Early detection system for simultaneously detecting various digestive system tumors

By detecting the methylation status of SEPT9, SDC2, SFRP2 and RNF180 genes, and using multiple algorithms integrated learning models, the problem of difficulty in detecting multiple digestive system tumors at the same time in the existing technology is solved, and a rapid and accurate diagnosis of multiple tumors is achieved. , reducing diagnostic costs and cycles.

CN118957069APending Publication Date: 2024-11-15SINGAPORE YIKUN DIAGNOSTICS PTE LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411121586.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-15
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

The prior art is difficult to detect multiple digestive system tumors at the same time, resulting in a long diagnosis cycle, high cost, and the inability to effectively trace the origin of multiple cancers.

Method used

Provide a premature detection system, including a kit system and a software analysis system, to detect the methylation status of SEPT9, SDC2, SFRP2 and RNF180 genes, and use the integrated learning model of multiple algorithms for joint diagnosis, so as to achieve the tumors of multiple digestive system. risk prediction and diagnosis.

Benefits of technology

The purpose of diagnosing multiple digestive system tumors at the same time through one test is achieved, reducing the diagnosis cycle and cost, and improving the accuracy and efficiency of diagnosis.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118957069A_ABST
    Figure CN118957069A_ABST
Patent Text Reader

Abstract

The invention provides an early detection system for simultaneously detecting multiple digestive system tumors, and relates to the field of biomedicine.The system comprises the steps of blood sample collection, cfDNA extraction and corresponding hydrosulphite conversion, and then methylation states of four genes SEPT9, SDC2, SFRP2 and RNF180 are detected and analyzed through a digestive system tumor early diagnosis reagent; and analyzing the optimal model obtained on the basis of machine learning training in combination with clinical data so as to evaluate the risks of various cancers, namely colon cancer, rectal cancer, liver cancer and gastric cancer, thereby achieving the purpose of simultaneously diagnosing various digestive system tumors through one-time detection.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the field of biomedicine, and in particular relates to an early detection system for simultaneously detecting multiple digestive system tumors. Background Art

[0002] Among many types of cancer, digestive system cancer is one of the most common malignant tumors. Currently, the screening and diagnosis of digestive system cancer in clinical practice are mainly through imaging methods such as endoscopy or ultrasound, as well as pathological testing. However, endoscopic examinations for colorectal cancer and gastric cancer are expensive, and patient compliance is poor. Although ultrasonic examinations for liver cancer are simple, convenient and inexpensive, they have a high rate of missed diagnosis due to limited imaging clarity and the inability to perform tissue biopsies immediately. In addition, traditional tumor marker screening methods, including PSA testing, PG testing, and AFP testing, have many limitations, such as low accuracy, insufficient specificity, and high false negative and false positive rates.

[0003] In recent years, liquid biopsy technology has provided new ideas for early detection of tumors. Liquid biopsy refers to the technology of detecting and diagnosing cancer through body fluids such as blood, saliva, and urine. Research and clinical practice have shown that the method of detecting tumor-derived cells or DNA in non-tumor tissue samples such as blood, urine, and feces has the advantages of non-invasiveness, high efficiency, and high sensitivity. Therefore, liquid biopsy technology has been widely used in early detection of cancer. In the screening method of digestive system cancer, the main technical route is to perform NGS or qPCR detection based on the detection of methylated cfDNA in blood or feces. This type of detection method usually selects multiple sites that are highly methylated in tumor tissues but low or no methylation in normal tissues. Then, specific operations are performed on the detected methylation levels, and the results of the operations are used for diagnosis. Since most of the cfDNA in the blood comes from hematopoietic cells, and the content of cfDNA released from tumor cells is low, the signal-to-noise ratio of the detection results of cfDNA methylation is low. How to distinguish the differences in methylation signals between tumor patients and non-tumor patients is the key to accurately diagnosing cancer patients. In recent years, new patents and products have emerged to detect and analyze cfDNA specific gene methylation for the purpose of accurately diagnosing cancer, such as: In the patent application with publication number CN113611404A, the applicant provided a mathematical model for liver cancer diagnosis. The method performs binary logistic regression analysis on the results of gender, age, and liver cancer triple test (including alpha-fetoprotein allotypes, alpha-fetoprotein, and abnormal prothrombin), and on this basis, establishes a multi-factor logistic regression model. After the diagnostic result is brought into the model, a score is obtained, and the score is compared with the preset threshold to achieve the diagnosis of liver cancer. The diagnostic sensitivity of this method for primary hepatocellular carcinoma is 86.9%, and the specificity is 80.2%.

[0004] In the patent application with publication number CN105200157B, the applicant provides a method for diagnosing gastric cancer, including a kit for quantitative detection of 5 microRNAs in plasma. The method quantitatively analyzes the CT value obtained by PCR, substitutes it into a mathematical model also based on logistic regression, and compares the score obtained by the model with a preset threshold, thereby achieving the diagnosis of colorectal cancer. The method has achieved a sensitivity of 84.0% and a specificity of 84.0% in clinical practice.

[0005] The patent with publication number CN106399570A provides an auxiliary diagnosis system for early colorectal cancer, including a PCR detection kit for the methylation level of BMP3 and NDRG4 genes in stool samples, a PCR detection kit for KRAS gene mutations in stool samples, and a fecal occult blood detection kit. The diagnosis result is achieved by substituting multiple test results into a mathematical model based on logistic regression, and comparing the score obtained by the model with the preset threshold, thereby achieving the diagnosis of colorectal cancer. This method has achieved a sensitivity of 76.4% and a specificity of 88.10% in clinical practice.

[0006] From the above cases, it can be seen that the current diagnosis of single cancer types based on the detection of clinical samples by detection kits and the mathematical model analysis of the test results has almost reached a sensitivity of >80% and a specificity of >80%. However, these patents and corresponding products have certain limitations, that is, these products are mostly used to diagnose a single tumor. However, in many cases, those at risk of cancer need to be diagnosed with multiple potential cancers at the same time. There are certain challenges in how to use these existing technologies or products to screen for multiple digestive system cancers in high-risk populations at the same time and trace the detected cancers. Because if these existing products and technologies are used, different kits need to be used to test the tester multiple times, and more samples need to be collected from patients. This will inevitably lead to a long diagnosis cycle, high costs, and inconvenience to patients. Therefore, this will greatly limit the application of cancer screening tests. Summary of the invention

[0007] Terms and Claims of the Present Invention: DNA:Deoxyribonucleic acid:deoxyribonucleic acid.

[0008] cfDNA: free deoxyribonucleic acid.

[0009] NGS: Next Generation Sequencing: second generation sequencing.

[0010] qPCR: Quantitative Polymerase Chain Reaction: quantitative polymerase chain reaction.

[0011] CT:Cycle Threshold: cycle threshold.

[0012] NC:Negative Control: negative control.

[0013] PC: Positive Control: positive control.

[0014] In order to solve the above problems, the present invention provides an early detection system for simultaneously detecting multiple digestive system tumors.

[0015] On the one hand, the present invention provides an early detection system for simultaneously detecting the methylation status of SEPT9, SDC2, SFRP2 and RNF180 genes to predict the risk of digestive system tumors.

[0016] Specifically, the early detection system can be divided into a reagent kit system and a software analysis system.

[0017] More specifically, the kit system includes but is not limited to: a sample collection tube, a cfDNA extraction kit, a DNA bisulfite conversion kit and a digestive system methylation detection kit. The cfDNA extraction kit, the DNA bisulfite conversion kit and the digestive system methylation detection kit can be commercially available or laboratory-made.

[0018] More specifically, the digestive system methylation detection kit includes primers, probes or blocking primers for detecting the methylation status of four genes: SEPT9, SDC2, SFRP2 and RNF180.

[0019] In certain specific embodiments of the present invention, the sequences of the primers, probes and blocking primers are as shown in DEQ ID NO: 1-29.

[0020] More specifically, the digestive system methylation detection kit also includes primers or probes for detecting control genes.

[0021] More specifically, the control genes include: Actin, ACTB and / or GAPDH.

[0022] In certain embodiments of the present invention, the control gene is Actin.

[0023] In certain specific embodiments of the present invention, the sequences of primers and probes for detecting Actin are shown in SEQ ID NOs: 30-32.

[0024] Specifically, the software analysis system includes: integrating different types of clinical data, based on multi-algorithm ensemble learning, and a model capable of jointly diagnosing multiple cancer types.

[0025] More specifically, the model includes one or a combination of a logistic regression model, a random forest model, a SVM, a Bayesian model, a decision tree model, and a neural network model.

[0026] In some specific embodiments of the present invention, the model is an integrated learning model of algorithms such as logistic regression, random forest, and support vector machine.

[0027] On the other hand, the present invention provides the use of the early detection system in preparing risk prediction or diagnosis kits for various digestive system tumors.

[0028] Specifically, the above-mentioned various digestive system tumors include: colon cancer, rectal cancer, liver cancer and gastric cancer.

[0029] More specifically, the steps for using the early detection system to predict the risk of various digestive system tumors are as follows: (1) Collect samples; (2) extraction of cfDNA and corresponding bisulfite conversion; (3) The methylation status of four genes, SEPT9, SDC2, SFRP2 and RNF180, was detected and analyzed in combination with clinical data using the optimal model obtained through machine learning training.

[0030] Specifically, in step (1), the sample includes: serum, plasma, tissue, cultured cells, blood, feces, saliva, urine, and cfDNA.

[0031] Specifically, the methylation level detection method includes: BSP first-generation sequencing, NGS second-generation sequencing, qPCR, pyrophosphate sequencing, bisulfite conversion sequencing, methylation chip method, MSP methylation-specific PCR, digital PCR method, third-generation sequencing method, whole genome methylation sequencing method, DNA enrichment detection method, simplified bisulfite sequencing technology, HPLC method, methylation-specific PCR or a combination thereof.

[0032] In certain specific embodiments of the present invention, in step (3), the detection step is: 1) Set up qPCR reaction according to the number of samples; 2) Collect qPCR test CT values; In certain specific embodiments of the present invention, in step (3), the analysis step is: A. Bring the processed quantitative data into the multi-classification logic operation formula for calculation to obtain the probability value P of a given sample belonging to each type of tumor or health; B. Find the classification result with the largest P value among all K classifications; C. Predict cancer risk using P value.

[0033] Specifically, the logic operation formula is:

[0034] in, Processed quantitative data; For this category The weight vector of is the weight vector for each category in all K categories, It is optimized based on clinical data through machine learning methods.

[0035] Compared with the prior art, the present invention has the following beneficial effects: The main purpose of the present invention is to provide an advanced cancer risk prediction system that can diagnose multiple digestive system tumors simultaneously through a single test. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 The figure is a working flow chart of the digestive system tumor early detection system of the present invention.

[0037] Figure 2 This is the reagent kit analysis process for the digestive system tumor early detection system.

[0038] Figure 3 Analyze the process for multi-classification models based on machine learning.

[0039] Figure 4 Fluorescence quantitative PCR amplification curves of four targets and internal reference for digestive system detection reagents.

[0040] Figure 5 Results of clinical testing of an early detection system for digestive system tumors.

[0041] Figure 6 To provide an early detection system for digestive system tumors and predict staging results in clinical testing. DETAILED DESCRIPTION

[0042] The present invention will be further described in detail below in conjunction with specific examples. The following examples are not intended to limit the present invention, but are only intended to illustrate the present invention. The experimental methods used in the following examples are generally conventional, unless otherwise specified, and the materials, reagents, etc. used in the following examples are commercially available, unless otherwise specified.

[0043] CT conversion reagent components: The CT converter of the present invention is a mixed solution of 1.5-4.0M sodium pyrosulfite and 0.01-0.06M hydroquinone. The two drugs in the CT converter are mixed in proportion and stored in the form of dry powder, which is prepared as needed when used.

[0044] DNA protectant ingredients: The DNA protective agent of the present invention is a 30%-70% dimethylformamide solution.

[0045] CT reagent conditioning solution components: The CT reagent regulating solution of the present invention comprises a 0.5-3.0M sodium hydroxide solution.

[0046] Magnetic bead suspension components: The magnetic bead suspension of the present invention is a hydroxyl microsphere with a particle size of 150nm-1.5μm, a solid content of 2.5%, and a surface coated with silicon dioxide / ferrositol.

[0047] Binding solution components: The binding solution component of the present invention is 3.0-7.5M guanidine hydrochloride solution.

[0048] Rinse liquid composition: The rinse solution component of the present invention is 40%-80% ethanol solution.

[0049] Desulfurization liquid components: The desulfurization liquid component of the present invention is a mixed solution of 0.3-0.7M sodium hydroxide and 20%-60% ethanol.

[0050] Eluent composition: The eluent component of the present invention is a mixture of 0.01M trihydroxymethane and 0.001M ethylenediaminetetraacetic acid, and the pH value is adjusted to 7.8-8.2 by hydrochloric acid.

[0051] Example 1 Detailed working process of the early detection system for digestive system tumors 1. Collection and processing of blood samples: 1.1. Collect 5 mL of whole blood into a blood collection tube treated with EDTA anticoagulant.

[0052] 1.2. Centrifuge the whole blood sample at 1,000 g for 15 minutes and collect the upper plasma.

[0053] 1.3. Transfer 2 mL of plasma to a 15 mL centrifuge tube.

[0054] 2. Extraction and purification of cfDNA from plasma: 2.1. Add 2 μL of carrier RNA, 100 μL of proteinase K and 100 μL of SDS into a 15 mL centrifuge tube containing 2 mL of plasma and mix well.

[0055] 2.2. Incubate at 55°C for 30 minutes, mixing by inverting several times.

[0056] 2.3. Incubate at room temperature for 5 minutes to cool the sample.

[0057] 2.4. Add 3.5 mL of binding solution and 100 μL of magnetic beads to the sample. Mix by inversion or vortexing for 10 minutes at room temperature. Place the tube on the magnetic stand for 5 minutes until the beads form a compact pellet. Discard the supernatant after this time.

[0058] 2.5. Add 1000 µL of Wash Buffer 1 and vortex for 15 seconds to resuspend the beads. Transfer the resuspended solution to a 1.5 mL microcentrifuge tube and place the tube on a magnetic stand for 1 minute until the beads form a compact pellet. Discard the supernatant.

[0059] 2.6. Add 1000 µL of Wash Buffer 1 and vortex for 15 seconds to resuspend the beads. Place the tube on a magnetic stand for 1 minute until the beads form a compact pellet, after which time discard the supernatant.

[0060] 2.7. Add 1000 µL of Wash Buffer 2 and vortex for 15 seconds to resuspend the beads. Place the tube on a magnetic stand for 1 minute until the beads form a compact pellet, after which time discard the supernatant and repeat this wash step.

[0061] 2.8. Briefly centrifuge to collect the liquid in the tube, place the collection tube on the magnetic rack and discard the liquid.

[0062] 2.9. Dry the pellet at 37°C for 15 minutes.

[0063] 2.10. Add 25 µL of elution buffer and vortex for 5 minutes at 37°C to resuspend the beads. 2.11. Place the collection tube on the magnetic stand for 1 minute.

[0064] 2.12. Transfer the supernatant containing purified cfDNA to a clean 1.5 mL microcentrifuge tube.

[0065] 2.13. The extracted cfDNA is used for subsequent downstream bisulfite conversion process.

[0066] Bisulfite conversion of cfDNA Prepare CT conversion solution as follows: 3.1.1. Add deoxygenated purified water to CT dry powder and make up to 10 mL.

[0067] 3.1.2. Add 2 mL of CT reagent conditioning solution and 500 µL of DNA protective agent in sequence.

[0068] 3.1.3. Vortex for 5 minutes until completely dissolved.

[0069] 3.2. Prepare the bisulfite reaction system in a 200 μL PCR tube. The specific preparation system is shown in Table 1: Table 1

[0070] 3.3. After the reaction system is prepared, set the conversion program on the PCR instrument for sulfite conversion. The program is shown in Table 2: Table 2 stage temperature time Number of cycles transsexual 98℃ 10min 1 Incubation 64℃ 30min 2 Keep 4℃ 10min 1 3.4. Purified cfDNA was treated with sulfite by the following procedure: 3.4.1. Take a new centrifuge tube, add 600μL binding solution and 20μL magnetic beads, transfer the conversion solution obtained from the reaction into the binding solution, and shake to mix.

[0071] 3.4.2. Leave at room temperature for 5 minutes and centrifuge briefly.

[0072] 3.4.3. Place the centrifuge tube on the magnetic rack for 5 minutes, remove the waste liquid, and keep the magnetic beads for later use.

[0073] 3.4.4. Add 400 μL of washing solution, shake and mix, centrifuge briefly, place on a magnetic rack for 3 minutes, and remove the waste liquid.

[0074] 3.4.5. Add 200 μL of desulfurization solution, shake and mix, then place at room temperature for 15 minutes, centrifuge briefly, place on the magnetic rack for 3 minutes, and remove the waste liquid.

[0075] 3.4.6. Add 400 μL of washing solution and shake to mix, then centrifuge briefly and place on the magnetic rack for 3 minutes before removing the waste liquid.

[0076] 3.4.7. Add 400 μL of washing solution and shake to mix, then centrifuge briefly. After 3 minutes on the magnetic rack, remove the waste liquid and centrifuge briefly to remove as much waste liquid as possible.

[0077] 3.4.8. Open the lid and dry at room temperature for 10 minutes, add 25μL of elution solution to the centrifuge tube, wash the magnetic beads on the tube wall to the bottom of the tube, and close the lid tightly.

[0078] 3.4.9. Vortex for 5 seconds and leave at room temperature for 4 minutes.

[0079] 3.4.10. Collect the converted cfDNA by magnetic suction for 1 min.

[0080] 4. qPCR detection using the Digestive System Cancer Morning Kit: The nucleic acid sequences of the primers, probes and blockers used are shown in Table 3: Table 3

[0081] 4.1. Set up qPCR reactions according to the number of samples. Prepare the qPCR reaction mixture according to the following composition for each qPCR reaction: Enzyme premix (Meridian Bioscience, MDX082), 5 μL; Methylation reaction mixture, 5 μL.

[0082] 4.2. Aliquot 10 μL of qPCR reaction mixture into qPCR reaction tubes. Prepare two extra tubes for each test as negative and positive controls for methylation reactions.

[0083] 4.3. Add 10 μL of bisulfite-converted cfDNA to the above mixture and mix thoroughly. Take 10 μL of NC and PC in the kit and add them to the negative control and positive control reaction tubes.

[0084] 4.4. Place the reaction tube in the qPCR and set the following PCR program for qPCR reaction.

[0085] Table 4

[0086] 5. Data machine model analysis and calculation test results: 5.1. Collect the CT value and tumor marker value of qPCR.

[0087] 5.2. Data processing of CT values ​​and tumor markers of given samples.

[0088] 5.3. The processed quantitative data is brought into the multi-classification logic operation formula for calculation to obtain the probability value P of a given sample belonging to each type of tumor or health. The logic operation formula is as follows:

[0089] in, Processed quantitative data; For this category The weight vector of is the weight vector for each category in all K categories, It is optimized based on clinical data through machine learning methods.

[0090] 5.4. Find the classification result from all K classifications with the largest P value.

[0091] 5.5. Predict cancer outcomes using P values.

[0092] The workflow of the early detection system for digestive system tumors is shown in Figure 1 , the sample collection and processing procedures are shown in Figure 2 , machine learning and auxiliary diagnosis process see Figure 3 .

[0093] Example 2 ROC curve of early detection system The various reagents in the early detection system for digestive system tumors fully demonstrate the detection of gene methylation in combined use. Through the collection of blood samples, extraction of cfDNA, and conversion of bisulfite. Figure 4 As shown in the figure, in the process of reagent detection, the digestive system tumor early detection kit can detect various methylated genes to be detected, including SEPT9, SDC2, SFRP2 and RNF180. At the same time, the PCR of the digestive system tumor early detection reagent contains an internal reference gene. The internal reference gene is used as a standard to evaluate whether the PCR is successful, and also as a reference for evaluating the sample size. Figure 4 It can be seen that a representative fluorescence quantitative PCR amplification curve of the digestive system tumor early detection reagent shows that the qPCR can detect four targets and internal references.

[0094] In order to verify the performance of the system in detecting various digestive system tumors, a clinical experiment involving 602 subjects was conducted. The subjects included 330 patients diagnosed with one of the following cancers: colon cancer, rectal cancer, liver cancer, or gastric cancer, 165 healthy subjects, and 107 patients diagnosed with one of the following precancerous lesions: colorectal cancer, liver cancer, or gastric cancer. The clinical diagnosis of cancer was made by clinical diagnostic methods such as gastroscopy, proctoscopy, and / or pathology.

[0095] The sample inclusion, exclusion and removal criteria are as follows: This study intends to include samples from patients with stage I-VI primary malignant tumors, including colorectal cancer, gastric cancer or hepatocellular carcinoma, other solid tumors (interference group) and healthy people (healthy group). The specific group definitions are as follows: Colorectal cancer group: colorectal cancer of different stages diagnosed by colonoscopy or pathology.

[0096] Colorectal precancerous lesions group: colorectal polypoid adenoma, chronic ulcerative colitis, progressive adenoma, non-progressive adenoma, adenomatous polyps, etc.

[0097] Gastric cancer group: gastric cancer of different stages diagnosed by gastroscopy or pathology.

[0098] Gastric precancerous lesions group: chronic atrophic gastritis, gastric ulcer, gastric polyps, residual stomach after surgery, hypertrophic gastritis, high-grade gastric intraepithelial neoplasia, low-grade gastric intraepithelial neoplasia, etc.

[0099] Hepatocellular carcinoma group: patients with liver cancer of different stages who meet the diagnostic criteria of the 2022 edition of the "Guidelines for the Diagnosis and Treatment of Primary Liver Cancer".

[0100] Liver cancer precancerous lesion group: high-grade dysplastic nodules of the liver, low-grade dysplastic nodules of the liver, cirrhosis, adenomatous hyperplasia of cirrhosis, etc.

[0101] Interference group: patients with other solid tumors except colorectal cancer, gastric cancer or hepatocellular carcinoma.

[0102] Healthy group: Those who have excluded diseases such as digestive system diseases and tumor diseases and can provide blood samples.

[0103] Training set sample inclusion criteria: a. Healthy people aged 40 or above, patients with precancerous lesions, and patients with tumors at different stages can be included in the training set; b. For early stage cancer (stage I-II), the patient must be newly diagnosed and treated, and the patient specimens must be provided within one month before or after the diagnosis of the disease; c.BCT blood collection volume is about 10mL; d. Those who are willing and sign the informed consent form; e. The relevant information of the sample is complete, including: sample type, age, gender, sample collection time, clinical diagnosis, pathological information, and the following clinical information:

[0104] Training set sample exclusion criteria: a. Samples with plasma volume less than 4 mL after separation; b. The sample is severely hemolyzed.

[0105] Training set sample removal criteria: a. Samples that cannot be tested or fail to be tested due to improper collection or storage; b. Samples with insufficient sample volume to complete the test; c. Samples with missing clinical data or irregular records.

[0106] Validation set sample inclusion, exclusion and removal criteria: Validation set sample inclusion criteria: a. Healthy people aged 40 years or older, patients with precancerous lesions, and patients with tumors at different stages can be included in the validation set; b. Having one or more risk factors for digestive tract tumors: such as positive stool occult blood, positive Helicobacter pylori, chronic hepatitis B carrier, etc.; c. Laparoscopic examination and tumor marker examination will be performed in the near future; d. The BCT blood collection volume is approximately 10 mL.

[0107] Validation set sample exclusion criteria: a. Samples with plasma volume less than 4 mL after separation; b. The sample is severely hemolyzed.

[0108] Validation set sample removal criteria: a. Samples that cannot be tested or fail to be tested due to improper collection or storage; b. Samples with insufficient sample volume to complete the test; c. Samples with missing clinical data or irregular records.

[0109] The system of the present invention is used to obtain and process sample data, verify and assess the effectiveness of the digestive system tumor early detection system, and thus evaluate the clinical performance of the system.

[0110] Table 5

[0111] When the results were analyzed by binary classification, as shown in Table 5, the system was able to fully distinguish between health / cancer; health / digestive tract tumors; health / liver cancer; health / colorectal cancer; health / gastric cancer; and liver cancer / digestive tract tumors. Among them, the system was best at distinguishing between health / digestive tract tumors, with the area under the ROC curve reaching 0.92.

[0112] Sensitivity refers to the ratio of the number of samples correctly predicted as positive to the number of samples that are actually positive; specificity refers to the ratio of the number of samples correctly predicted as negative to the number of samples that are actually negative; accuracy refers to the percentage of the number of samples predicted correctly to the total number of samples.

[0113] According to Table 6, Figure 5 , Figure 6 The experimental results show that the system can accurately identify and diagnose a variety of digestive system tumors, including those in the early stages. Its overall sensitivity is 83.5%, specificity is 87.8%, and the area under the ROC curve is 84.7%; the overall accuracy rate reaches 85.8%.

[0114] Table 6

[0115] Finally, it should be noted that the above content is only used to illustrate the technical solution of the present invention, rather than to limit the scope of protection of the present invention. Simple modifications or equivalent substitutions of the technical solution of the present invention by ordinary technicians in this field do not deviate from the essence and scope of the technical solution of the present invention.

Claims

1. An early detection system for simultaneously detecting multiple digestive system tumors, characterized in that: The early detection system diagnoses digestive system tumors based on the combined detection of SEPT9, SDC2, SFRP2 and RNF180 methylation status.

2. The early detection system according to claim 1, characterized in that: The early detection system can be divided into a reagent kit system and a software analysis system.

3. The early detection system according to claim 2, characterized in that: The kit system includes: a sample collection tube, a cfDNA extraction kit, a DNA bisulfite conversion kit and a digestive system methylation detection kit. The cfDNA extraction kit, the DNA bisulfite conversion kit and the digestive system methylation detection kit are commercially available or laboratory-made.

4. The early detection system according to claim 3, characterized in that: The digestive system methylation detection kit comprises primers, probes or blocking primers for detecting the methylation status of four genes: SEPT9, SDC2, SFRP2 and RNF180.

5. The early detection system according to claim 4, characterized in that: The digestive system methylation detection kit comprises primers, probes and blocking primers whose sequences are shown in SEQ ID NOs: 1-29.

6. The early detection system according to claim 3, characterized in that: The digestive system methylation detection kit also includes primers, probes or blocking primers for detecting control genes.

7. The early detection system according to claim 6, characterized in that: The control genes include: Actin, ACTB and / or GAPDH.

8. The early detection system according to claim 7, characterized in that: The control gene is Actin, and the primer and probe sequences are shown in SEQ ID NOs: 30-32.

9. The early detection system according to claim 2, characterized in that: The software analysis system includes: integrating different types of clinical data, based on multi-algorithm integrated learning, and a model for joint diagnosis of multiple cancer types.

10. The early detection system according to claim 9, characterized in that: The model includes one of a logistic regression model, a random forest model, a support vector machine, a Bayesian model, a decision tree model, and a neural network model, or a combination thereof.

11. Use of the early detection system according to any one of claims 1 to 10 in the preparation of a digestive system tumor risk prediction or diagnosis kit.

12. The use according to claim 11, characterized in that: The digestive system tumors include: colon cancer, rectal cancer, liver cancer or gastric cancer.

13. The use according to claim 11, characterized in that: The steps of diagnosing various digestive system tumors by the early detection system include: (1) Collect samples; (2) extraction of cfDNA and corresponding bisulfite conversion; (3) The methylation status of four genes, SEPT9, SDC2, SFRP2 and RNF180, was detected and analyzed in combination with clinical data using the optimal model obtained through machine learning training.

14. The use according to claim 13, characterized in that: In step (1), the sample includes: serum, plasma, tissue, cultured cells, blood, feces, saliva, urine or cfDNA.

15. The use according to claim 13, characterized in that: The methylation level detection method includes: BSP first-generation sequencing, NGS second-generation sequencing, qPCR, pyrophosphate sequencing, bisulfite conversion sequencing, methylation chip method, MSP methylation-specific PCR, digital PCR method, third-generation sequencing method, whole genome methylation sequencing method, DNA enrichment detection method, simplified bisulfite sequencing technology, HPLC method, methylation-specific PCR or a combination thereof.

16. The use according to claim 13, characterized in that: In step (3), the detection step is: 1) Set up qPCR reaction according to the number of samples; 2) Collect CT values ​​for qPCR detection.

17. The use according to claim 13, characterized in that: In step (3), the analysis step is: A. Bring the processed quantitative data into the multi-classification logic operation formula for calculation to obtain the probability value P of a given sample belonging to each type of tumor or health; B. Find the classification result with the largest P value among all K classifications; C. Predict cancer risk using P value.

18. The use according to claim 17, characterized in that The logic operation formula is: 。