A system for predicting the risk of tumor cell metastasis

Abstract

The present invention belongs to the field of bioinformatics, and specifically relates to a method and system for predicting the risk of tumor cell metastasis. The method of the present invention predicts the S of the tumor cell to be detected based on the gene expression information of the tumor cell to be detected. EMT , S Tra and S comm ; Among them, S EMT Used to indicate the epithelial-mesenchymal transition ability of tumor cells, S Travel Indicates the survival ability of tumor cells in blood circulation, S comm Indicates the ability of tumor cells to invade tissues or organs; then calculate the risk factor S of the tumor cells to be tested for metastasis Merta The method of the present invention can assist doctors in timely diagnosis and targeted treatment of cancer patients in the early stage of cancer metastasis, and can guide the research and development of new anti-cancer drugs, which is conducive to improving the prevention and treatment rate and cure rate of cancer.

Description

Technical Field

[0001] The present invention belongs to the field of bioinformatics, and in particular relates to a method and system for predicting the risk of tumor cell metastasis. Background Art

[0002] Cancer metastasis refers to the process of cancer spreading from the primary site to other tissues and organs. There is no doubt that cancer metastasis is the main cause of cancer-related deaths. Due to the defects of detection technology, cancer metastasis is difficult to detect by clinical means in the early stages of cancer metastasis, which means that the best time for treatment is often missed when cancer metastasis is discovered. Therefore, the prediction of the risk of cancer metastasis and the metastatic site through statistical and analytical means is of great significance to the treatment of cancer.

[0003] Currently, many clinical DNA sequencing technologies have been used to sequence the genomes of cancer patients. At the same time, existing technologies have also proposed metastasis prediction models based on copy number variation, tumor load factors, etc. However, these studies ignore the complex heterogeneity of the tumor microenvironment and lack research on the functions of different cell types in the process of cancer metastasis. Summary of the invention

[0004] With the development of second-generation sequencing technology (especially single-cell sequencing technology), the present invention proposes a more in-depth and interpretable model based on the transcriptome sequencing data of more primary and metastatic cancer patients, so as to predict the metastasis risk and organ preference of all cancer types and have a deeper understanding of the mechanism of cancer metastasis.

[0005] Specifically, based on the mechanism of cancer metastasis, the present invention sorted out the key influencing factors of cancer metastasis, and collected and mined multiple cancer transcriptome sequencing data. After statistical analysis and verification, a systems biology method was proposed to predict the possibility and distribution of cancer metastasis. The method of the present invention is the first prediction method based on new transcriptome sequencing data and applicable to all types of cancer. The prediction results obtained using the method of the present invention can accurately predict the risk of cancer metastasis in patients, as well as the risk coefficient of metastasis to different organs in patients.

[0006] Based on the above findings, the present invention first provides a method for predicting the risk of tumor cell metastasis, which comprises:

[0007] Obtaining gene expression information of tumor cells to be tested;

[0008] Predicting the S of the tumor cell to be tested based on the gene expression information EMT , S Travel and S comm ; Among them, S EMTUsed to indicate the epithelial-mesenchymal transition (EMT) ability of tumor cells. Travel Indicates the survival ability of tumor cells in blood circulation, S comm Indicates the ability of tumor cells to invade tissues or organs;

[0009] Then the risk factor S of the tumor cells to be tested for metastasis is calculated using the following formula: Meta :

[0010]

[0011] When the S Meta When it is higher, it is predicted to have a higher risk of metastasis.

[0012] The present invention further provides a method for predicting the organ metastasis tendency of tumor cells, which comprises: using the method for predicting the risk of tumor cell metastasis as described above;

[0013] Among them, the S of tumor cells invading different tissues or organs to be tested is obtained. comm When the tumor cells to be tested metastasize to a certain tissue or organ, comm When it is higher, it is predicted that the tumor cells have a higher tendency to metastasize to that tissue or organ.

[0014] The present invention further provides a method for predicting the prognosis of a tumor patient, comprising: using the method for predicting the risk of tumor cell metastasis as described above;

[0015] Among them, the S of the tumor cells to be tested is obtained Meta , when S Meta When it is higher, the tumor patient is predicted to have a worse prognosis.

[0016] In a second aspect, the present invention further provides a tumor cell detection system, comprising:

[0017] An acquisition module configured to acquire gene expression information of tumor cells to be tested;

[0018] A calculation module is configured to predict the S of the tumor cell to be tested based on the gene expression information. EMT , S Travel and S comm ; Among them, S EMT Used to indicate the epithelial-mesenchymal transition (EMT) ability of tumor cells. Travel Indicates the survival ability of tumor cells in blood circulation, S comm Indicates the ability of tumor cells to invade tissues or organs; then the risk factor S of the tumor cells to be tested for metastasis is calculated using the following formula Meta :

[0019] And, the output module is configured to be Meta When it is higher, it is predicted to have a higher risk of metastasis.

[0020] In a third aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the instructions are run on a computing device, the computer program is executed by a processor to implement: a method for predicting the risk of tumor cell metastasis as described above, or a method for predicting the organ metastasis tendency of tumor cells as described above, or a method for predicting the prognosis of tumor patients as described above.

[0021] The present invention further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for predicting the risk of tumor cell metastasis as described above, or the method for predicting the organ metastasis tendency of tumor cells as described above, or the method for predicting the prognosis of tumor patients as described above is implemented.

[0022] The present invention further provides a computer program product, which includes a computer program. When the computer program is read and executed by a computing device, the computing device executes: the method for predicting the risk of tumor cell metastasis as described above, or the method for predicting the organ metastasis tendency of tumor cells as described above, or the method for predicting the prognosis of tumor patients as described above.

[0023] The method of the present invention has at least the following advantages:

[0024] 1) Comprehensiveness: The present invention can widely predict the metastasis patterns of various different cancers to various different organs at the level of all cancer types, which is more comprehensive than previous studies on a single cancer type.

[0025] 2) Explainability: Based on the method of the present invention, it is possible to trace back cells and genes related to cancer metastasis that conform to biological laws, which has important biological research significance and value.

[0026] 3) Accuracy: The present invention has a high consistency with many years of clinical statistics and cognitive results in the prediction of most cancers and patients.

[0027] The method of the present invention can assist doctors in timely diagnosis and targeted treatment of cancer patients in the early stages of cancer metastasis, and can guide the development of new anti-cancer drugs, which is beneficial to improving the prevention and cure rates of cancer. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0029] Figure 1 Schematic diagram of the prediction principle in the embodiment of the present invention; in the figure, Figure A is a schematic diagram of the metastasis path of tumor cells; Figure B is the design idea of ​​the embodiment of the present invention.

[0030] Figure 2 1 is the distribution of the overall metastasis tendency of the population in the embodiment of the present invention; in the figure, Figure A is the predicted metastasis ratio; Figure B is the comparison between the predicted results and the clinical statistical results.

[0031] Figure 3 The results of the metastasis analysis of lung adenocarcinoma patients in the embodiment of the present invention are shown in Figure A. The metastasis of four patients predicted by the method of the present invention is shown in Figure B. The proportion of patients with high, medium and low metastasis tendency for each type of metastasis is shown in Figure C. The results of the survival analysis of people in different groups are shown in Figure C.

[0032] Figure 4 It is the analysis result of the single cell data set of colorectal cancer liver metastasis in the embodiment of the present invention; in the figure, Figure A is the classification result of the CD45- cell population; Figure B is the copy number variation of the cell population; Figure C and Figure D are the interaction analysis results of two cancer cell data sets with the normal liver microenvironment; Figure E is the ligand-receptor pair that plays an important interaction between cancer cells and endothelial cells, macrophages and neutrophils.

[0033] Figure 5 The results of the analysis of the single-cell dataset of lung adenocarcinoma liver metastasis in the embodiment of the present invention; in the figure, Figure A is the classification result of a single cell; Figure B is the copy number variation of the cell group; Figure C is the information of a sample; Figure D is the difference analysis results of the interaction between two cancer cell datasets and the normal bone microenvironment; Figure E is the ligand-receptor pair that plays an important interaction between cancer cells and endothelial cells, macrophages and neutrophils. Figure F is a comparison of the prediction results. DETAILED DESCRIPTION

[0034] The specific embodiments of the present invention are described in detail below. It should be understood that the specific embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention. Those skilled in the art may make various modifications and changes to the present invention without departing from the scope or spirit of the present invention. For example, a feature described or illustrated as part of one embodiment may be used in another embodiment to produce a further embodiment.

[0035] Unless otherwise indicated, the meaning of all terms (including technical and scientific terms) used to disclose the present invention is the same as that commonly understood by those of ordinary skill in the art to which the present invention belongs. By way of further guidance, the following definitions are used to better understand the teachings of the present invention. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention.

[0036] The terms "and / or", "or / and", and "and / or" used in this article include any one of two or more related listed items, and also include any and all combinations of related listed items, and the arbitrary and all combinations include any two related listed items, any more related listed items, or a combination of all related listed items. It should be noted that when at least three items are connected by at least two conjunctions selected from "and / or", "or / and", and "and / or", it should be understood that in this application, the technical solution undoubtedly includes technical solutions that are all connected by "logical and", and undoubtedly includes technical solutions that are all connected by "logical or". For example, "A and / or B" includes three parallel solutions of A, B and A+B. For example, the technical solution of "A, and / or, B, and / or, C, and / or, D" includes any one of A, B, C, and D (that is, the technical solution that is all connected by "logical OR"), and also includes any and all combinations of A, B, C, and D, that is, the combination of any two or any three of A, B, C, and D, and also includes the combination of four of A, B, C, and D (that is, the technical solution that is all connected by "logical AND").

[0037] As used herein, the terms "comprising", "including" and "comprising" are synonymous and are inclusive or open-ended and do not exclude additional, unrecited members, elements or method steps.

[0038] Numerical ranges expressed as endpoints herein include all numbers and fractions subsumed within the range, as well as the recited endpoints.

[0039] In the present invention, descriptions such as "plurality" and "multiple" refer to quantities greater than or equal to 2 unless otherwise specified.

[0040] In the present invention, the technical features described in an open manner include closed technical solutions composed of the listed features, and also include open technical solutions containing the listed features.

[0041] In the present invention, "preferably", "better", "more preferably", and "suitably" are only used to describe implementation methods or examples with better effects, and it should be understood that they do not constitute a limitation on the scope of protection of the present invention. In the present invention, "optionally", "optional", and "optional" refer to being optional, that is, any one of the two parallel schemes of "yes" or "no". If multiple "options" appear in a technical solution, unless otherwise specified and there is no contradiction or mutual restriction, each "optional" is independent of each other.

[0042] In the present invention, the term "tumor" refers to a new growth formed by the proliferation of local tissue cells under the action of various tumorigenic factors. Because this new growth is mostly a space-occupying block-like protrusion, it is also called a neoplasm. The "tumor" described in the present invention includes "benign tumors" and "malignant tumors". However, benign tumors generally do not metastasize and have a generally good prognosis. Therefore, the present invention can be more applied to the detection of malignant tumor cells.

[0043] The "malignant tumor" mentioned in this invention is also called "cancer", which is called cancer or tumor in traditional Chinese medicine. It refers to abnormal cell proliferation, and these proliferating cells may invade other parts of the body. It is a disease caused by the abnormal control of cell division and proliferation. In addition to uncontrolled division, cancer cells can also locally invade surrounding normal tissues and even metastasize to other parts of the body through the body's circulatory system or lymphatic system.

[0044] The "tumor cell metastasis" mentioned in the present invention refers to the process in which tumor cells invade the circulatory system from the original site of occurrence and migrate to other parts of the body to continue to grow. Usually, benign tumors will not produce distant metastasis, and the prognosis of patients with metastasis is very poor. Because cancer cells metastasize to various parts of the body, cancer treatment becomes more difficult.

[0045] In the present invention, the term "tumor microenvironment" refers to the surrounding microenvironment of tumor cells, including surrounding blood vessels, immune cells, fibroblasts, bone marrow-derived inflammatory cells, various signaling molecules and extracellular matrix. Tumors and the surrounding environment are closely related and constantly interact with each other. Tumors can affect their microenvironment by releasing cell signaling molecules, promoting tumor angiogenesis and inducing immune tolerance, while immune cells in the microenvironment can affect the growth and development of cancer cells.

[0046] Methods for predicting the risk of tumor cell metastasis

[0047] The present invention found that the following three steps have a key impact on the metastasis of tumor cells (such as Figure 1 As shown in A): First, tumor cells invade the surrounding area from the primary site and enter the blood circulation through the epithelial-mesenchymal transition (EMT) process. Many transcription factors related to the EMT process regulate this important process. Then the tumor cells enter the blood circulation system, hide from immune cells and survive, and reach other organs with the blood circulation. Finally, when the tumor cells reach other organs, they need to adapt to and invade the microenvironment in the new organ to complete the metastasis. According to several steps of cancer metastasis, the present invention further sorts out several key factors that determine cancer metastasis, such as epithelial cells, blood cells, tissue microenvironment, etc. for further research. By collecting data from multiple databases for statistical analysis, a technical prediction model is established, and the collected clinical statistical results are compared through the prediction model to verify the accuracy of the prediction results.

[0048] Based on the above findings, the present invention first provides a method for predicting the risk of tumor cell metastasis, which comprises:

[0049] Obtaining gene expression information of tumor cells to be tested;

[0050] Predicting the S of the tumor cell to be tested based on the gene expression information EMT , S Travel and S comm ; Among them, S EMT Used to indicate the epithelial-mesenchymal transition (EMT) ability of tumor cells. Travel Indicates the survival ability of tumor cells in blood circulation, S comm Indicates the ability of tumor cells to invade tissues or organs;

[0051] Then the risk factor S of the tumor cells to be tested for metastasis is calculated using the following formula: Meta :

[0052]

[0053] When the S Meta When it is higher, it is predicted to have a higher risk of metastasis.

[0054] In some embodiments, the S EMT It is obtained based on the following method: the gene expression information of the tumor cells to be tested is compared with the EMT-related gene set, and the gene expression similarity is calculated to obtain the S EMT .

[0055] In some specific embodiments, the EMT-related gene set is derived from the dbEMT 2.0 database.

[0056] In some specific embodiments, the gene expression similarity between tumor cells and the EMT-related gene set is calculated by the ssGSEA algorithm.

[0057] In some embodiments, the S Travel The method is as follows: based on the gene expression information of the tumor cells to be tested and the gene expression information of the cells in the blood, the interaction strength between the tumor cells to be tested and the different immune cells in the blood is predicted, the weights are set based on the existence ratio of each immune cell in the blood, and the S of the tumor cells to be tested is obtained by weighted summation. Travel .

[0058] In some preferred embodiments, the immune cells in the blood include one or more of neutrophils, platelets and natural killer cells. Travel It is positively correlated with the interaction intensity between tumor cells and neutrophils. Travel It is positively correlated with the interaction strength between tumor cells and platelets. Travel Negatively correlated with the intensity of interaction between tumor cells and natural killer cells.

[0059] Among them, neutrophils and platelets have the ability to protect the survival of cancer cells, while natural killer cells kill cancer cells.

[0060] In some embodiments, the S Travel Based on the following method:

[0061] Based on the existing tumor metastasis dataset, tumor non-metastasis dataset and normal blood immune cell dataset, we generated a dataset that combined the gene expression information of tumor cells with S Travel an associated first prediction model;

[0062] Using the first prediction model, key data is obtained from the gene expression information of the tumor cells to be tested and weights are assigned to calculate the S Travel .

[0063] In some embodiments, the S comm Based on the following method:

[0064] Based on the gene expression information of the tumor cells to be tested and the gene expression information of the main cells of the tissue microenvironment, the interaction between the tumor cells and the main cells of the tissue microenvironment is predicted, and the weights are set based on the proportions of the main cells in the tissue microenvironment. The weighted sum of the interaction strengths between the tumor cells and the main cells of the tissue microenvironment is subtracted from the weighted sum of the interaction strengths between the main cells in the tissue microenvironment, thereby obtaining the S of the tumor cells to be tested.comm .

[0065] During specific implementation, those skilled in the art can identify the main cell types of each tissue microenvironment based on common sense.

[0066] In some specific embodiments, the major cell types within the tissue microenvironment are confirmed by integrating human single-cell atlas datasets from different studies using the R software package Harmony, and re-annotating the cell types of each tissue or organ separately to obtain the major cell types of each tissue microenvironment.

[0067] As examples, the tissues or organs include bones, brain, lungs, liver and other common metastatic sites.

[0068] In some embodiments, the S comm Based on the following method:

[0069] Based on the existing tumor metastasis dataset, tumor non-metastasis dataset and normal tissue cell dataset, a new dataset was generated that combines the gene expression information of tumor cells with S comm an associated second prediction model;

[0070] Using the second prediction model, key data is obtained from the gene expression information of the tumor cells to be tested and weights are assigned to calculate the S comm .

[0071] In a specific implementation, the weighted sum of the interactions mentioned in the present invention can be determined by combining the corresponding scalar weights of the actual scales of different values, and further combined into a single scalar value through a linear or nonlinear, algebraic, trigonometric or related method via an algebraic algorithm, a statistical learning algorithm, a Bayesian algorithm, a regression algorithm or a similar algorithm, and together with the decision function derived from the mathematical derivation of the scalar value, the first prediction model or the second prediction model is provided, and the gene expression information from the tumor cells to be tested can be parsed into the corresponding S through the prediction model. Travel Value or S comm Those skilled in the art can also develop the first prediction model or the second prediction model with a set of representative expression profiles of an existing tumor metastasis data set (or representative expression profiles of other relevant features) by learning weights and decision thresholds under cross-validation, bootstrap or similar sampling techniques to optimize sensitivity, specificity, negative and positive predictive values, hazard ratios or any combination thereof.

[0072] The invention refers to a weighted sum of interactions, wherein the individual weights can be positive or negative.

[0073] Due to sequencing technology, some existing data sets contain many different types of cells for each sample, and removing the gene expression information of other cells (including epithelial cells, immune cells, fibroblasts, etc.) will help improve the accuracy of the prediction model. Therefore, in some embodiments, when obtaining the gene expression information of tumor cells from the existing tumor metastasis data set and the tumor non-metastasis data set, only one sample is selected for each individual, and only the gene expression information of primary tumor cells is retained, while the gene expression information of other cells is removed.

[0074] In some preferred embodiments, the gene expression information of other cells is screened out by using marker genes and / or copy number variation.

[0075] In some cases, it is not possible to distinguish tumor cells from other cells by marker genes. In this case, copy number variation can be used to distinguish them.

[0076] In some specific embodiments, the cibersortx algorithm is used to remove the gene expression information of other cells.

[0077] In some specific embodiments, the InferCNV algorithm is used to infer the copy number variation of cells, and cells with larger copy number variation are inferred to be tumor cells.

[0078] The “existing tumor metastasis dataset” mentioned in the present invention includes The Cancer Genome Atlas (TCGA) data.

[0079] When detecting a specific type of cancer, those skilled in the art may also introduce existing "tumor metastasis datasets" and "tumor non-metastasis datasets" of the corresponding type. For example, in a method for the metastasis risk of liver cancer cells, the existing "liver cancer metastasis dataset" and "liver cancer non-metastasis dataset" may be combined to generate or improve a corresponding prediction model.

[0080] The "normal blood immune cell dataset" mentioned in the present invention refers to a dataset containing normal blood immune cell expression information. The "normal tissue cell dataset" mentioned in the present invention refers to a dataset containing normal tissue cell expression information. In some embodiments, those skilled in the art can simultaneously obtain the above information based on the known human normal cell dataset.

[0081] In some embodiments, the method further comprises:

[0082] Provide the S Meta thresholds associated with transfer risk;

[0083] The S of the tumor cells to be tested Meta Compared with the threshold, when the S MetaWhen the value is higher than the threshold, it is predicted that the tumor cells to be tested have a risk of metastasis.

[0084] In some specific embodiments, the threshold value can be confirmed based on comparative analysis of existing tumor metastasis datasets and tumor non-metastasis datasets.

[0085] In some embodiments, multiple cutoff values ​​can be determined based on comparative analysis of existing tumor metastasis data sets and tumor non-metastasis data sets, and the S of the tumor cells to be tested can be set to Meta By comparing with the cutoff value, a conclusion is obtained that the tumor cells to be tested have different degrees of metastasis risk (such as low metastasis risk, medium metastasis risk or high metastasis risk, etc.).

[0086] In some embodiments, the gene expression information is transcriptome information.

[0087] Those skilled in the art can identify specific methods for obtaining gene expression information based on common knowledge, such as microarray, Q-PCR, immunohistochemistry, ELISA, or other techniques capable of quantifying mRNA or protein expression.

[0088] In some specific embodiments, the gene expression information in the present invention is obtained by RNA sequencing or single cell sequencing.

[0089] The "ribonucleic acid sequencing" mentioned in the present invention is referred to as RNA sequencing, also known as Whole Transcriptome Shotgun Sequencing. It is a transcriptomics research method based on the second-generation sequencing technology. RNA sequencing is a technology that uses the power of second-generation sequencing to reveal a snapshot of the presence and quantity of RNA from a genome at a given moment.

[0090] The "single-cell sequencing" mentioned in the present invention refers to the use of optimized next-generation sequencing technology (NGS) to detect the sequence of single cells, which can obtain cell sequence differences under specific microenvironments to facilitate the study of their functional differences, etc. DNA sequencing of individual cells can help understand the variation of small-scale cells in cancer, for example; RNA sequencing can help understand and identify different cell types and the genes they express, which is of great benefit to the study of developmental biology, etc.

[0091] In some embodiments, the tumor cells to be tested are derived from a biological sample of a subject to be tested, wherein the subject to be tested is suspected of having a tumor or has been diagnosed with a tumor.

[0092] The individuals “suspected of having a tumor” mentioned in the present invention include individuals who are in the observation period after the tumor is cured.

[0093] In some embodiments, nucleic acid may include RNA or DNA nucleic acid, for example, mRNA, cRNA, cDNA, etc., provided that the sample retains the expression information of the host cell or tissue from which it is obtained. Samples can be prepared by a variety of different methods known in the art, for example, by isolating mRNA from cells, as known in the field of differential gene expression, wherein the isolated mRNA is used as isolated, amplified, or used to prepare cDNA, cRNA, etc. Therefore, determining the level of the mRNA in the sample includes preparing cDNA or cRNA from mRNA, and then measuring cDNA or cRNA. Samples are usually prepared using standard protocols from cells or tissues harvested from an object in need of treatment, wherein the cell type or tissue from which the nucleic acid can be produced includes any tissue in which the expression type of the phenotype to be determined exists, including but not limited to, sick cells or tissues, body fluids, etc.

[0094] "Biological sample", "sample" and "test sample" are used interchangeably herein to refer to any substance, biological fluid, tissue or cell obtained or otherwise taken from an individual. It includes blood (including whole blood, white blood cells, peripheral blood mononuclear cells, buffy coat, plasma and serum), sputum, tears, mucus, nasal washes, nasal aspirates, breath samples, urine, semen, saliva, meningeal fluid, amniotic fluid, glandular fluid, lymph fluid, nipple aspirates, bronchial aspirates, synovial fluid, joint aspirates, ascites, cells, cell extracts and cerebrospinal fluid. It also includes all of the above experimentally separated fractions. For example, a blood sample can be fractionated into serum or fractions containing specific types of blood cells, such as red blood cells or white blood cells (leukocytes). If desired, the sample can be a combination of samples from an individual, such as a combination of tissue and fluid samples. The term "biological sample" also includes substances containing homogeneous solid matter, such as substances from stool samples, tissue samples or biopsy tissues. The term "biological sample" also includes substances derived from tissue culture or cell culture. Any suitable method for obtaining a biological sample may be utilized; exemplary methods include, for example, phlebotomy, swabbing (e.g., oral swabbing), and fine needle aspiration biopsy procedures. Samples may also be collected, for example, by microdissection (e.g., laser capture microdissection (LCM) or laser microdissection (LMD)), bladder washing, smears (e.g., PAP smears), or catheter lavage. A "biological sample" obtained or derived from an individual includes any such sample that has been processed in any suitable manner after being obtained from the individual.

[0095] In some embodiments, the tumor comprises one or more of breast cancer, colorectal cancer, lung cancer, liver cancer, pancreatic cancer, and skin cancer.

[0096] Method for predicting organ metastasis tendency of tumor cells

[0097] The present invention also provides a method for predicting the organ metastasis tendency of tumor cells, which comprises: using the method for predicting the risk of tumor cell metastasis as described above;

[0098] Among them, the S of tumor cells invading different tissues or organs to be tested is obtained. comm When the tumor cells to be tested metastasize to a certain tissue or organ, comm When it is higher, it is predicted that the tumor cells have a higher tendency to metastasize to that tissue or organ.

[0099] In some embodiments, the method further comprises:

[0100] Provide the S comm Thresholds associated with propensity to metastasize;

[0101] The S of the tumor cells to be tested comm Compared with the threshold, when the S comm When the value is higher than the threshold, it is predicted that the tumor cells to be detected have a tendency to metastasize to the tissue or organ.

[0102] In some specific embodiments, the threshold value can be confirmed based on comparative analysis of existing tumor metastasis datasets and tumor non-metastasis datasets.

[0103] In some embodiments, multiple cutoff values ​​can be determined based on comparative analysis of existing tumor metastasis data sets and tumor non-metastasis data sets, and the S of the tumor cells to be tested can be set to comm By comparing with the cutoff value, a conclusion is obtained that the tumor cells to be tested have different degrees of metastasis to the tissue or organ (such as high metastasis tendency, medium metastasis tendency or low metastasis tendency, etc.).

[0104] Methods for predicting the prognosis of tumor patients

[0105] The present invention also provides a method for predicting the prognosis of a tumor patient, comprising: using the method for predicting the risk of tumor cell metastasis as described above;

[0106] Among them, the S of the tumor cells to be tested is obtained Meta , when S Meta When it is higher, the tumor patient is predicted to have a worse prognosis.

[0107] In some embodiments, the method further comprises:

[0108] Provide the S Meta Thresholds associated with prognostic conditions;

[0109] The S of the tumor cells to be tested Meta Compared with the threshold, when the S MetaWhen the value is higher than the threshold, it is predicted that the tumor cells to be tested have a poor prognosis.

[0110] In some specific embodiments, the threshold value can be confirmed based on comparative analysis of existing tumor metastasis datasets and tumor non-metastasis datasets.

[0111] In some embodiments, multiple cutoff values ​​can be determined based on comparative analysis of existing tumor metastasis data sets and tumor non-metastasis data sets, and the S of the tumor cells to be tested can be set to Meta By comparing with the cutoff value, a conclusion is obtained that the tumor cells to be tested have different degrees of prognosis (such as better prognosis, poor prognosis or worse prognosis, etc.).

[0112] Other applications

[0113] The method for predicting the risk of tumor cell metastasis of the present invention can be applied to auxiliary diagnosis of tumor patients. In clinical practice, the method of the present invention can be used to determine the metastasis risk of tumor cells, and combined with the age, medical history, clinical symptoms and related detection indicators of the tumor patient, to comprehensively diagnose the patient's tumor metastasis risk, organ tendency and prognosis.

[0114] The method for predicting the risk of tumor cell metastasis of the present invention can be applied to the research and development of new anticancer drugs. By measuring the risk of tumor cell metastasis before and after drug addition (or setting up a drug addition experimental group and a non-drug addition control group), the effect of the drug in reducing the risk of tumor cell metastasis can be measured.

[0115] Tumor cell detection system

[0116] The present invention further provides a tumor cell detection system, which comprises:

[0117] An acquisition module configured to acquire gene expression information of tumor cells to be tested;

[0118] A calculation module is configured to predict the S of the tumor cell to be tested based on the gene expression information. EMT , S Travel and S comm ; Among them, S EMT Used to indicate the epithelial-mesenchymal transition (EMT) ability of tumor cells. Travel Indicates the survival ability of tumor cells in blood circulation, S comm Indicates the ability of tumor cells to invade tissues or organs; then the risk factor S of the tumor cells to be tested for metastasis is calculated using the following formula Meta :

[0119] And, the output module is configured to beMeta When it is higher, it is predicted to have a higher risk of metastasis.

[0120] In some embodiments, the computing module includes S EMT The computing unit, the S EMT The calculation unit is configured to compare the gene expression information of the tumor cells to be tested with the EMT-related gene set, calculate the gene expression similarity, and thus obtain the S EMT .

[0121] In some embodiments, the computing module includes S Travel The computing unit, the S Travel The calculation unit is configured to predict the interaction strength between the tumor cells to be tested and different immune cells in the blood based on the gene expression information of the tumor cells to be tested and the gene expression information of the cells in the blood, set the weights based on the existence ratio of each immune cell in the blood, and obtain the S of the tumor cells to be tested by weighted summation. Travel .

[0122] In some preferred embodiments, the immune cells in the blood include one or more of neutrophils, platelets and natural killer cells. Travel It is positively correlated with the interaction intensity between tumor cells and neutrophils. Travel It is positively correlated with the interaction strength between tumor cells and platelets. Travel Negatively correlated with the intensity of interaction between tumor cells and natural killer cells.

[0123] In some embodiments, the computing module includes S comm The computing unit, the S comm The calculation unit is configured to predict the interaction between the tumor cells and the main cells of the tissue microenvironment based on the gene expression information of the tumor cells to be tested and the gene expression information of the main cells of the tissue microenvironment, set the weights based on the proportions of the main cells in the tissue microenvironment, and subtract the weighted sum of the interaction strengths between the main cells in the tissue microenvironment from the weighted sum of the interaction strengths between the tumor cells and the main cells of the tissue microenvironment, so as to obtain the S of the tumor cells to be tested. comm .

[0124] In some embodiments, the computing module includes a model generation unit; and S Travel Computational units and / or S comm Computational unit;

[0125] The model generation unit is configured to generate a first prediction model based on an existing tumor metastasis data set, a tumor non-metastasis data set, and a normal blood immune cell data set, and / or to generate a second prediction model based on an existing tumor metastasis data set, a tumor non-metastasis data set, and a normal tissue cell data set;

[0126] The first prediction model combines the gene expression information of tumor cells with S Travel associated;

[0127] The second prediction model combines the gene expression information of tumor cells with S comm associated;

[0128] The S Travel The calculation unit is configured to use the first prediction model to obtain key data from the gene expression information of the tumor cells to be tested and assign weights to calculate the S of the tumor cells to be tested. Travel ;

[0129] The S comm The calculation unit is configured to use the second prediction model to obtain key data from the gene expression information of the tumor cells to be tested and assign weights to calculate the S of the tumor cells to be tested. comm .

[0130] In some embodiments, when obtaining gene expression information of tumor cells from existing tumor metastasis datasets and tumor non-metastasis datasets, only one sample is selected for each individual, and only the gene expression information of primary tumor cells is retained, while the gene expression information of other cells is removed; more preferably, the gene expression information of other cells is screened out by marker genes and / or copy number variations.

[0131] In some embodiments, the output module includes a first output unit, the first output unit is configured to provide the S Meta The threshold value associated with the risk of metastasis; the S Meta Compared with the threshold, when the test score is higher than the threshold, it is predicted that the tumor cell to be tested has a risk of metastasis.

[0132] In some embodiments, the output module further includes a second output unit, wherein the second output unit is configured to obtain the S of the tumor cells to be detected invading different tissues or organs. comm When the tumor cells to be tested metastasize to a certain tissue or organ, comm When it is higher, it is predicted that the tumor cells have a higher tendency to metastasize to that tissue or organ.

[0133] In some embodiments, the output module further includes a third output unit, wherein the third output unit is configured to obtain the S of the tumor cell to be detected. Meta, when S Meta When it is higher, the tumor patient is predicted to have a worse prognosis.

[0134] In some embodiments, the tumor cell detection system further comprises:

[0135] The detection end is used to detect the gene expression information of tumor cells;

[0136] The detection end is connected to the acquisition module via communication.

[0137] In specific implementation, those skilled in the art can set corresponding detection ends in combination with conventional gene expression information detection methods such as microarray, Q-PCR, immunohistochemistry, ELISA or other technologies capable of quantifying mRNA or protein expression.

[0138] In some preferred embodiments, the gene expression information is transcriptome information.

[0139] In some preferred embodiments, the detection end includes a RNA sequencing device or a single cell sequencing device.

[0140] In some specific embodiments, the tumor cells to be tested are derived from a biological sample of a subject to be tested, wherein the subject to be tested is suspected of having a tumor or has been diagnosed with a tumor.

[0141] In some specific embodiments, the tumor comprises one or more of breast cancer, colorectal cancer, lung cancer, liver cancer, pancreatic cancer, and skin cancer.

[0142] In some optional embodiments, the tumor cell detection system further comprises:

[0143] The device control module is used to control the operation of the detection end, the acquisition module, the calculation module and the output module, and the device control module is connected with the detection end, the acquisition module, the calculation module and the output module through communication.

[0144] In specific implementation, those skilled in the art can also configure a corresponding tumor cell detection system based on the specific implementation schemes introduced in the methods and applications section above.

[0145] Non-transitory computer readable storage medium

[0146] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the instructions are run on a computing device, the computer program is executed by a processor to implement: the method for predicting the risk of tumor cell metastasis as described above, or the method for predicting the organ metastasis tendency of tumor cells as described above, or the method for predicting the prognosis of tumor patients as described above.

[0147] Electronic devices

[0148] The present invention further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for predicting the risk of tumor cell metastasis as described above, or the method for predicting the organ metastasis tendency of tumor cells as described above, or the method for predicting the prognosis of tumor patients as described above is implemented.

[0149] Computer program product

[0150] The present invention further provides a computer program product, which includes a computer program. When the computer program is read and executed by a computing device, the computing device executes: the method for predicting the risk of tumor cell metastasis as described above, or the method for predicting the organ metastasis tendency of tumor cells as described above, or the method for predicting the prognosis of tumor patients as described above.

[0151] Embodiments of the present invention will be described in detail below in conjunction with examples. It should be understood that these examples are only used to illustrate the present invention and are not intended to limit the scope of the present invention. The experimental methods in the following examples that do not specify specific conditions are preferably referred to the guidance provided in the present invention, and can also be based on the experimental manual or normal conditions in this area, and can also be based on other experimental methods known in the art, or according to the conditions recommended by the manufacturer.

[0152] In the following specific embodiments, the measured parameters of raw material components may have slight deviations within the range of weighing accuracy unless otherwise specified. For temperature and time parameters, acceptable deviations caused by instrument test accuracy or operation accuracy are allowed.

[0153] Example 1

[0154] 1. Research Methods

[0155] The research method of the present invention is mainly divided into the following steps (such as Figure 1 First, the US Cancer Genome Atlas (TCGA) data and human single-cell atlas data were collected as gene expression data for cancer cells and normal tissue cells, respectively. Then, different scores were proposed based on several important steps in the occurrence of metastasis to measure the possibility of metastasis. Finally, the model was applied to the single-cell datasets of colorectal cancer liver metastasis and breast cancer, respectively, to compare the differences between patients with and without metastasis.

[0156] 1.1TCGA data collection and deconvolution

[0157] Transcriptome sequencing data and clinical information of patients were downloaded from the official website of the U.S. Cancer Genome Atlas (https: / / portal.gdc.cancer.gov / ). According to the known types of cancer that often metastasize, 7 important cancer types were screened for prediction, including breast cancer (BRCA), colorectal cancer (COAD), lung cancer (LUAD, LUSC), liver cancer (LIHC), pancreatic cancer (PAAD), and skin cancer (SKCM). For each cancer, only the primary cancer was retained for the prediction of metastasis, and only one sample was selected for each patient. Due to sequencing technology, each sample contains many different types of cells. In order to obtain a purer expression of tumor cells, the expression of epithelial cells, immune cells, and fibroblasts in the sample was removed by using the cibersortx algorithm, and the remaining tumor cell expression was used for subsequent research. In addition, the clinical data of these patients were also collected, including the size or invasion range of the tumor itself, whether lymph node invasion and metastasis occurred, distant metastasis, etc.

[0158] 1.2 Human single-cell atlas collection and preprocessing

[0159] In order to understand the transcriptome expression of the internal microenvironment of different tissues and organs under normal conditions, the expression of tissues and organs that are common metastatic sites was collected from several important human single-cell atlas studies, including bones, brain, lungs, liver, etc. The single-cell atlas datasets from different studies were integrated using the R software package Harmony. Then, the cell types of each tissue and organ were re-annotated and the main cell types of each tissue microenvironment were obtained.

[0160] 1.3 Transfer probability score calculation

[0161] The epithelial-mesenchymal transition (EMT) process plays a vital role in tumor metastasis. From the dbEMT 2.0 database, we found important gene sets related to EMT. We calculated the gene expression similarity of tumor cells using the single gene enrichment analysis algorithm ssGSEA, denoted as S EMT .

[0162] Another factor that determines tumor metastasis is that cancer cells can travel to different tissues and organs through blood circulation. Here, we calculate the interaction between cancer cells and immune cells in the blood to determine whether cancer cells can survive in the blood circulation, denoted as S TravelSpecifically, neutrophils and platelets in the blood help protect cancer cells from surviving in the blood, while natural killer cells kill cancer cells. Therefore, the strength of the interaction between cancer cells and these cells is considered separately. Based on the gene expression information of the tumor cells to be tested and the gene expression information of cells in the blood, the strength of the interaction between the tumor cells to be tested and the different immune cells in the blood is predicted. The weights are set based on the proportion of each immune cell in the blood, and then their weighted sum is used to obtain S. Travel .

[0163] Finally, the present invention believes that when the interaction between cancer cells and cells inside the tissue microenvironment is stronger than the original effect of cells in the tissue, cancer cells can invade into the new microenvironment. Based on the gene expression information of the tumor cells to be tested and the gene expression information of the main cells of the tissue microenvironment, the interaction between cancer cells and the main cells of the tissue microenvironment and the interaction inside the microenvironment are calculated respectively. The weights are set based on the proportions of the main cells inside the tissue microenvironment, and the weighted sum of the interaction strengths between the main cells inside the tissue microenvironment is subtracted from the weighted sum of the interaction strengths between the main cells inside the tissue microenvironment. The intensity of the interaction between tumor cells and tissue cells is compared to see whether it is stronger than the interaction strength between cells inside the tissue, and the ability of cancer cells to invade tissues and organs is calculated, recorded as S. comm .

[0164] Since cancer metastasis is determined by these three effects, S Meta Represents the risk factor of metastasis for a patient.

[0165]

[0166] Because it is calculated for each tissue and organ separately, this score can also measure the tendency of organ metastasis. Using all samples of a certain cancer in the TCGA database to simulate all patients with that cancer, the expectation of metastasis to different locations can be calculated to predict the overall metastasis ratio of the population.

[0167] 2. Research Results

[0168] 2.1 Prediction of overall metastatic organ preference in the population

[0169] The model was applied to predict the metastatic organ tendency of six common cancers (breast cancer, lung cancer, liver cancer, colorectal cancer, pancreatic cancer, skin cancer) to seven frequently metastatic tissues and organs (brain, bone marrow, liver, lung, bladder, kidney, lymph node). Furthermore, in order to illustrate the accuracy of the prediction, a recent research data was selected to obtain the distribution of various cancer metastases to different organs in clinical statistics. This clinical statistical data was used as the gold standard for evaluating the algorithm. Figure 2A uses a Sankey diagram to show the predicted metastasis ratio. The left side represents a primary cancer and the right side represents a metastatic site. From this diagram, we can see that lung cancer is the most metastatic cancer, followed by colorectal cancer and pancreatic cancer. On the other hand, the lungs and liver are the most likely sites for cancer metastasis, and cancers related to the digestive system are more likely to metastasize to the liver. Figure 2 The comparison between the prediction results and clinical statistical results is shown in B. Figure 2 In B, each point represents the probability of a cancer metastasizing to an organ, where the color of the point represents a primary cancer, and the shape of the point represents metastasis to a tissue organ. The X-axis coordinate represents the results of clinical statistics, while the Y-axis coordinate represents the predicted results. Intuitively, most of the results are located near the diagonal; the calculated average error is 13.62%, and the paired Wilcoxon test p value is 0.854, which shows that the predicted results are consistent with the statistics in the actual population.

[0170] 2.2 Analysis of metastasis in patients with lung adenocarcinoma

[0171] Let’s take lung adenocarcinoma patients as an example. For each lung adenocarcinoma patient in the TCGA database, we can calculate the S Meta The risk scores of metastasis to different tissues and organs were calculated. Figure 3 A shows the metastasis of the first four patients. They all tend to metastasize to the bone marrow, liver and lymph nodes, while the possibility of brain and kidney metastasis is lower. At the same time, the metastasis tendency of patients is also quite different. The first patient tends to metastasize to the lymph nodes while the second patient is more likely to metastasize to the liver. According to the metastasis risk coefficient score, the overall patients are divided into three categories: high metastasis tendency (greater than 0.4), medium metastasis tendency (less than 0.4 and greater than 0.2) and low metastasis tendency (less than 0.2). The proportion of patients in the three categories with each type of metastasis is as follows: Figure 3 As shown in B. It can be seen that most people may have at least one metastasis, and in comparison, there are more people with low risk of bone metastasis and kidney metastasis. Combined with clinical data, survival analysis was performed on people in different groups ( Figure 3 C). It can be clearly seen that patients with a high risk of metastasis have a poorer prognosis, while patients with a low risk of metastasis have a longer survival period. This shows that the classification can reasonably distinguish the metastasis risk of patients and screen out patients with a low risk of metastasis.

[0172] 2.3 Data analysis of colorectal cancer liver metastasis

[0173] In order to verify that the model of the present invention can also be applied to single-cell data sets, data from two recent studies on colorectal cancer were selected. Among them, no metastasis was found in all patients in the study (Zhang, Lei, et al. Cell 181.2 (2020): 442-459), while all patients in (Liu, Yedan, et al. Cancer Cell 40.4 (2022): 424-437) had liver metastasis. The present invention attempts to compare the differences in primary cancer data of the two articles. To this end, the CD45- cell population in (Liu, Yedan, et al. Cancer Cell 40.4 (2022): 424-437) was first selected for clustering, and these cells were divided into epithelial / cancer cells, endothelial cells, fibroblasts, and other 4 major categories ( Figure 4 A). Since cancer cells are largely mixed with epithelial cells and have no specific marker genes, the present invention uses the InferCNV algorithm to infer their copy number variation and compares them with endothelial cells ( Figure 4 B). The cells with larger copy number variation are considered cancer cells. Through the screening in the first two steps, pure cancer cells were found from CD45- cells. The cancer cells in the dataset (Zhang, Lei, et al. Cell 181.2 (2020): 442-459) were found by the same method. The interaction analysis of the two cancer cell datasets with the normal liver microenvironment was performed respectively. The results are as follows Figure 4 C, Figure 4 D. By comparison, it can be found that liver metastasis will interact more with macrophages, monocytes, neutrophils, etc. in the liver. This shows that the interaction between cancer cells and them may help cancer cells invade the liver and promote tumor metastasis. Further, through Cellchat calculation, it is possible to find out the ligand-receptor pairs that play an important role in the interaction between cancer cells and endothelial cells, macrophages and neutrophils ( Figure 4 E). It can be seen that ANXA family, COL family ligands and ITGAM receptors play a key role in the interaction between cancer cells and liver microenvironment, which is consistent with previous reports. These important ligand-receptors are of great significance for further revealing the mechanism of tumor metastasis.

[0174] 2.4 Analysis of lung adenocarcinoma metastasis data

[0175] In order to study the applicability of the method of the present invention in different types of cancer, this example further collected a data set of lung adenocarcinoma research in recent years (Laughney, Ashley M., et al. Nature medicine 26.2 (2020): 259-269). Similarly, in this example, the main cell types of the data are first found using a single-cell clustering algorithm ( Figure 5 A). The Copykat algorithm was used to find the copy number variation and infer the proportion of cancer cells ( Figure 5 B). Compared with the 16 patient samples in the original data, 10 were from primary lung cancer, 4 were from brain metastasis, and 2 were from bone metastasis, all of which were from different samples ( Figure 5 C) By studying the interactions between cells, the differences between the interactions between cancer samples from the primary site and metastatic samples were compared ( Figure 5 D). It can be seen that compared with primary cancer samples, cancer cells in metastatic samples have more significant interactions with the microenvironment, especially with neutrophils and macrophages. This interaction prompts cancer cells to metastasize to the brain or bones more quickly. Among them, the ligand receptor pairs CCL2-CXCR1 and ITGA2-IGBRE2 are particularly prominent in the interaction between cancer cells and macrophages ( Figure 5 E). In order to verify the rationality of the method of the present invention, the data of the primary samples and the samples with known metastasis were predicted respectively, and their metastasis probabilities to the bone and brain were estimated. The prediction results showed that the predicted probability of bone metastasis in the bone metastasis samples was 57.4%, which was much higher than the 32.1% metastasis probability predicted for the samples without bone metastasis. Similarly, the predicted probability of brain metastasis in the brain metastasis samples was 62.3%, which was higher than the 36.7% ( Figure 5 F) These results show that the prediction method of the present invention can well obtain the possibility of metastasis to different locations in patients, and provide help for the prediction of cancer metastasis.

[0176] The above-mentioned embodiments only express several implementation methods of the present invention, and the description is relatively specific and detailed, but it cannot be understood as limiting the scope of the invention patent. It should be pointed out that for ordinary technicians in this field, several modifications and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A tumor cell detection system, comprising: An acquisition module configured to acquire gene expression information of tumor cells to be tested; A calculation module is configured to predict the S of the tumor cell to be tested based on the gene expression information. EMT , S Travel and S comm ; Among them, S EMT Used to indicate the epithelial-mesenchymal transition ability of tumor cells, S Travel Indicates the survival ability of tumor cells in blood circulation, S comm Indicates the ability of tumor cells to invade tissues or organs; then the risk factor S of the tumor cells to be tested for metastasis is calculated using the following formula Meta : The calculation module includes S EMT Computing unit, S Travel Computational unit and S comm Computing unit; wherein the S EMT The calculation unit is configured to compare the gene expression information of the tumor cells to be tested with the gene set related to the epithelial cell-mesenchymal cell transition ability, calculate the gene expression similarity, and thus obtain the S EMT ; The S Travel The calculation unit is configured to predict the interaction strength between the tumor cells to be tested and different immune cells in the blood based on the gene expression information of the tumor cells to be tested and the gene expression information of the cells in the blood, set the weights based on the existence ratio of each immune cell in the blood, and obtain the S of the tumor cells to be tested by weighted summation. Travel Among them, immune cells in the blood include neutrophils, platelets and natural killer cells; S Travel It is positively correlated with the interaction intensity between tumor cells and neutrophils. Travel It is positively correlated with the interaction strength between tumor cells and platelets. Travel It is negatively correlated with the interaction intensity between tumor cells and natural killer cells; the S comm The calculation unit is configured to predict the interaction between the tumor cells and the main cells of the tissue microenvironment based on the gene expression information of the tumor cells to be tested and the gene expression information of the main cells of the tissue microenvironment, set the weights based on the proportions of the main cells in the tissue microenvironment, and subtract the weighted sum of the interaction strengths between the main cells in the tissue microenvironment from the weighted sum of the interaction strengths between the tumor cells and the main cells of the tissue microenvironment, so as to obtain the S of the tumor cells to be tested. comm ; And, the output module is configured to be Meta When it is higher, it is predicted to have a higher risk of metastasis.

2. The tumor cell detection system according to claim 1, wherein: The calculation module includes a model generation unit; And, S Travel Computational unit and S comm Computational unit; The model generation unit is configured to generate a first prediction model based on an existing tumor metastasis data set, a tumor non-metastasis data set, and a normal blood immune cell data set, and to generate a second prediction model based on an existing tumor metastasis data set, a tumor non-metastasis data set, and a normal tissue cell data set; The first prediction model combines the gene expression information of tumor cells with S Travel associated; The second prediction model combines the gene expression information of tumor cells with S comm associated; The S Travel The calculation unit is configured to use the first prediction model to obtain key data from the gene expression information of the tumor cells to be tested and assign weights to calculate the S of the tumor cells to be tested. Travel ; The S comm The calculation unit is configured to use the second prediction model to obtain key data from the gene expression information of the tumor cells to be tested and assign weights to calculate the S of the tumor cells to be tested. comm .

3. The tumor cell detection system according to claim 2, wherein: When obtaining the gene expression information of tumor cells from the existing tumor metastasis dataset and tumor non-metastasis dataset, only one sample is selected for each individual, and only the gene expression information of primary tumor cells is retained, while the gene expression information of other cells is removed.

4. The tumor cell detection system according to claim 3, wherein: Gene expression information of other cells can be filtered out by marking genes and / or copy number variations.

5. The tumor cell detection system according to any one of claims 1 to 4, wherein: The output module includes a first output unit, the first output unit is configured to provide the S Meta The threshold value associated with the risk of metastasis; the S Meta Compared with the threshold, when the S Meta When the value is higher than the threshold, it is predicted that the tumor cells to be tested have a risk of metastasis.

6. The tumor cell detection system according to any one of claims 1 to 4, wherein: The output module includes a second output unit, which is configured to obtain the S of the tumor cells to be detected invading different tissues or organs. comm When the tumor cells to be tested metastasize to a certain tissue or organ, comm When it is higher, it is predicted that the tumor cells have a higher tendency to metastasize to that tissue or organ.

7. The tumor cell detection system according to claim 5, wherein: The output module also includes a second output unit, which is configured to obtain the S of the tumor cells to be detected invading different tissues or organs. comm When the tumor cells to be tested metastasize to a certain tissue or organ, comm When it is higher, it is predicted that the tumor cells have a higher tendency to metastasize to that tissue or organ.

8. The tumor cell detection system according to any one of claims 1 to 4 and 7, wherein: The output module further includes a third output unit, and the third output unit is configured to obtain the S of the tumor cell to be detected. Meta , when S Meta When it is higher, the tumor patient is predicted to have a worse prognosis.

9. The tumor cell detection system according to claim 5, wherein: The output module further includes a third output unit, and the third output unit is configured to obtain the S of the tumor cell to be detected. Meta , when S Meta When it is higher, the tumor patient is predicted to have a worse prognosis.

10. The tumor cell detection system according to claim 6, wherein: The output module further includes a third output unit, and the third output unit is configured to obtain the S of the tumor cell to be detected. Meta , when S Meta When it is higher, the tumor patient is predicted to have a worse prognosis.

11. The tumor cell detection system according to any one of claims 1 to 4, 7, 9 to 10, further comprising: The detection end is used to detect the gene expression information of tumor cells; The detection end is connected to the acquisition module via communication.

12. The tumor cell detection system according to claim 5, further comprising: The detection end is used to detect the gene expression information of tumor cells; The detection end is connected to the acquisition module via communication.

13. The tumor cell detection system according to claim 6, further comprising: The detection end is used to detect the gene expression information of tumor cells; The detection end is connected to the acquisition module via communication.

14. The tumor cell detection system according to claim 8, further comprising: The detection end is used to detect the gene expression information of tumor cells; The detection end is connected to the acquisition module via communication.

15. The tumor cell detection system according to claim 11, wherein: The detection end includes a RNA sequencing device or a single cell sequencing device.

16. The tumor cell detection system according to any one of claims 12 to 14, wherein: The detection end includes a RNA sequencing device or a single cell sequencing device.

17. A non-transitory computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the method of predicting the risk of tumor cell metastasis, or the method of predicting the organ metastasis tendency of tumor cells, or the method of predicting the prognosis of tumor patients; in, The method for predicting the risk of tumor cell metastasis comprises: Obtaining gene expression information of tumor cells to be tested; Predicting the S of the tumor cell to be tested based on the gene expression information EMT , S Travel and S comm ; Among them, S EMT It is used to indicate the epithelial-mesenchymal transition ability of tumor cells. EMT It is obtained based on the following method: the gene expression information of the tumor cells to be tested is compared with the gene set related to epithelial cell-mesenchymal cell transition, and the gene expression similarity is calculated to obtain the S of the tumor cells to be tested. EMT ; S Travel Indicates the viability of tumor cells in blood circulation. The S Travel The method is as follows: based on the gene expression information of the tumor cells to be tested and the gene expression information of the cells in the blood, the interaction strength between the tumor cells to be tested and the different immune cells in the blood is predicted, and the weights are set based on the existence ratio of each immune cell in the blood, and the S of the tumor cells to be tested is obtained by weighted summation. Travel Among them, immune cells related to cancer metastasis in the blood include neutrophils, platelets and natural killer cells; S Travel It is positively correlated with the interaction intensity between tumor cells and neutrophils. Travel It is positively correlated with the interaction strength between tumor cells and platelets. Travel Negatively correlated with the interaction strength between tumor cells and natural killer cells; S comm Indicates the ability of tumor cells to invade tissues or organs. comm The S of the tumor cells to be tested is obtained based on the gene expression information of the tumor cells to be tested and the gene expression information of the main cells of the tissue microenvironment, the interaction between the tumor cells and the main cells of the tissue microenvironment is predicted, the weights are set based on the proportions of the main cells in the tissue microenvironment, and the weighted sum of the interaction strengths between the main cells in the tissue microenvironment is subtracted from the weighted sum of the interaction strengths between the main cells in the tissue microenvironment, thereby obtaining the S of the tumor cells to be tested. comm ; Then the risk factor S of the tumor cells to be tested for metastasis is calculated using the following formula: Meta : When the S Meta When it is higher, it is predicted to have a higher risk of metastasis; The method for predicting the organ metastasis tendency of tumor cells comprises: Using the method for predicting the risk of tumor cell metastasis; Among them, the S of tumor cells invading different tissues or organs to be tested is obtained. comm When the tumor cells to be tested metastasize to a certain tissue or organ, comm When it is higher, it is predicted that the tumor cells have a higher tendency to metastasize to that tissue or organ; The method for predicting the prognosis of a tumor patient comprises: Using the method for predicting the risk of tumor cell metastasis; Among them, the S of the tumor cells to be tested is obtained Meta , when S Meta When it is higher, the tumor patient is predicted to have a worse prognosis.

18. The non-transitory computer readable storage medium of claim 17, wherein: The S of the tumor cells to be tested Travel Based on the following method: Based on the existing tumor metastasis dataset, tumor non-metastasis dataset and normal blood immune cell dataset, we generated a dataset that combined the gene expression information of tumor cells with S Travel an associated first prediction model; Using the first prediction model, key data is obtained from the gene expression information of the tumor cells to be tested and weights are assigned to calculate the S Travel .

19. The non-transitory computer-readable storage medium of claim 17, wherein: The S of the tumor cells to be tested comm Based on the following method: Based on the existing tumor metastasis dataset, tumor non-metastasis dataset and normal tissue cell dataset, a new dataset was generated that combines the gene expression information of tumor cells with S comm an associated second prediction model; Using the second prediction model, key data is obtained from the gene expression information of the tumor cells to be tested and weights are assigned to calculate the S comm .

20. The non-transitory computer readable storage medium according to claim 18 or 19, wherein: When obtaining the gene expression information of tumor cells from the existing tumor metastasis dataset and tumor non-metastasis dataset, only one sample is selected for each individual, and only the gene expression information of primary tumor cells is retained, while the gene expression information of other cells is removed.

21. The non-transitory computer readable storage medium of claim 20, wherein: Gene expression information of other cells can be filtered out by marking genes and / or copy number variations.

22. The non-transitory computer-readable storage medium according to any one of claims 17 to 19 and 21, wherein: The method for predicting the risk of tumor cell metastasis also includes: Provide the S Meta thresholds associated with transfer risk; The S of the tumor cells to be tested Meta Compared with the threshold, when the S Meta When the value is higher than the threshold, it is predicted that the tumor cells to be tested have a risk of metastasis.

23. The non-transitory computer readable storage medium of claim 20, wherein: The method for predicting the risk of tumor cell metastasis also includes: Provide the S Meta thresholds associated with transfer risk; The S of the tumor cells to be tested Meta Compared with the threshold, when the S Meta When the value is higher than the threshold, it is predicted that the tumor cells to be tested have a risk of metastasis.

24. The non-transitory computer readable storage medium of claim 17, wherein: The gene expression information is transcriptome information.

25. The non-transitory computer readable storage medium of claim 17, wherein: The tumor cells to be tested are derived from a biological sample of a subject to be tested, and the subject to be tested is suspected of having a tumor or has been diagnosed with a tumor.

26. The non-transitory computer readable storage medium of claim 25, wherein: The tumor includes one or more of breast cancer, colorectal cancer, lung cancer, liver cancer, pancreatic cancer, and skin cancer.

27. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method of predicting the risk of tumor cell metastasis, or the method of predicting the organ metastasis tendency of tumor cells, or the method of predicting the prognosis of tumor patients is implemented; in, The method for predicting the risk of tumor cell metastasis comprises: Obtaining gene expression information of tumor cells to be tested; Predicting the S of the tumor cell to be tested based on the gene expression information EMT , S Travel and S comm ; Among them, S EMT It is used to indicate the epithelial-mesenchymal transition ability of tumor cells. EMT It is obtained based on the following method: the gene expression information of the tumor cells to be tested is compared with the gene set related to epithelial cell-mesenchymal cell transition, and the gene expression similarity is calculated to obtain the S of the tumor cells to be tested. EMT ; S Travel Indicates the viability of tumor cells in blood circulation. The S Travel The method is as follows: based on the gene expression information of the tumor cells to be tested and the gene expression information of the cells in the blood, the interaction strength between the tumor cells to be tested and the different immune cells in the blood is predicted, and the weights are set based on the existence ratio of each immune cell in the blood, and the S of the tumor cells to be tested is obtained by weighted summation. Travel Among them, immune cells related to cancer metastasis in the blood include neutrophils, platelets and natural killer cells; S Travel It is positively correlated with the interaction intensity between tumor cells and neutrophils. Travel It is positively correlated with the interaction strength between tumor cells and platelets. Travel Negatively correlated with the interaction strength between tumor cells and natural killer cells; S comm Indicates the ability of tumor cells to invade tissues or organs. comm The S of the tumor cells to be tested is obtained based on the gene expression information of the tumor cells to be tested and the gene expression information of the main cells of the tissue microenvironment, the interaction between the tumor cells and the main cells of the tissue microenvironment is predicted, the weights are set based on the proportions of the main cells in the tissue microenvironment, and the weighted sum of the interaction strengths between the main cells in the tissue microenvironment is subtracted from the weighted sum of the interaction strengths between the main cells in the tissue microenvironment, thereby obtaining the S of the tumor cells to be tested. comm ; Then the risk factor S of the tumor cells to be tested for metastasis is calculated using the following formula: Meta : When the S Meta When it is higher, it is predicted to have a higher risk of metastasis; The method for predicting the organ metastasis tendency of tumor cells comprises: Using the method for predicting the risk of tumor cell metastasis; Among them, the S of tumor cells invading different tissues or organs to be tested is obtained. comm When the tumor cells to be tested metastasize to a certain tissue or organ, comm When it is higher, it is predicted that the tumor cells have a higher tendency to metastasize to that tissue or organ; The method for predicting the prognosis of a tumor patient comprises: Using the method for predicting the risk of tumor cell metastasis; Among them, the S of the tumor cells to be tested is obtained Meta , when S Meta When it is higher, the tumor patient is predicted to have a worse prognosis.

28. The electronic device according to claim 27, wherein: The S of the tumor cells to be tested Travel Based on the following method: Based on the existing tumor metastasis dataset, tumor non-metastasis dataset and normal blood immune cell dataset, we generated a dataset that combined the gene expression information of tumor cells with S Travel an associated first prediction model; Using the first prediction model, key data is obtained from the gene expression information of the tumor cells to be tested and weights are assigned to calculate the S Travel .

29. The electronic device according to claim 27, wherein: The S of the tumor cells to be tested comm Based on the following method: Based on the existing tumor metastasis dataset, tumor non-metastasis dataset and normal tissue cell dataset, a new dataset was generated that combines the gene expression information of tumor cells with S comm an associated second prediction model; Using the second prediction model, key data is obtained from the gene expression information of the tumor cells to be tested and weights are assigned to calculate the S comm .

30. The electronic device according to claim 28 or 29, wherein: When obtaining the gene expression information of tumor cells from the existing tumor metastasis dataset and tumor non-metastasis dataset, only one sample is selected for each individual, and only the gene expression information of primary tumor cells is retained, while the gene expression information of other cells is removed.

31. The electronic device according to claim 30, wherein: Gene expression information of other cells can be filtered out by marking genes and / or copy number variations.

32. The electronic device according to any one of claims 27 to 29 and 31, wherein: The method for predicting the risk of tumor cell metastasis also includes: Provide the S Meta thresholds associated with transfer risk; The S of the tumor cells to be tested Meta Compared with the threshold, when the S Meta When the value is higher than the threshold, it is predicted that the tumor cells to be tested have a risk of metastasis.

33. The electronic device according to claim 30, wherein: The method for predicting the risk of tumor cell metastasis also includes: Provide the S Meta thresholds associated with transfer risk; The S of the tumor cells to be tested Meta Compared with the threshold, when the S Meta When the value is higher than the threshold, it is predicted that the tumor cells to be tested have a risk of metastasis.

34. The electronic device according to claim 27, wherein: The gene expression information is transcriptome information.

35. The electronic device according to claim 27, wherein: The tumor cells to be tested are derived from a biological sample of a subject to be tested, and the subject to be tested is suspected of having a tumor or has been diagnosed with a tumor.

36. The electronic device according to claim 35, wherein: The tumor includes one or more of breast cancer, colorectal cancer, lung cancer, liver cancer, pancreatic cancer, and skin cancer.

37. A computer program product, comprising a computer program, which, when read and executed by a computing device, causes the computing device to execute: a method for predicting the risk of tumor cell metastasis, or a method for predicting the organ metastasis tendency of tumor cells, or a method for predicting the prognosis of a tumor patient; in, The method for predicting the risk of tumor cell metastasis comprises: Obtaining gene expression information of tumor cells to be tested; Predicting the S of the tumor cell to be tested based on the gene expression information EMT , S Travel and S comm ; Among them, S EMT It is used to indicate the epithelial-mesenchymal transition ability of tumor cells. EMT It is obtained based on the following method: the gene expression information of the tumor cells to be tested is compared with the gene set related to epithelial cell-mesenchymal cell transition, and the gene expression similarity is calculated to obtain the S of the tumor cells to be tested. EMT ; S Travel Indicates the viability of tumor cells in blood circulation. The S Travel The method is as follows: based on the gene expression information of the tumor cells to be tested and the gene expression information of the cells in the blood, the interaction strength between the tumor cells to be tested and the different immune cells in the blood is predicted, and the weights are set based on the existence ratio of each immune cell in the blood, and the S of the tumor cells to be tested is obtained by weighted summation. Travel Among them, immune cells related to cancer metastasis in the blood include neutrophils, platelets and natural killer cells; S Travel It is positively correlated with the interaction intensity between tumor cells and neutrophils. Travel It is positively correlated with the interaction strength between tumor cells and platelets. Travel Negatively correlated with the interaction strength between tumor cells and natural killer cells; S comm Indicates the ability of tumor cells to invade tissues or organs. comm The S of the tumor cells to be tested is obtained based on the gene expression information of the tumor cells to be tested and the gene expression information of the main cells of the tissue microenvironment, the interaction between the tumor cells and the main cells of the tissue microenvironment is predicted, the weights are set based on the proportions of the main cells in the tissue microenvironment, and the weighted sum of the interaction strengths between the main cells in the tissue microenvironment is subtracted from the weighted sum of the interaction strengths between the main cells in the tissue microenvironment, thereby obtaining the S of the tumor cells to be tested. comm ; Then the risk factor S of the tumor cells to be tested for metastasis is calculated using the following formula: Meta : When the S Meta When it is higher, it is predicted to have a higher risk of metastasis; The method for predicting the organ metastasis tendency of tumor cells comprises: Using the method for predicting the risk of tumor cell metastasis; Among them, the S of tumor cells invading different tissues or organs to be tested is obtained. comm When the tumor cells to be tested metastasize to a certain tissue or organ, comm When it is higher, it is predicted that the tumor cells have a higher tendency to metastasize to that tissue or organ; The method for predicting the prognosis of a tumor patient comprises: Using the method for predicting the risk of tumor cell metastasis; Among them, the S of the tumor cells to be tested is obtained Meta , when S Meta When it is higher, the tumor patient is predicted to have a worse prognosis.

38. The computer program product of claim 37, wherein: The S of the tumor cells to be tested Travel Based on the following method: Based on the existing tumor metastasis dataset, tumor non-metastasis dataset and normal blood immune cell dataset, we generated a dataset that combined the gene expression information of tumor cells with S Travel an associated first prediction model; Using the first prediction model, key data is obtained from the gene expression information of the tumor cells to be tested and weights are assigned to calculate the S Travel .

39. The computer program product of claim 37, wherein: The S of the tumor cells to be tested comm Based on the following method: Based on the existing tumor metastasis dataset, tumor non-metastasis dataset and normal tissue cell dataset, a new dataset was generated that combines the gene expression information of tumor cells with S comm an associated second prediction model; Using the second prediction model, key data is obtained from the gene expression information of the tumor cells to be tested and weights are assigned to calculate the S comm .

40. A computer program product according to claim 38 or 39, wherein: When obtaining the gene expression information of tumor cells from the existing tumor metastasis dataset and tumor non-metastasis dataset, only one sample is selected for each individual, and only the gene expression information of primary tumor cells is retained, while the gene expression information of other cells is removed.

41. The computer program product of claim 40, wherein: Gene expression information of other cells can be filtered out by marking genes and / or copy number variations.

42. The computer program product according to any one of claims 37 to 39 and 41, wherein: The method for predicting the risk of tumor cell metastasis also includes: Provide the S Meta thresholds associated with transfer risk; The S of the tumor cells to be tested Meta Compared with the threshold, when the S Meta When the value is higher than the threshold, it is predicted that the tumor cells to be tested have a risk of metastasis.

43. The computer program product of claim 40, wherein: The method for predicting the risk of tumor cell metastasis also includes: Provide the S Meta thresholds associated with transfer risk; The S of the tumor cells to be tested Meta Compared with the threshold, when the S Meta When the value is higher than the threshold, it is predicted that the tumor cells to be tested have a risk of metastasis.

44. The computer program product of claim 37, wherein: The gene expression information is transcriptome information.

45. The computer program product of claim 37, wherein: The tumor cells to be tested are derived from a biological sample of a subject to be tested, and the subject to be tested is suspected of having a tumor or has been diagnosed with a tumor.

46. ​​The computer program product of claim 45, wherein: The tumor includes one or more of breast cancer, colorectal cancer, lung cancer, liver cancer, pancreatic cancer, and skin cancer.

Images