Marker, kit, and system for screening or diagnosis of early-stage liver cancer, and use

Abstract

A marker, kit, and system for screening or diagnosis of early-stage liver cancer, and a use. The marker comprises any one of AK2, DCTN2, DKK4, IFNGR1, TIMM10 or Galectin 4 protein, and a biomarker panel comprises any combination of AK2, DCTN2, DKK4, IFNGR1, TIMM10 or Galectin 4 protein. A model construction method comprises the following steps: S1, determining a training set, test set samples, and verification set samples, which each comprises healthy individual samples, liver cirrhosis patient samples, and liver cancer patient samples; S2, determining a protein biomarker / biomarker panel and a component weight; S3, using the training set and the test set samples to train a classification algorithm; S4, taking a Youden's index as an optimal threshold for distinguishing liver cancer and non-liver cancer individuals, and taking a Youden's index as an optimal threshold for distinguishing liver cirrhosis and extremely-early-stage liver cancer individuals; and S5, using the verification set samples to evaluate and optimize the performance of a classification algorithm model. The marker can achieve efficient and sensitive screening or diagnosis of early-stage liver cancer.

Description

Marker, kit, system and application for screening or diagnosing early liver cancer TECHNICAL FIELD

[0001] The present application relates to a marker, kit, system and application for screening or diagnosing early liver cancer, belonging to the technical field of biological medicine. BACKGROUND

[0002] Hepatocellular carcinoma (HCC) is a common primary liver cancer in China, which has high malignancy and is basically in the late stage when discovered, and has poor prognosis. In the early stage of liver cancer, the prognosis of patients is relatively good after treatment due to low malignancy. Existing researches show that the detection of high-risk groups of hepatocellular carcinoma will significantly improve the survival rate of HCC patients.

[0003] At present, the common detection methods for liver cancer are alpha fetoprotein (AFP), ultrasound (US), multi-phase dynamic contrast-enhanced computed tomography (CT) and / or dynamic contrast-enhanced magnetic resonance imaging (MRI), digital substraction angiography (DSA) and the like.

[0004] Among them, the continuous increase of serum marker AFP level usually indicates the occurrence of liver cancer, and is a biomarker of liver cancer. The liver cancer diagnosis model based on AFP is also an important method for early liver cancer diagnosis in the prior art.

[0005] The ASAP early liver cancer diagnosis model is composed of four variables of gender, age, AFP and PIVKA-II, and the ASAP early liver cancer diagnosis model has high diagnostic value in hepatocellular carcinoma (HCC) patients.

[0006] The GALAD model established by Johnson et al. includes AFP (alpha fetoprotein), AFP-L3 (alpha fetoprotein isoform L3), abnormal thrombin (DCP), gender, and age, and has very high diagnostic efficiency in liver cancer and early liver cancer patients, which is significantly higher than single or combined detection of AFP, AFP-L3 and DCP.

[0007] The present application aims to discover more markers for early liver cancer and further explore new diagnosis models to realize early screening and diagnosis of liver cancer. SUMMARY

[0008] The application provides six new biomarkers of early liver cancer and further provides a high-efficiency and sensitive classification algorithm model for screening or diagnosing early liver cancer, and provides a new index and method for screening and diagnosing early liver cancer.

[0009] To achieve the above object, the application provides the following technical scheme:

[0010] In the first aspect, the application provides an application of a capture reagent for detecting a protein biomarker / biomarker group in the preparation of a reagent for screening or diagnosing early liver cancer, characterized in that the protein biomarker comprises any one of AK2, DCTN2, DKK4, IFNGR1, TIMM10 or Galectin4 protein; and the protein biomarker group comprises any combination of the above proteins.

[0011] In the application, a group of plasma marker proteins of early liver cancer is provided, including AFP, AK2, DCTN2, DKK4, IFNGR1, TIMM10 and Galectin4 protein, wherein AFP is a known liver cancer marker, and AK2, DCTN2, DKK4, IFNGR1, TIMM10 and Galectin4 are newly discovered plasma markers of liver cancer.

[0012] Preferably, the protein biomarker group further comprises AFP.

[0013] Further preferably, the protein biomarker group comprises AK2, DCTN2, DKK4, IFNGR1, TIMM10, Galectin4 and AFP protein.

[0014] Preferably, the detection comprises detecting the concentration of the biomarker / biomarker group in a biological sample in vitro, and the biological sample comprises peripheral blood and plasma.

[0015] Further preferably, the biological sample is a plasma sample.

[0016] In the second aspect, the application provides a kit for screening or diagnosing early liver cancer, comprising a capture reagent for detecting a protein biomarker / biomarker group, wherein the protein biomarker comprises any one of AK2, DCTN2, DKK4, IFNGR1, TIMM10 or Galectin4 protein; and the protein biomarker group comprises any combination of the above proteins.

[0017] Preferably, the protein biomarker group further comprises AFP.

[0018] Further preferably, the protein biomarker group comprises AK2, DCTN2, DKK4, IFNGR1, TIMM10, Galectin4 and AFP protein.

[0019] Preferably, the capture reagent comprises biotinylated antibody.

[0020] Preferably, the detection comprises detecting the concentration of the biomarker / biomarker group in the biological sample in vitro, wherein the biological sample comprises peripheral blood, plasma.

[0021] Further preferably, the biological sample is a plasma sample.

[0022] Preferably, the kit further comprises reagents and devices for detecting plasma proteins, and can simultaneously detect the concentration of multiple plasma proteins, including sample collection, sample plasma concentration detection device, sample plasma protein concentration real-time display device, etc.

[0023] In a third aspect, the present application provides a method for constructing a classification algorithm model for screening or diagnosing early liver cancer, comprising the following steps:

[0024] S1, determining training set, test set sample and validation set sample, all comprising healthy person sample, liver cirrhosis patient sample and liver cancer patient sample;

[0025] S2, determining protein biomarker / biomarker group and component weight;

[0026] S3, training the classification algorithm using the training set and test set sample;

[0027] S4, making ROC curve of liver cancer patients and non-liver cancer individuals in the test set, and using Youden index as the best threshold value for distinguishing liver cancer and non-liver cancer individuals; making ROC curve of very early liver cancer and liver cirrhosis individuals in the prospective cohort, and using Youden index as the best threshold value for distinguishing liver cirrhosis and very early liver cancer individuals;

[0028] S5, using the validation set sample to evaluate and optimize the performance of the classification algorithm model.

[0029] Preferably, the classification algorithm model is a random forest model, and step S3 comprises: establishing 100-200 decision trees in the training set, using R language as the program, using randomForest as the program package, setting the number of decision trees to 100-200, and setting the number of branches to 3-5.

[0030] Preferably, the protein biomarker comprises any one of AFP, AK2, DCTN2, DKK4, IFNGR1, TIMM10 or Galectin4 protein; the protein biomarker group comprises any combination of AFP, AK2, DCTN2, DKK4, IFNGR1, TIMM10 or Galectin4 protein; and the weight ratio of AFP, AK2, DCTN2, DKK4, IFNGR1, TIMM10, Galectin4 is (3-2) : (2-1) : (2-1) : (2.5-1.5) : (1.5-0.8) : (1.5-1) : (1.5-1).

[0031] Further preferably, the protein biomarker group comprises AK2, DCTN2, DKK4, IFNGR1, TIMM10, Galectin4 and AFP protein.

[0032] Further preferably, the weight ratio of AFP, AK2, DCTN2, DKK4, IFNGR1, TIMM10, Galectin4 is 2.11:1.52:1.32:1.89:0.94:1.03:1.17.

[0033] Preferably, the optimal threshold value for distinguishing liver cancer and non-liver cancer individuals is 0.329, and the optimal threshold value for distinguishing liver cirrhosis and very early liver cancer individuals is 0.338.

[0034] In a fourth aspect of the present application, an early liver cancer diagnosis system based on a classification algorithm is provided, comprising: a processor and a memory, the processor can process sample information according to the classification algorithm model of any one of claims 8-11, including inputting the parameters of the protein biomarker / biomarker group of the individual to be tested into the classification algorithm model in the processor; the classification algorithm model in the processor evaluates the liver cancer risk value of the sample to be tested, and compares with the threshold value to output the risk evaluation result; the memory is used to store threshold data information and evaluation result information.

[0035] Preferably, the memory includes but is not limited to Random Access Memory (RAM), Read Only Memory (ROM), Programmable Read-Only Memory (PROM), Erasable Programmable Read-Only Memory (EPROM), Electric Erasable Programmable Read-Only Memory (EEPROM) and the like.

[0036] Preferably, the processor can be an integrated circuit chip with signal processing capability. The processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; or can be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FGPA) or other programmable logic device, a discrete gate or a crystal logic device, a discrete hardware component.

[0037] Preferably, the diagnosis system can be a server, a cloud platform, a mobile phone, a tablet computer, a notebook computer, an ultra-mobile personal computer (UMPC), a handheld computer, a netbook, a personal digital assistant (PDA), a wearable electronic device, a virtual reality device, etc.

[0038] Compared with the prior art, the application has the advantages and significant progress that the application uses AFP combined with other six indicators to establish an early warning diagnosis model for early liver cancer, has high sensitivity and good specificity compared with the prior art, and can realize early warning for very early liver cancer. Compared with the existing liver cancer markers or early diagnosis models, the warning time can be advanced by more than half a year. BRIEF DESCRIPTION OF DRAWINGS

[0039] To make the technical solutions of the application clearer, the drawings used in the embodiments of the application will be briefly introduced below.

[0040] Figure 1 is the plasma concentration of seven markers of healthy people, cirrhosis patients and early liver cancer patients in Example 1.

[0041] Figure 2 is a flow chart of the construction of the early liver cancer detection model in Example 3.

[0042] Figure 3 is the weight of each marker in the early liver cancer diagnosis model in Example 3.

[0043] Figure 4 is the number of random forest classification trees and their error values in Example 3.

[0044] Figure 5 is the ROC curve and the optimal threshold of liver cancer and non-liver cancer in Example 3.

[0045] Figure 6 is the ROC curve and the optimal threshold of cirrhosis and very early stage liver cancer in Example 3;

[0046] Figure 7 is the Risk Score of the diagnosis model of healthy people, cirrhosis patients, very early stage liver cancer patients and early stage liver cancer patients in Example 3;

[0047] Figure 8 is a column chart of the statistical AUC of the classification results of liver cancer in Example 4 using the integrated model in the test set and the validation set compared with the liver cancer related index (AFP);

[0048] Figure 9 is the AUC value comparison of the prediction effect of very early stage liver cancer of the result statistics of the integrated model in Example 5 compared with known liver cancer prediction models or clinical indicators. Including known ASAP liver cancer early diagnosis model, PIVKA-II index, Galad model and AFP index. According to the AUC value, it can be known that the RiskScore integrated model has the highest accuracy;

[0049] Figure 10 is the positive rate comparison of the prediction of the result statistics of the integrated model in Example 6 compared with known liver cancer prediction models or clinical indicators for liver cirrhosis later stage cancer. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical scheme and advantages of the embodiments of the present application more clear, the present application will be further described below in combination with specific examples. It should be understood that these examples are only used to illustrate the present application and are not used to limit the scope of the present application. The experimental methods in the following examples are not specified, which are usually according to the conventional conditions or according to the conditions suggested by the manufacturers. Unless otherwise specified, percentages and parts are weight percentages and weight parts. The experimental materials and reagents used in the following examples are commercially available unless otherwise specified.

[0051] Unless otherwise indicated, the technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. It should be noted that the terms used herein are only for the purpose of describing the specific embodiments of the present application, and are not intended to limit the exemplary embodiments of the present application.

[0052] In order to more fully understand the present application, the professional terms in the present application are explained as follows.

[0053] AFP, is alpha-fetoprotein, which is a special protein contained in the blood during the embryonic period, only in the fetal liver cells can be synthesized, especially in 16-20 weeks of pregnancy, the highest content of fetal protein, 100 ml of blood can reach 300-400 mg, then gradually reduced, after birth a week completely disappeared. Because it only exists in the blood of the fetus, so the full name is called alpha-fetal protein. Normal liver cells do not produce alpha-fetoprotein, but when you get liver cancer, because the liver cells are immature and grow indefinitely, they have resumed the ability to synthesize alpha-fetoprotein, sometimes even up to 1000 mg in 100 ml of blood, which is the reason for detecting alpha-fetoprotein to help diagnose liver cancer.

[0054] AK2, is adenosine kinase 2 protein.

[0055] DCTN2, is dynein activator protein 2.

[0056] DKK4, is dickkopf WNT signaling pathway inhibitor 4.

[0057] IFNGR1, is interferon gamma receptor 1.

[0058] TIMM10, is mitochondrial inner membrane translocase 10.

[0059] Galectin4, is galectin 4.

[0060] Kit, refers to the box used to hold the detection of chemical composition, drug residues or virus species and other chemical reagents, generally more in hospitals, inspection and quarantine or pharmaceutical companies. Reagent kit is an important tool in chemical analysis experiment, its role is used to detect the substances contained in the sample. In biochemical detection, reagent kit is also an important method for discovering diseases. Reagent kit can be used for enzyme labeled method, immunoassay method and other detection methods.

[0061] Example 1 Screening of early liver cancer diagnosis markers

[0062] This example detects the protein concentration in the plasma of three types of people (healthy people, liver cirrhosis patients, and early liver cancer patients). And screening 7 kinds of biomarkers related to early liver cancer.

[0063] 1.1, plasma separation: peripheral blood collection, using disposable non-heat, non-endotoxin test tube (EDTA, citrate, heparin anticoagulation can be used), avoid the use of hemolysis, hyperlipidemia samples, specimen suspension should be centrifuged to remove, make the specimen clear and transparent. The plasma to be tested should be detected as soon as possible, stored at 2-8℃ for 48 hours, longer time should be frozen (-20℃ or -80℃) storage, avoid repeated freeze-thaw.

[0064] 1.2, Collect the peripheral blood samples at 3000 rpm for 10 minutes.

[0065] 1.3, The resulting plasma is divided into 1.5 ml EP tubes.

[0066] 1.4, Standard preparation: Take 8 1.5 ml centrifuge tubes, respectively marked S1, S2, S3, S4, S5, S6, S7, blank, add 900ul of standard / sample diluent to the first tube S1, add 200ul of standard / sample diluent to the second to eighth tubes, add 100ul of standard solution (100.0ng / ml) to the first tube, mix well on the vortex mixer, then use the pipette to remove 200ul, move to the second tube, repeat the same for S7.

[0067] 1.5, Biotinylated antibody working solution configuration: 20 minutes before use, dilute 100x biotinylated antibody to 1x working solution with biotinylated antibody diluent, configure according to the required amount, use it on the same day, and discard the rest.

[0068] 1.6, TMB color developing solution configuration: 10 minutes before use, mix TMB color developing solution A and B 1:1, avoid light, and prepare for use.

[0069] 1.7, Detection procedure:

[0070] 1) Sample addition: add 50ul of standard / sample diluent to the blank well, and add 50ul of standard or sample to be tested to the rest of the wells, mix the reaction plate, and place it at 37℃ for 50 minutes;

[0071] 2) Plate washing: wash the reaction plate with 1x washing solution for 3 times, add 300ul of 1x washing solution to each well, shake / saturate for 1-2 minutes each time, and dry on filter paper;

[0072] 3) Add 100ul of biotinylated antibody diluent to the blank well, and add 100ul of 1x biotinylated antibody working solution to the rest of the wells, mix well, and place it at 37℃ for 50 minutes;

[0073] 4) Plate washing: same as above;

[0074] 5) Add 100ul of SABC working solution to each well, mix well, and place it at 37℃ for 30 minutes;

[0075] 6) Plate washing: same as above;

[0076] 7) Add 100ul of TMB mixed solution prepared in advance to each well, mix well, and place it at 37℃ in the dark for 10-20 minutes (the specific color developing time is determined according to the color developing result);

[0077] 8) Add 50ul of stop solution to each well, mix well, and measure the absorbance at 450nm within 30 minutes.

[0078] 1.8、Result calculation

[0079] All OD values are suggested to be calculated after subtracting the blank well value. If the blank well OD is lower than 0.1, it can also be calculated directly. Draw the standard curve by using the standard concentration as the horizontal coordinate and the OD value as the vertical coordinate, or use software to draw the standard curve. Calculate the corresponding content according to the sample OD value.

[0080] In this example, 7 biomarkers related to early liver cancer were screened, which were AFP, AK2, DKK4, DCTN2, IFNGR1, Galentin4 and TIMM10. The contents of the 7 biomarkers in each group of people are shown in Figure 1.

[0081] Example 2: Early liver cancer prediction based on 7 biomarker groups of classification algorithm model

[0082] This example provides a method for predicting early liver cancer using the 7 biomarker groups (AFP, AK2, DKK4, DCTN2, IFNGR1, Galentin4 and TIMM10) screened in Example 1, which includes the following steps:

[0083] 2.1, Obtain the plasma of the person to be detected. When collecting blood, a disposable non-heat source, non-endotoxin test tube (EDTA, citrate, heparin anticoagulant) treated anticoagulant vessel should be used for collection;

[0084] 2.2, centrifugation at 3000rpm for 10 minutes is required. (The blood sample should be centrifuged within 1h after collection at room temperature; the blood sample can be processed within 4h after collection on ice), take the supernatant;

[0085] 2.3, detect the concentrations of AFP, AK2, DCTN2, IFNGR1, TIMM10, Galectin4 and DKK4 in the plasma;

[0086] 2.4, input the results obtained in step 2.3 into the classification algorithm model to obtain a score;

[0087] 2.5, compare the score obtained in step 2.4 with the set threshold value, and the comparison result is used to evaluate the risk of liver cancer of the detected person.

[0088] In the present embodiment, any one of classification algorithms in random forest, vector machine (SVM), logistic regression, ridge regression and deep neural network can be used to construct a model for the 7 classification indexes.

[0089] Example 3 Training of Random Forest Model

[0090] As shown in Figure 2, the present embodiment uses the peripheral blood samples of the subjects and the 7 biomarker groups screened in Example 1 to train the random forest model.

[0091] 3.1, 518 subjects were collected, including 188 healthy people, 155 patients with cirrhosis, 175 patients with liver cancer, and 77 patients with very early liver cancer. The plasma samples were obtained by centrifugation, and the concentrations of AFP, AK2, DKK4, DCTN2, IFNGR1, Galentin4 and TIMM10 in the plasma were detected. The samples recruited were randomly divided into training set and test set according to the ratio of 7:3, and the training set contained a total of 361 samples, including 120 healthy people, 111 patients with cirrhosis and 130 patients with liver cancer;

[0092] 3.2, as shown in Figure 3, the weights of the 7 proteins were determined, and the weight ratio of AFP, AK2, DCTN2, DKK4, IFNGR1, TIMM10 and Galectin4 was 2.11:1.52:1.32:1.89:0.94:1.03:1.17;

[0093] 3.3, the plasma protein concentration of each marker in the training set sample was used as input data to construct a random forest model, and the algorithm process of the random forest was as follows:

[0094] 1) N decision trees were established in the training set, and the number of classification trees and their error values are shown in Figure 4;

[0095] 2) According to the model error value, the number of decision trees was 100;

[0096] 3) The program used was R language, the program package used was randomForest, the number of decision trees was set to 100 (ntree=100), and the number of branches was set to 3 (mtry=3);

[0097] 3.3, the output result of the model is the liver cancer risk score of the sample;

[0098] 3.4 In addition, plasma from 126 patients with early-stage liver cancer, 92 patients with cirrhosis, and 126 healthy individuals was collected as an external validation set. Plasma from 78 patients with very early-stage liver cancer was collected as a prospective cohort to validate the model.

[0099] 3.5 Threshold Classification: ROC curves were plotted for hepatocellular carcinoma (HCC) patients and non-HCC individuals in the test set (as shown in Figure 5). The Youden Index was used as the optimal threshold for distinguishing between HCC and non-HCC individuals. ROC curves were plotted for very early-stage HCC and cirrhosis individuals in the prospective cohort (as shown in Figure 6). The Youden Index was used as the optimal threshold for distinguishing between cirrhosis and very early-stage HCC individuals.

[0100] 3.6. Compare the obtained liver cancer risk score with the known thresholds to obtain the liver cancer risk level. As shown in Figure 7, the risk level assessment rules of the liver cancer detection model are as follows: a risk score below 0.329 indicates a low-risk group; a score above 0.329 but below 0.338 indicates a medium-risk group with a higher probability of developing liver cancer within six months; and a risk score above 0.338 indicates a high-risk group, and regular monitoring of liver lesions is recommended.

[0101] Example 4: Comparison Experiment between the Integrated Model of 7 Markers and the Single Marker Model

[0102] 4.1 The blood samples to be tested were divided into test set samples and validation set samples using a method similar to that in Example 3;

[0103] 4.2 The construction of the integrated model is shown in Example 3. Using a method similar to Example 3, a random forest model is trained using only the AFP single marker to obtain the AFP group model;

[0104] 4.3 The model evaluation metrics AUC for the test set and validation set samples in the integrated model and the AFP group model are shown in Figure 8. Figure 8 clearly shows that the integrated model outperforms the single AFP group model.

[0105] This embodiment also compares the AUC values ​​of the integrated model with those of other individual markers on the test set. The results are shown in Table 1 below.

[0106] Table 1

[0107] Based on the AUC of the model in Table 1, it can be determined that the AUC of the integrated score is significantly greater than the AUC of other classification indicators. This embodiment verifies that the combination of the seven biomarkers of this invention is significantly more effective in diagnosing early-stage liver cancer than the existing techniques that use only one biomarker.

[0108] Example 5: Comparative Experiment of Several Liver Cancer Screening Methods

[0109] The present embodiment compares the liver cancer screening model provided in Embodiment 3 of the present application with the ASAP liver cancer early diagnosis model and the ASAP model and the single index model thereof in the prior art.

[0110] The specific operation steps are similar to those in Embodiment 4, and the AUC values of different models are compared to determine the accuracy of the models.

[0111] The results are shown in FIG. 9. The model provided in the present application is more sensitive and efficient than the ASAP liver cancer early diagnosis model and the ASAP model in the prior art.

[0112] Embodiment 6

[0113] In the present embodiment, the plasma of 78 individuals with cirrhosis is collected, and these individuals are diagnosed with cancer within six months to three years after blood sampling. The protein concentration in the plasma is detected and introduced into the model.

[0114] The results are shown in FIG. 10. When the threshold value of the method (risk score) in the present application is 0.338, 55 cases are detected as positive, and the positive rate is 70.5%. The positive rate of AFP is 16.7%, the positive rate of PIVAK-II is 11.5%, the positive rate of Galad model is 37.2%, and the positive rate of ASAP model is 31.6%. It is shown that the early liver cancer prediction diagnosis model has a high accuracy in detecting very early liver cancer, and the early warning time can be advanced by more than half a year.

[0115] Embodiment 7

[0116] The present embodiment is an early liver cancer diagnosis system, which comprises a processor and a memory. The memory is used to store one or more programs, and when the programs are executed by the processor, the processor realizes the construction of a random forest model.

[0117] The electronic device of the present embodiment can include a memory, a processor, a bus and a communication interface, which are directly or indirectly electrically connected to each other to realize the interaction and transmission of data.

[0118] In the embodiment, the memory can be, but is not limited to, a random access memory (RAM), a read only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), and the like.

[0119] In the embodiment, the processor can be an integrated circuit chip with a signal processing capability. The processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), and the like; or can be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FGPA) or other programmable logic device, a discrete gate or a crystal logic device, a discrete hardware component.

[0120] In actual application, the diagnostic system can be an electronic device, which can be a server, a cloud platform, a mobile phone, a tablet computer, a notebook computer, an ultra-mobile personal computer (UMPC), a handheld computer, a netbook, a personal digital assistant (PDA), a wearable electronic device, a virtual reality device, and the like. Therefore, the type of the electronic device is not limited in the embodiment.

[0121] In the embodiment, a device for detecting plasma proteins is also provided, which can detect the concentrations of multiple plasma proteins at the same time, including sample acquisition, sample plasma concentration detection device, and sample plasma protein concentration real-time display device.

[0122] In the embodiment, a computer readable medium having a computer program stored thereon is also provided, and the computer program can be executed by the processor to implement the construction of the random forest model. The computer readable medium includes a U disk, a mobile hard disk, a read only memory, a random access memory, a magnetic disk, a floppy disk, an optical disk, and the like, which can store program codes.

[0123] Applicant states that during the description of the above specification:

[0124] The description of the terms "the embodiments", "the embodiments of the present application", "as shown", "further", "further improved technical solutions" and the like means that the specific features, structures, materials or characteristics described in the embodiments or examples are contained in at least one embodiment or example of the present application; in the present specification, the illustrative description of the above terms is not necessarily directed to the same embodiment or example, and the specific features, structures, materials or characteristics described can be combined or combined in any one or more embodiments or examples in a suitable manner; in addition, the person skilled in the art can combine or combine the different embodiments or examples described in the specification and the features of the different embodiments or examples without causing contradiction.

[0125] Finally, it should be noted that:

[0126] The above embodiments are only used to illustrate the technical solutions of the present application, but not to limit them;

[0127] Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part or all of the technical features, and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application, and the non-essential improvements and adjustments or replacements made by the person skilled in the art according to the content of the specification are all within the scope of the present application.

Claims

1. Use of a capture reagent for detecting a protein biomarker / biomarker panel in the manufacture of a reagent for screening or diagnosing early stage liver cancer, characterized in that, The protein biomarker comprises any one of AK2, DCTN2, DKK4, IFNGR1, TIMM10 or Galectin4 protein; and the protein biomarker group comprises any combination of AK2, DCTN2, DKK4, IFNGR1, TIMM10 or Galectin4 protein.

2. Use according to claim 1, characterized in that, The protein biomarker group further comprises AFP.

3. Use according to claim 1, characterized in that, The detection comprises in-vitro detection of the concentration of the biomarker / biomarker group in a biological sample, wherein the biological sample comprises peripheral blood, plasma.

4. A kit for screening or diagnosing early stage liver cancer, characterized by, The capture reagent for detecting the protein biomarker / biomarker group comprises any one of AK2, DCTN2, DKK4, IFNGR1, TIMM10 or Galectin4 protein; and the protein biomarker group comprises any combination of AK2, DCTN2, DKK4, IFNGR1, TIMM10 or Galectin4 protein.

5. The kit for screening or diagnosing early stage liver cancer according to claim 4, wherein The protein biomarker group further comprises AFP.

6. The kit for screening or diagnosing early stage liver cancer according to claim 4, wherein The capture reagent comprises a biotinylated antibody.

7. The kit for screening or diagnosing early stage liver cancer according to claim 4, wherein The detection comprises in-vitro detection of the concentration of the biomarker / biomarker group in a biological sample, wherein the biological sample comprises peripheral blood, plasma. 8.A method for constructing a classification algorithm model for screening or diagnosing early liver cancer, characterized in that, The method comprises the following steps: S1, determining training set, test set sample and validation set sample, which all comprise healthy person sample, liver cirrhosis patient sample and liver cancer patient sample; S2, determining protein biomarker / biomarker group and component weight; S3, training classification algorithm by using the training set and test set sample; S4, making ROC curve of liver cancer patient and non-liver cancer individual in the test set, and using Youden index as the best threshold value for distinguishing liver cancer and non-liver cancer individual; making ROC curve of liver cirrhosis individual and extremely early liver cancer individual in the prospective cohort, and using Youden index as the best threshold value for distinguishing liver cirrhosis and extremely early liver cancer individual; S5, evaluating and optimizing the performance of the classification algorithm model by using the validation set sample. 9.The method of claim 8, wherein the method is characterized by, The classification algorithm model is a random forest model, and step S3 further comprises: establishing 100-200 decision trees in the training set, using R language as the program, using randomForest as the program package, setting the number of decision trees to 100-200, and setting the number of branches to 3-5.

10. The construction method of the classification algorithm model for screening or diagnosing early liver cancer according to claim 8, characterized in that, The protein biomarker comprises any one of AFP, AK2, DCTN2, DKK4, IFNGR1, TIMM10 or Galectin4 protein; and the protein biomarker group comprises any combination of AFP, AK2, DCTN2, DKK4, IFNGR1, TIMM10 or Galectin4 protein. The protein biomarker comprises any one of AFP, AK2, DCTN2, DKK4, IFNGR1, TIMM10 or Galectin4 protein; and the protein biomarker group comprises any combination of AFP, AK2, DCTN2, DKK4, IFNGR1, TIMM10 or Galectin4 protein. and the weight ratio of AFP, AK2, DCTN2, DKK4, IFNGR1, TIMM10, Galectin4 is (3-2) : (2-1) : (2-1) : (2.5-1.5) : (1.5-0.8) : (1.5-1) : (1.5-1). 11.The method of claim 8, wherein the method is characterized by, The optimal threshold value for distinguishing liver cancer and non-liver cancer individuals is 0.329, and the optimal threshold value for distinguishing liver cirrhosis and very early liver cancer individuals is 0.

338.

12. A system for early diagnosis of liver cancer based on a classification algorithm, characterized by, It comprises: a processor and a memory, The processor can process sample information according to the classification algorithm model of any one of claims 8-11, including inputting the protein biomarker / biomarker group parameters of the individual to be tested into the classification algorithm model in the processor; the classification algorithm model in the processor evaluates the liver cancer risk value of the sample to be tested, and compares with the threshold value to output the risk evaluation result; The memory is used to store threshold data information and evaluation result information.

Images