnosis and treatment of liver malignancies, surgical treatment is only suitable for a small number of patients, and the 5-year survival rate of patients after surgery is still unsatisfactory; other treatments include chemotherapy, targeted therapy, and immunotherapy, but the results are still unsatisfactory. The main reason for this phenomenon is metastasis and frequent recurrence, so there is a need for indicators that reflect early metastasis and recurrence and can serve as a basis for precision treatment.

[0003] Vascular structure is an important part of the tumor microenvironment. Microvascular invasion, also known as microvascular cancer thrombus, refers to the nesting of cancer cells in vascular endothelial cells. The presence of microvascular invasion is a key factor in early metastasis and recurrence of cancer. It is considered an independent prognostic factor for many cancers and can also serve as an important reference for treatment options such as chemotherapy, radiotherapy, transcatheter arterial chemoembolization, radionuclide therapy, and endoscopic resection.

[0004] Since microvascular invasion is a histopathological finding, it can only be diagnosed through postoperative surgical specimens. However, a considerable number of patients are not eligible for surgery at the time of initial onset, so specimens must be obtained through puncture to diagnose the disease. In addition, due to the limited information obtained from these specimens and the existence of sampling errors, it is challenging to evaluate the microvascular invasion status of these patients, which ultimately affects treatment decisions and clinical outcomes, and has an unsatisfactory impact on the selection of therapeutic drugs for patients receiving neoadjuvant therapy or advanced patients receiving radiotherapy, as well as the prediction of the efficacy of interventional therapy. Artificial intelligence is currently showing a trend of deep penetration into the medical field, playing an auxiliary role that cannot be ignored. In order to overcome the limitations of microvascular invasion assessment in the above-mentioned cases, using artificial intelligence to build a model from molecular features related to microvascular invasion to predict the occurrence of microvascular invasion may be a better solution. Summary of the invention

[0005] In order to overcome the deficiencies of the prior art, the object of the present invention is to provide a method and system for identifying microvascular invasion of liver malignant tumors.

[0006] To achieve the above object, the present invention provides the following solutions:

[0007] A method for identifying microvascular invasion of liver malignant tumors, comprising:

[0008] Obtaining a tissue sample, and performing transcriptome sequencing or proteome sequencing on the tissue sample to obtain initial sequencing data, and sorting the initial sequencing data by gene expression value to obtain gene expression rank, and determining the gene expression rank or the original value of the gene expression data as the sample sequencing data;

[0009] Perform differential analysis on all gene expression ranks or gene expression data in the sample sequencing data to obtain differential genes, and then select data of differential genes from all genes in the sample sequencing data as a sample set;

[0010] Randomly allocating the sample set according to a preset ratio to obtain a training set and a validation set;

[0011] Performing a ten-fold cross validation on the training set to obtain an out-of-bag error rate result, and selecting corresponding model hyperparameters according to the out-of-bag error rate result;

[0012] Select variables according to the preset importance ranking to obtain the target gene;

[0013] Sorting the expression values ​​of the target genes to obtain gene expression ranks;

[0014] A random forest model is constructed using the gene expression rank and the model hyperparameters, and the validation set is used to validate the results to obtain a validated recognition model;

[0015] The sample to be tested is input into the recognition model to obtain a recognition result.

[0016] Preferably, the target genes include: RANBP1, ARMCX3, RAMP3, CDH17, CTHRC1, TARS, MET, ASRGL1, SLC16A2, NDRG1, COG1, VPS45, PPP1R14A, RAN, TAF9, PCK1, CPXM2, ZNF280D, ST6GAL1, PCM1, CSTA, JMJD6, RCL1, UCK2, ZNF687, GTF2F2, RCCD1, CCDC127, FRK, NAT10, C1orf43, FRYLDUS1L, TTC7B.

[0017] Preferably, the preset ratio is 7:3.

[0018] A system for identifying microvascular invasion of liver malignant tumors, comprising:

[0019] A sample acquisition module is used to acquire tissue samples, and perform transcriptome sequencing or proteome sequencing on the tissue samples to obtain initial sequencing data, and sort the initial sequencing data by gene expression value to obtain gene expression rank, and determine the gene expression rank or the original value of the gene expression data as sample sequencing data;

[0020] A difference analysis module is used to perform a difference analysis on all gene expression ranks or gene expression data in the sample sequencing data to obtain differential genes, and then select data of differential genes from all genes in the sample sequencing data as a sample set;

[0021] A data set allocation module is used to randomly allocate the sample set according to a preset ratio to obtain a training set and a validation set;

[0022] A parameter selection module performs a ten-fold cross validation on the training set to obtain an out-of-bag error rate result, and selects corresponding model hyperparameters according to the out-of-bag error rate result;

[0023] The gene acquisition module is used to select variables according to the preset importance ranking to obtain the target gene;

[0024] A gene ranking module, used to sort the expression values ​​of the target genes to obtain gene expression ranks;

[0025] A model building module, used to build a random forest model using the gene expression rank and the model hyperparameters, and use the validation set to verify the results to obtain a verified recognition model;

[0026] The recognition module is used to input the sample to be tested into the recognition model to obtain the recognition result.

[0027] An electronic device, comprising:

[0028] at least one processor; and

[0029] a memory communicatively connected to the at least one processor; wherein,

[0030] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the above-mentioned method for identifying microvascular invasion of liver malignant tumors.

[0031] A non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to execute the above-mentioned method for identifying microvascular invasion of liver malignant tumors.

[0032] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0033] The present invention provides a method and system for identifying microvascular invasion of malignant liver tumors. The method comprises: obtaining a tissue sample, and performing transcriptome sequencing or proteome sequencing on the tissue sample to obtain initial sequencing data, and respectively sorting the initial sequencing data according to the expression value of a gene to obtain a gene expression rank, and determining the gene expression rank or the original value of the gene expression data as the sample sequencing data; performing differential analysis on all gene expression ranks or gene expression data in the sample sequencing data to obtain differential genes, and then selecting data of differential genes from all genes in the sample sequencing data as a sample set; randomly allocating the sample set according to a preset ratio to obtain a training set and a validation set; performing a ten-fold cross validation on the training set to obtain an out-of-bag error rate result, and selecting a corresponding model hyperparameter according to the out-of-bag error rate result; selecting variables according to a preset importance ranking to obtain a target gene; sorting the expression value of the target gene to obtain a gene expression rank; using the gene expression rank and the model hyperparameter to construct a random forest model, and using the validation set to perform result validation to obtain a validated recognition model; inputting the sample to be tested into the recognition model to obtain a recognition result. The present invention does not rely on a specific detection platform or a specific detection method, so the invention can calculate the expression data of multiple gene combinations obtained by different sequencing methods to obtain microvascular invasion discrimination. The invention provides a basis for timely intervention and treatment of patients with a high risk of metastasis in clinical practice. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0035] Figure 1 A flow chart of a method provided by an embodiment of the present invention;

[0036] Figure 2 A schematic diagram of the system structure provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0037] 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.

[0038] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0039] Figure 1 A flow chart of a method provided by an embodiment of the present invention, such as Figure 1 As shown, the present invention provides a method for identifying microvascular invasion of liver malignant tumors, comprising:

[0040] Step 100: obtaining a tissue sample, and performing transcriptome sequencing or proteome sequencing on the tissue sample to obtain initial sequencing data, and sorting the initial sequencing data by gene expression value to obtain gene expression rank, and determining the gene expression rank or the original value of the gene expression data as sample sequencing data;

[0041] Step 200: performing differential analysis on all gene expression ranks or gene expression data in the sample sequencing data to obtain differential genes, and then selecting data of differential genes from all genes in the sample sequencing data as a sample set;

[0042] Step 300: randomly distribute the sample set according to a preset ratio to obtain a training set and a validation set;

[0043] Step 400: performing a ten-fold cross validation on the training set to obtain an out-of-bag error rate result, and selecting corresponding model hyperparameters according to the out-of-bag error rate result;

[0044] Step 500: Select variables according to the preset importance ranking to obtain target genes;

[0045] Step 600: sorting the expression values ​​of the target genes to obtain gene expression ranks;

[0046] Step 700: constructing a random forest model using the gene expression rank and the model hyperparameters, and using the validation set to verify the results to obtain a verified recognition model;

[0047] Step 800: Input the sample to be tested into the recognition model to obtain a recognition result.

[0048] The present invention provides a multi-gene combination for identifying microvascular invasion of liver malignant tumors and its application, and the multi-gene combination includes the following genes:

[0049] RANBP1, ARMCX3, RAMP3, CDH17, CTHRC1, TARS, MET, ASRGL1, SLC16A2, NDRG1, COG1, VPS45, PPP1R14A, RAN, TAF9, PCK1, CPXM2, ZNF280D, ST6GAL1, PCM1, CSTA, JMJD6, RCL1, UCK2, ZNF687, GTF2F2, RCCD1, CCDC127, FRK, NAT10, C1orf43, FRYLDUS1L, TTC7B.

[0050] The present invention further provides a specific method for obtaining the expression level of this multi-gene combination in a patient's tissue specimen: a fresh tissue sample of the tumor area obtained by surgery or puncture, or a paraffin tissue section sample or a paraffin-embedded sample of the patient's tumor site, and the sample is subjected to transcriptome sequencing or proteome sequencing. Model construction process: All samples are randomly assigned in a ratio of 7:3, with most samples used as training sets and other samples used as validation sets. According to the out-of-bag error rate results of the ten-fold cross-validation, the corresponding parameters (such as the depth of the tree, the number of trees, and the feature selection strategy) are selected, and the importance of the sample features is obtained according to the feature importance assessment, and the variables, i.e., the genes in the above table, are selected based on the importance ranking. The expression values ​​of these genes are sorted to obtain the gene expression rank. A random forest model was constructed using the gene expression rank and parameters, and the validation set was used to verify the results.

[0051] The present invention also provides a device for identifying and applying microvascular invasion of liver malignant tumors, which includes (1) providing a gene expression level analysis module. Obtaining the expression value of the above gene detected in the sample; (2) providing a sample discrimination module. Sorting the above gene expression levels to obtain gene expression ranks, and then incorporating them into a random forest model to obtain a prediction result, which is ultimately microvascular invasion or no microvascular invasion.

[0052] The present invention also provides a memory for storing a computer program.

[0053] The present invention also provides a processor for executing a computer program, which can implement the above technical content.

[0054] The present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the content of the above technology can be implemented.

[0055] The present invention also provides a device for detecting the expression of these genes, including a detection device targeting transcriptome and a detection device targeting proteome.

[0056] Corresponding to the above method, such as Figure 2As shown, this embodiment also provides a system for identifying microvascular invasion of liver malignant tumors, including:

[0057] A sample acquisition module is used to acquire tissue samples, and perform transcriptome sequencing or proteome sequencing on the tissue samples to obtain initial sequencing data, and sort the initial sequencing data by gene expression values ​​to obtain gene expression ranks, and determine the gene expression ranks or the original values ​​of the gene expression data as sample sequencing data;

[0058] A difference analysis module is used to perform a difference analysis on all gene expression ranks or gene expression data in the sample sequencing data to obtain differential genes, and then select data of differential genes from all genes in the sample sequencing data as a sample set;

[0059] A data set allocation module is used to randomly allocate the sample set according to a preset ratio to obtain a training set and a validation set;

[0060] A parameter selection module performs a ten-fold cross validation on the training set to obtain an out-of-bag error rate result, and selects corresponding model hyperparameters according to the out-of-bag error rate result;

[0061] The gene acquisition module is used to select variables according to the preset importance ranking to obtain the target gene;

[0062] A gene ranking module, used to sort the expression values ​​of the target genes to obtain gene expression ranks;

[0063] A model building module, used to build a random forest model using the gene expression rank and the model hyperparameters, and use the validation set to verify the results to obtain a verified recognition model;

[0064] The recognition module is used to input the sample to be tested into the recognition model to obtain the recognition result.

[0065] Specifically, this embodiment uses the gene expression data of liver malignant tumors with microvascular invasion and non-microvascular invasion to obtain differential genes, and further incorporates them into subsequent analysis steps. The differential genes are used to perform ten-fold cross validation on the training set to obtain the out-of-bag error rate results, and the corresponding model hyperparameters are selected according to the out-of-bag error rate results.

[0066] Furthermore, the gene expression data of microvascular invasion and non-microvascular invasion are the sample sequencing data mentioned above. That is, the expression values ​​or expression ranks of all genes are analyzed for differences, differential genes are obtained, and then the data of differential genes selected from all genes of the sample (gene expression values ​​or re-ranked gene expression ranks) are included in the subsequent steps.

[0067] The present invention provides a multi-gene combination for distinguishing microvascular invasion in patients with liver malignancies. The data of gene expression levels of patient sample tissues and a random forest algorithm based on gene expression ranks are used to predict microvascular invasion in liver malignancies. This method does not rely on a specific detection platform or a specific detection method, so the invention can calculate the expression data of multi-gene combinations obtained by different sequencing methods to obtain microvascular invasion discrimination. The invention provides a basis for timely intervention and treatment of patients with a high risk of metastasis in clinical practice.

[0068] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0069] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for identifying microvascular invasion of liver malignant tumors, characterized in that: include: Obtaining a tissue sample, and performing transcriptome sequencing or proteome sequencing on the tissue sample to obtain initial sequencing data, and sorting the initial sequencing data by gene expression value to obtain gene expression rank, and determining the gene expression rank or the original value of the gene expression data as the sample sequencing data; Perform differential analysis on all gene expression ranks or gene expression data in the sample sequencing data to obtain differential genes, and then select data of differential genes from all genes in the sample sequencing data as a sample set; Randomly allocating the sample set according to a preset ratio to obtain a training set and a validation set; Performing a ten-fold cross validation on the training set to obtain an out-of-bag error rate result, and selecting corresponding model hyperparameters according to the out-of-bag error rate result; Select variables according to the preset importance ranking to obtain the target gene; Sorting the expression values ​​of the target genes to obtain gene expression ranks; A random forest model is constructed using the gene expression rank and the model hyperparameters, and the validation set is used to validate the results to obtain a validated recognition model; The sample to be tested is input into the recognition model to obtain a recognition result.

2. The method for identifying microvascular invasion of liver malignant tumors according to claim 1, characterized in that: The target genes include: RANBP1, ARMCX3, RAMP3, CDH17, CTHRC1, TARS, MET, ASRGL1, SLC16A2, NDRG1, COG1, VPS45, PPP1R14A, RAN, TAF9, PCK1, CPXM2, ZNF280D, ST6GAL1, PCM1, CSTA, JMJD6, RCL1, UCK2, ZNF687, GTF2F2, RCCD1, CCDC127, FRK, NAT10, C1orf43, FRYLDUS1L, and TTC7B.

3. The method for identifying microvascular invasion of liver malignant tumors according to claim 1, characterized in that: The preset ratio is 7:

3.

4. A system for identifying microvascular invasion of liver malignant tumors, characterized in that: include: A sample acquisition module is used to acquire tissue samples, and perform transcriptome sequencing or proteome sequencing on the tissue samples to obtain initial sequencing data, and sort the initial sequencing data by gene expression values ​​to obtain gene expression ranks, and determine the gene expression ranks or the original values ​​of the gene expression data as sample sequencing data; A difference analysis module is used to perform a difference analysis on all gene expression ranks or gene expression data in the sample sequencing data to obtain differential genes, and then select data of differential genes from all genes in the sample sequencing data as a sample set; A data set allocation module is used to randomly allocate the sample set according to a preset ratio to obtain a training set and a validation set; A parameter selection module performs a ten-fold cross validation on the training set to obtain an out-of-bag error rate result, and selects corresponding model hyperparameters according to the out-of-bag error rate result; The gene acquisition module is used to select variables according to the preset importance ranking to obtain the target gene; A gene ranking module, used to sort the expression values ​​of the target genes to obtain gene expression ranks; A model building module, used to build a random forest model using the gene expression rank and the model hyperparameters, and use the validation set to verify the results to obtain a verified recognition model; The recognition module is used to input the sample to be tested into the recognition model to obtain the recognition result.

5. An electronic device, characterized in that: include: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method for identifying microvascular invasion of liver malignant tumors according to any one of claims 1 to 3.

6. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to enable a computer to execute the method for identifying microvascular invasion of liver malignant tumors according to any one of claims 1 to 3.

Images