Application of biomarker UGT2B15 in colorectal adenoma diagnosis

Abstract

The invention discloses an application of a biomarker UGT2B15 in colorectal adenoma diagnosis, and provides an application of a biomarker for diagnosing or predicting colorectal adenoma or / and a detection reagent thereof in preparation of a product for diagnosing or predicting colorectal adenoma. The invention provides a kit for diagnosing colorectal adenoma or predicting colorectal adenoma in a subject, provides an electronic device for diagnosing colorectal adenoma or predicting colorectal adenoma, and provides a screening method of a biomarker for diagnosing colorectal adenoma or predicting colorectal adenoma. The invention provides a method for constructing a diagnostic model of colorectal adenoma, and provides a processor, a computer medium and a device.

Description

Technical Field

[0001] The invention belongs to the field of biotechnology and relates to the application of a biomarker UGT2B15 in the diagnosis of colorectal adenoma. Background Art

[0002] Colorectal adenoma is considered to be the main precancerous lesion of colorectal cancer. It is generally believed that it takes 7 to 10 years for adenoma to develop into cancer, which leaves enough time and space for precise intervention and treatment of colorectal cancer. Therefore, early detection of colorectal adenoma and timely and effective intervention are crucial.

[0003] At present, there are mainly the following methods for screening colorectal adenomas: colonoscopy is an invasive procedure with low patient compliance. At the same time, colonoscopy resources are limited and affected by endoscopic examination technology, resulting in a low detection rate and a high missed diagnosis rate of colorectal adenomas; stool testing is easily affected by food, drugs or other intestinal diseases, resulting in a high false positive rate; tumor marker testing is often used for screening of colorectal cancer, but has low sensitivity and specificity for colorectal adenomas; CT colon imaging technology has radiation hazards and there is currently no standardized method, resulting in low performance in diagnosing adenomas. Summary of the invention

[0004] In order to solve the technical problems existing in the prior art, the present invention provides the following technical solutions:

[0005] The present invention provides the use of a UGT2B15 biomarker for diagnosing or predicting colorectal adenoma or / and a detection reagent thereof in preparing a product for diagnosing or predicting colorectal adenoma.

[0006] Furthermore, the products include reagents, kits, primers, chips, antibodies, and test strips.

[0007] The term "UGT2B15" as used herein is intended to encompass fragments, variants (e.g., allelic variants), and derivatives thereof. Representative human UGT2B15 cDNAs and human UGT2B15 protein sequences are well known in the art and are publicly available from the NCBI website. For example, at least one human UGT2B15 isoform is known: human UGT2B15, Gene ID: 7366. Nucleic acid and polypeptide sequences of human UGT2B15 orthologs in organisms other than humans are well known.

[0008] Furthermore, the sample types detected by the detection reagent include tissues, blood, plasma, serum, and blood cells.

[0009] Furthermore, the sample types include tissues, blood, and plasma.

[0010] Furthermore, the sample types are tissues and blood.

[0011] Furthermore, the sample type is tissue.

[0012] The term "sample" used in the present invention refers to any composition containing nucleic acid or protein isolated from a subject. In some embodiments, the sample can be selected from at least one of blood, tissue, blood cells, bone marrow, ascites, fine needle biopsy samples, body fluids containing cells, free floating nucleic acids, sputum, saliva, urine, semen, cerebrospinal fluid, peritoneal fluid, pleural fluid, feces, lymph, skin swab, oral swab, nasal swab or lavage. In some embodiments, the sample can be blood or tissue.

[0013] Furthermore, the detection reagent is used to detect the content of the biomarker.

[0014] Furthermore, the detection reagent includes a reagent for detecting the gene expression amount or protein expression level of the biomarker.

[0015] The present invention provides a kit for diagnosing colorectal adenoma or predicting colorectal adenoma in a subject, wherein the kit comprises a detection reagent for detecting the content of a biomarker, wherein the biomarker comprises UGT2B15.

[0016] The present invention provides an electronic device for diagnosing or predicting colorectal adenoma, the electronic device comprising:

[0017] An acquisition and detection module is configured to acquire a sample, detect the sample, and obtain the content of the biomarker in the sample;

[0018] The diagnosis or prediction module uses a machine learning method and the content of the biomarker to establish a regression equation, construct a model, and then outputs a diagnosis or prediction result according to the model.

[0019] The term "electronic device" as used in the present invention refers to any suitable computing or processing device, or other device constructed or modified to store data or information. Examples of electronic devices suitable for use with embodiments of the present invention include stand-alone computing devices; networks, including local area networks (LANs), wide area networks (WANs), the Internet, corporate intranets, and extranets; electronic appliances such as personal digital assistants (PDAs), cell phones, page managers, etc.; and local and distributed processing systems.

[0020] Furthermore, the machine learning method includes a linear regression algorithm model, a support vector machine algorithm model, a nearest neighbor / k-nearest neighbor algorithm model, a logistic regression algorithm model, a decision tree algorithm model, a k-means algorithm model, a random forest algorithm model, a naive Bayes algorithm model, a dimensionality reduction algorithm model, and a gradient enhancement algorithm model.

[0021] The term "machine learning" used in the present invention refers to an algorithm that gives a computer learning ability without explicit programming, including an algorithm that learns from data and makes predictions about the data. The machine learning algorithms used in the embodiments disclosed herein may include (but are not limited to) random forest (RF), least absolute shrinkage and selection operator (LASSO) logistic regression, regularized logistic regression, XGBoost, decision tree learning, artificial neural network (ANN), deep neural network (DNN), support vector machine, rule-based machine learning, etc. Algorithms such as linear regression or logistic regression can be used as part of the machine learning process.

[0022] Furthermore, the sample includes tissue, blood, plasma, serum, and blood cells.

[0023] Furthermore, the sample includes tissue, blood, and plasma.

[0024] Furthermore, the sample is tissue or blood.

[0025] Furthermore, the sample is tissue.

[0026] Further, the biomarker includes UGT2B15.

[0027] In some embodiments, the method for detecting the expression level of a gene or protein in a sample is a sequencing method, and the sequencing method is not limited to a second generation sequencing method or a third generation sequencing method. The means for sequencing is not particularly limited, and sequencing by a second generation or third generation sequencing method can achieve fast and efficient sequencing. The high throughput and deep sequencing characteristics of sequencing can be utilized, which is beneficial to the analysis of subsequent data, especially the accuracy of statistical test analysis.

[0028] The present invention provides a method for screening biomarkers for diagnosing or predicting colorectal adenoma, the screening method comprising:

[0029] Obtaining the colorectal adenoma disease conditions and samples of a sample group, testing the samples to obtain the gene or protein expression levels of the samples, wherein the sample group includes healthy individuals and colorectal adenoma patients;

[0030] According to the test results of the sample, screening genes or proteins with significant differences between the healthy individuals and the colorectal adenoma patients by statistical methods;

[0031] Using the contents of genes or proteins with significant differences screened out from the sample population and the prevalence of colorectal adenoma in the sample population, a prediction model is constructed through a machine learning training method;

[0032] The gene or protein that appears more than a threshold number of times in the prediction model is the biomarker.

[0033] Furthermore, the selection of the threshold in the screening method needs to be subjected to random segmentation, the number of random segmentations is multiple, the number of prediction models constructed is multiple, and the prediction model needs to be cross-validated multiple times.

[0034] Furthermore, the statistical method includes any one or more of T test, generalized linear model, rank test, logistic regression, multiple difference, multiple hypothesis testing correction or Kruskal-Wallis test.

[0035] The above statistical methods for judging the significant difference detection results include but are not limited to T-test, Wilcoxontest, logistic regression, KW test, FoldChange, Bonferroni correction or one or more of other existing statistical methods. When constructing a prediction model, the colorectal adenoma prevalence of each sample and the expression of genes or proteins in the significant difference detection results are summarized to obtain the summarized data of different individuals, and then machine learning training is performed, and cross-validation is used to construct a prediction model.

[0036] Furthermore, the sample includes tissue, blood, plasma, serum, and blood cells.

[0037] Furthermore, the sample includes tissue, blood, and plasma.

[0038] Furthermore, the sample is tissue or blood.

[0039] Furthermore, the sample is tissue.

[0040] Furthermore, the process of constructing the prediction model includes dividing the sample group into a training set and a test set, establishing a mapping relationship with the colorectal adenoma prevalence, obtaining a training data set or a test data set respectively, and obtaining the prediction model through a machine learning training method and cross-validation.

[0041] In the above-mentioned biomarker screening method, samples including healthy individuals and colorectal adenoma patients and the prevalence of colorectal adenoma are first obtained. The gene or protein expression level of the sample is detected, and a certain number of genes or proteins are selected as the target to be detected. After the detection, the content of the gene or protein to be detected in each sample is obtained. Statistical methods are used to test the genes or proteins that are different between colorectal adenoma patients and healthy individuals, which is the significant difference detection result. The significant difference detection result shows that the expression level of the genes or proteins of these test results indicates the number of existence in the sample population, which is related to the prevalence of colorectal adenoma. In the healthy group and colorectal adenoma patients, the expression level of the genes or proteins of the significant difference detection results is significantly different, but it is impossible to confirm the effect of the significant difference detection results on diagnosis or prediction, and it is not clear how to diagnose / predict.

[0042] In some embodiments, in order to improve and ensure the effectiveness of diagnosing / predicting colorectal adenoma, machine learning methods are used to model the content of significant difference detection results of sample groups and the prevalence of colorectal adenoma, and a predictive model capable of diagnosing / predicting colorectal adenoma is constructed. The genes or proteins obtained from the predictive model are all genes or proteins that have been screened out more than a threshold number of times and have good effects, that is, the final biomarkers.

[0043] In some embodiments, in the process of establishing a prediction model, one or more analysis methods including but not limited to LASSO regression, BIC (Bayesian Information Criterion), and stepwise regression are used to analyze the established prediction model and select a prediction model with high effectiveness.

[0044] The present invention provides a method for constructing a diagnostic model for colorectal adenoma, wherein the construction method uses the aforementioned screening method to obtain biomarkers, and applies a machine learning training method to construct a diagnostic model.

[0045] Further, the biomarker includes UGT2B15.

[0046] The present invention provides a processor, which is used to run a program, and when the program is run, it executes the above-mentioned screening method, the above-mentioned construction method, or the above-mentioned diagnosis model.

[0047] The present invention provides a computer medium, which includes a storage medium, and the storage medium is used to store a program. When the program is running, the device connected to the storage medium is controlled to execute the screening method, the construction method, or the diagnostic model described above.

[0048] In some embodiments, the present invention can be used in a variety of general or special computing system environments, such as personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, and the like.

[0049] In some embodiments, the above modules or steps can achieve the effect through a computing device. In some embodiments, the modules or steps can be concentrated on a single computing device. In some embodiments, the modules or steps can be distributed on multiple computing devices, and multiple computing devices are linked by a network or other linking methods. In some embodiments, the computing device stores a program code for executing the above modules or steps, and the program code is stored in a storage device. In some embodiments, the modules or steps can be prepared into integrated circuits respectively, or combined into a single integrated circuit. In the specific embodiments of the present invention, no implementation method is limited, whether it is hardware or software.

[0050] The present invention provides the use of biomarkers in constructing a system for automatically diagnosing or predicting colorectal adenoma, wherein the biomarkers include UGT2B15.

[0051] The present invention provides a device for screening biomarkers for diagnosing colorectal adenoma or predicting colorectal adenoma, the device comprising:

[0052] An acquisition module is configured to acquire the colorectal adenoma disease conditions and samples of a sample group, detect the samples, and obtain the gene or protein expression levels of the samples, wherein the sample group includes healthy individuals and colorectal adenoma patients;

[0053] A statistical module is configured to use the test results of the sample to screen genes or proteins that are significantly different between the healthy individual and the colorectal adenoma patient by statistical methods;

[0054] A model building module is configured to build a prediction model by using the contents of genes or proteins with significant differences screened out from the sample population and the colorectal adenoma morbidity of the sample population through a machine learning training method;

[0055] The biomarker determination module is configured to define a gene or protein that appears in the prediction model more than a threshold value as the biomarker.

[0056] The term "module" used in the present invention refers to a combination of software and / or hardware that can implement a predetermined function. Implementation of hardware, software, or a combination of hardware and software is conceivable.

[0057] The term "diagnosis" used in the present invention refers to determining the health status of a subject, including detecting the presence or absence of a disease, response to treatment, recurrence risk assessment, assessment of cancer risk and degree of cancer, prognosis, etc. In some cases, the term "diagnosis" refers to a single factor used to determine, verify or confirm the clinical status of a patient.

[0058] In some specific embodiments, the device for screening biomarkers for diagnosing colorectal adenoma or predicting colorectal adenoma can screen biomarkers for diagnosing colorectal adenoma or predicting colorectal adenoma by detecting the gene or protein content in the sample of the sample population. In some embodiments, the device for screening biomarkers for diagnosing colorectal adenoma or predicting colorectal adenoma can perform new screening of biomarkers or partially update biomarkers when the biomarker prediction effect is affected by factors such as the region and environment of the subject.

[0059] Furthermore, the gene or protein expression levels detected in the sample include UGT2B15.

[0060] The term "ROC curve" used in the present invention is a receiver operating characteristic curve, which is a coordinate diagram composed of a false positive probability (1-specificity) as the horizontal axis and a true positive probability (sensitivity) as the vertical axis, and a curve drawn by different results obtained by the subject under specific stimulation conditions due to different judgment criteria. Select the best diagnostic limit value. The closer the ROC curve is to the upper left corner, the higher the accuracy of the test. The point of the ROC curve closest to the upper left corner is the best threshold with the least error, and its total number of false positives and false negatives is the least. Comparison of disease recognition ability of two or more different diagnostic tests. When comparing two or more diagnostic methods for the same disease, the ROC curves of each test can be drawn into the same coordinate to intuitively distinguish the pros and cons. The subject represented by the ROC curve close to the upper left corner works most accurately. It can also be compared by calculating the area under the ROC curve (AUC) of each test respectively. Which test has the largest AUC, then which test has the best diagnostic value.

[0061] Advantages and beneficial effects of the present invention:

[0062] The present invention uses whole transcriptome analysis technology to explore a new biomarker UGT2B15 related to colorectal adenoma. The biomarker UGT2B15 can screen patients with colorectal adenoma with high sensitivity and specificity. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 It is the HE staining image of the test set of colorectal adenoma tissue and adjacent normal tissue; CRA is colorectal adenoma tissue, NOR is adjacent normal tissue;

[0064] Figure 2 This is the differential expression result diagram of UGT2B15 in colorectal adenoma tissue and adjacent normal tissue in the test set;

[0065] Figure 3 This is the ROC curve of UGT2B15 in the test set for colorectal adenoma diagnosis;

[0066] Figure 4 It is to verify the HE staining of colorectal adenoma tissue and adjacent normal tissue; CRA is colorectal adenoma tissue, NOR is adjacent normal tissue;

[0067] Figure 5 This is the immunohistochemical staining of UGT2B15; CRA is adenoma tissue, NOR is normal tissue adjacent to the tumor; positive expression is yellow-brown staining of the cytoplasm and blue staining of the nucleus;

[0068] Figure 6 It is a bar graph of the average optical density values ​​of UGT2B15 in the validation set in colorectal adenoma tissue and adjacent normal tissue;

[0069] Figure 7 This is the ROC curve of UGT2B15 for the diagnosis of colorectal adenoma in the validation set. DETAILED DESCRIPTION

[0070] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present invention will be described in detail below in conjunction with the embodiments.

[0071] Example 1 Marker UGT2B15 is associated with the diagnosis of colorectal adenoma

[0072] 1. Clinical samples

[0073] The whole transcriptome sequencing (RNA sequencing, RNA-seq) test set used 12 cases of colorectal adenoma and 12 cases of adjacent normal tissues, a total of 24 cases, the specific information is shown in Table 1. The samples used were all formaldehyde-fixed paraffin-embedded tissues of adenomas that were removed by colonoscopy and surgery at Wangjing Hospital between 2019 and 2023. The research samples were again confirmed as colorectal adenomas and adjacent normal tissues by 2 senior pathologists ( Figure 1 ).

[0074] Table 1 Clinical information of patients in the RNA-seq test set

[0075] serial number gender age Pathological type Location Grading level 1 female 75 Tubular adenoma Hepatic flexure of colon Mild to moderate dysplasia 2 male 58 Tubular adenoma Sigmoid colon Mild dysplasia 3 female 75 Tubular adenoma Hepatic flexure of colon Mild to moderate dysplasia 4 female 73 Tubular adenoma Transverse colon Mild dysplasia 5 male 83 Tubular adenoma Sigmoid colon Mild dysplasia 6 male 63 Tubular adenoma B-level Moderate to severe dysplasia 7 female 52 Tubular adenoma Sigmoid colon Moderate to severe dysplasia 8 female 50 Tubular adenoma Transverse colon Mild dysplasia 9 male 54 Tubular adenoma Sigmoid colon Mild to moderate dysplasia 10 male 60 Tubular adenoma Sigmoid colon Mild to moderate dysplasia 11 male 57 Tubular adenoma rectum Moderate to severe dysplasia 12 male 59 Tubulovillary adenoma Zhiyi Junction Moderate to severe dysplasia

[0076] 2. Experimental equipment

[0077] Agilent 2100 Bioanalyzer (Agilent), Qubit 2.0 Fluorometer (Invitrogen), Veriti 96-well Multiplex Temperature-controlled Gradient Gene Amplifier (MJ Research, USA), Peltier Thermal Cycler PTC-225 (MJ Research, USA), DNA Quantification Kit (KAPABiosystem), Multi-sample Adapter Primer Kit (NEBNext), High Sensitivity DNA Detection Kit (Agilent), RNA 6000 Pico chip (Agilent).

[0078] 3. Experimental methods

[0079] Whole transcriptome sequencing was performed on 12 cases of colorectal adenoma tissues and 12 cases of adjacent normal tissues. The sequencing process mainly included RNA extraction, library construction, sequencing, data quality control, and data analysis.

[0080] 3.1 Library construction and sequencing

[0081] Total RNA from colorectal adenomas and adjacent normal tissues was extracted using Trizol, and RNA integrity was tested using the Agilent 2100 Bioanalyzer, with quality requirements of RIN ≥ 7, 28S / 18S ≥ 1.5:1. The starting Total RNA was accurately quantified using the QUBIT RNA Assay Kit, with a starting amount requirement of 0.1-1ug. Beads with Oligod (T) were used to purify mRNA in Total RNA, and NEB Next Reaction Buffer was used to fragment the mRNA to the target length range, and then the first complementary DNA chain (cDNA) was synthesized using random primers. The second cDNA chain was then synthesized under the action of DNA polymerase I, and the purified double-stranded cDNA was end-repaired, A-tailed, and connected to the sequencing adapter before PCR amplification. The PCR product was quality controlled using the 2100 Bioanalyzer chip to obtain the final library, and the constructed cDNA library was sequenced using the II Lumina Novasek 6000 sequencing platform.

[0082] 3.2 Data filtering and quality control

[0083] Base calling was performed on the raw files of IILumina sequencing and converted into raw sequencing sequences. The collection of raw sequencing sequences was called Raw Data. Raw Data was filtered to delete reads that did not meet the analysis criteria to obtain Clean Reads for subsequent analysis. The filtering conditions were as follows: (1) Remove contaminated samples; (2) If the reads contained adapter sequences, the adapter sequences were truncated first; (3) With a sliding window of 4 bases in length, the bases with an average base quality of <15 at both ends of the reads were truncated; (4) Read pairs with high N content were removed (if the N content in a read was >5%, the entire pair of reads was removed); (5) Read pairs with high content of low-quality bases (bases with a quality value ≤20) were removed (if the low-quality bases in a read were >30%, the entire pair of reads was removed); (6) If the length of the reads after truncation was <100 bp, the entire pair of reads was removed; (7) Unpaired reads were removed.

[0084] 3.3 Sequence alignment and gene quantification

[0085] The index file of the reference genome was downloaded from the official website. Clean Reads were aligned with the reference genome sequence using Hisat2 software. The number of reads aligned to the reference genome in Clean Reads was required to account for ≥70% of the total number of Clean Reads, and the number of reads aligned to multiple positions of the reference genome accounted for ≤10% of the total number of reads aligned to the reference genome. The readcount of gene expression was calculated using featureCounts and StringTie software. The number of fragments per kilobase of exon model per million mapped fragments (FPKM) was calculated taking into account the length and depth of the RNA sequence, and the FPKM value of gene expression was filtered: (1) genes with a length of less than 100 bp were removed; (2) if the FPKM of a gene was < 0.0001, it was considered to be expressed as 0, and genes with an expression of 0 in all samples were removed.

[0086] 3.4 Analysis of differentially expressed genes

[0087] The data after gene expression quantification were statistically analyzed, and the differentially expressed genes (DEGs) between colorectal tissues and adjacent normal tissues were screened out using the limma package and the DESeq package in R language. The screening criteria were: |log2FC|>1 and P<0.05, where FC represents the fold change (FC), and the P value represents the statistical significance of the difference (P-Value). At the same time, the genes expressed in colorectal adenoma tissues but almost not expressed in normal tissues were further screened out based on the FPKM value of gene expression.

[0088] 3.5 Statistical analysis

[0089] IBM SPSS Statistics 26.0 software was used for statistical analysis. The measured data were expressed as mean ± standard deviation (x ± s) when they were in accordance with normal distribution. The independent sample t test was used for comparison between the two groups. Nonparametric test was used for comparison when they were not in accordance with normal distribution. P < 0.05 indicated statistically significant differences.

[0090] 4. Experimental results

[0091] According to the analysis of the whole transcriptome sequencing results, the expression of UGT2B15 in colorectal adenoma tissue was significantly higher than that in normal tissue, and there was a very significant statistical difference (P < 0.001), as shown in Table 2 and Table 3. In the evaluation of its diagnostic efficacy, the ROC curve showed that the area under the curve (AUC) was 0.9514, at which point its sensitivity and specificity could reach 91.67%, as shown in Table 2 and Table 3. Figure 2 , Figure 3 .

[0092] Table 2 Differential expression of UGT2B15 in adenoma tissue and adjacent tissue

[0093]

[0094] Note: CRA is colorectal adenoma tissue, NOR is normal tissue adjacent to the tumor.

[0095] Table 3 Gene expression quantitative values ​​(FPKM) of UGT2B15 in each sample of the test set

[0096] Sample No. FPKM value of adenoma tissue FPKM value of peritumoral tissue 1 3.28283 0 2 1.27065 0 3 2.99115 0 4 0.85987 0 5 1.23642 0 6 0.24646 0 7 2.41035 0 8 1.72367 0.7081 9 2.75271 0.40397 10 1.92791 0 11 1.81842 1.40223 12 1.21561 0

[0097] Example 2 Immunohistochemical verification of the diagnostic performance of UGT2B15 for colorectal adenoma

[0098] 1. Clinical samples

[0099] The immunohistochemistry (IHC) validation set used 45 cases of colorectal adenomas and 45 cases of adjacent normal tissues, for a total of 90 cases (see Table 4). All samples used were formaldehyde-fixed paraffin-embedded adenoma tissues that were removed by colonoscopy and surgery at Wangjing Hospital between 2019 and 2023. The research samples were again confirmed as colorectal adenomas and adjacent normal tissues by 2 senior pathologists (see Figure 4 ).

[0100] Table 4 Clinical information of patients in the IHC validation set

[0102]

[0103]

[0104] 2. Experimental equipment

[0105] 4℃ / -20℃ refrigerator (BCD-196F) (Qingdao Haier Co., Ltd.), RM2235 slicer (LEICA, Germany), HI1210 slice spreader (LEICA, Germany), HI1210 slice baking machine (LEICA, Germany), high pressure steam sterilizer (YX280) (Hefei Huatai Medical Co., Ltd.), electric blower drying oven (101-0AB) (Linmao Technology Beijing Co., Ltd.), three-purpose constant temperature water bath (SH.W21.600) (Shanghai Shuli Instrument Co., Ltd.), BX51 microscope (OLYMPUS, Japan) Co., Ltd.), hematoxylin blue solution (Beijing Shiji Heli Biotechnology Co., Ltd.), DAB color development kit (Beijing Zhongshan Biotechnology Co., Ltd.), hydrogen peroxide solution (Beijing Haiderun Pharmaceutical Group Co., Ltd.), UGT2B15 antibody reagent (A16657) (Wuhan Abotek Biotechnology Co., Ltd.), citric acid repair solution (pH 6.0) (Beijing Zhongshan Jinqiao Biotechnology Co., Ltd.), EDTA antigen repair solution (pH 8.0) (Beijing Zhongshan Jinqiao Biotechnology Co., Ltd.), PBS buffer powder (pH 7.3) (Beijing Zhongshan Jinqiao Biotechnology Co., Ltd.).

[0106] 3. Experimental methods

[0107] The UGT2B15 gene obtained by whole transcriptome sequencing analysis was validated by immunohistochemistry, and the average optical density value of UGT2B15 was calculated using Image ProPlus 6.0 software for semi-quantitative analysis, ROC curve drawing, and diagnostic value evaluation. The specific immunohistochemical validation process is as follows:

[0108] (1) Tissue sections:

[0109] 1) CRA paraffin samples were sectioned at 2.5 μm thickness;

[0110] 2) Expand in a 48°C slide spreader;

[0111] 3) Bake the slices in a 63°C oven for 1 hour.

[0112] (2) Dewaxing and hydration:

[0113] 1) Soak the prepared sections in dewaxing solutions I and II for 15 min respectively;

[0114] 2) Soak in anhydrous ethanol I and II for 5 min respectively;

[0115] 3) Soak in 95% ethanol for 2 minutes and then soak in 80% ethanol for 1 minute;

[0116] 4) Rinse with distilled water for 3 times and soak in PBS for 5 min × 3 times.

[0117] (3) Antigen retrieval:

[0118] 1) Heat EDTA or citric acid repair solution in a pressure cooker until boiling and then put the slices in;

[0119] 2) After the pressure cooker is sprayed, start antigen retrieval;

[0120] 3) EDTA timing 3min, citric acid timing 2.5min;

[0121] 4) Cool the slices at room temperature for 20 minutes.

[0122] (4) Immune response:

[0123] 1) Soak in PBS for 5 min × 3 times and then soak in 3% H2O2 for 10 min;

[0124] 2) Soak again in PBS for 5 min × 3 times;

[0125] 3) Add primary antibody and incubate at room temperature at 37℃ for 2h;

[0126] 4) Soak in PBS for 5 min × 3 times;

[0127] 5) Add secondary antibody and incubate at 37℃ moisturizing box for 30min;

[0128] 6) Soak in PBS for 5 min × 3 times.

[0129] (5)Chemical dyeing:

[0130] 1) DAB staining, microscopic examination of staining degree, positive expression appears yellow-brown;

[0131] 2) Re-stain with hematoxylin for 5 minutes, separate with color-separating solution, and de-blue with anti-blue solution, then wash with water three times.

[0132] (6) Dehydration and sealing:

[0133] 1) Soak in 80% ethanol for 1 min, 95% ethanol for 2 min, and 90% ethanol II for 2 min;

[0134] 2) Soak in anhydrous ethanol I for 5 min and anhydrous ethanol II for 5 min in sequence;

[0135] 3) Soak in transparent liquid I for 1 min, transparent liquid II for 5 min, and transparent liquid III for 5 min in sequence;

[0136] 4) Fix the slide with neutral gum.

[0137] 4. Statistical analysis

[0138] The statistical analysis method is the same as in Example 1.

[0139] 5. Experimental results

[0140] The results of IHC staining of UGT2B15 were analyzed and it was found that the average optical density of UGT2B15 in colorectal adenoma tissue was significantly higher than that in normal tissue adjacent to the tumor, and there was a very significant statistical difference (P < 0.001). In the evaluation of its diagnostic efficacy, the ROC curve showed that the AUC was 0.918, at which point its sensitivity could reach 86.67% and its specificity could reach 91.11%. Figure 5 , Figure 6 , Figure 7 .

[0141] The description of the above embodiments is only used to understand the method and core idea of ​​the present invention. It should be pointed out that, for those skilled in the art, several improvements and modifications can be made to the present invention without departing from the principle of the present invention, and these improvements and modifications will also fall within the scope of protection of the claims of the present invention.

Claims

1. Use of the UGT2B15 biomarker or / and its detection reagent for diagnosing or predicting colorectal adenoma in the preparation of a product for diagnosing or predicting colorectal adenoma.

2. The use according to claim 1, wherein the sample types detected by the detection reagent include tissues, blood, plasma, serum, and blood cells; Preferably, the sample types include tissue, blood, and plasma; Preferably, the sample type is tissue or blood; Preferably, the sample type is tissue; Preferably, the detection reagent is used to detect the content of the biomarker.

3. A kit for diagnosing colorectal adenoma or predicting colorectal adenoma in a subject, the kit comprising a detection reagent for detecting the content of a biomarker, wherein the biomarker comprises UGT2B15.

4. An electronic device for diagnosing colorectal adenoma or predicting colorectal adenoma, the electronic device comprising: An acquisition and detection module is configured to acquire a sample, detect the sample, and obtain the content of the biomarker in the sample; A diagnosis or prediction module, wherein the diagnosis or prediction module uses a machine learning method and the content of the biomarker to establish a regression equation, construct a model, and then outputs a diagnosis or prediction result according to the model; Preferably, the machine learning method includes a linear regression algorithm model, a support vector machine algorithm model, a nearest neighbor / k-nearest neighbor algorithm model, a logistic regression algorithm model, a decision tree algorithm model, a k-means algorithm model, a random forest algorithm model, a naive Bayes algorithm model, a dimensionality reduction algorithm model, and a gradient enhancement algorithm model; Preferably, the sample includes tissue, blood, plasma, serum, blood cells; Preferably, the sample includes tissue, blood, or plasma; Preferably, the sample is tissue or blood; Preferably, the sample is a tissue; Preferably, the biomarker comprises UGT2B15.

5. A method for screening biomarkers for diagnosing or predicting colorectal adenoma, the screening method comprising: Obtaining the colorectal adenoma disease conditions and samples of a sample group, testing the samples to obtain the gene or protein expression levels of the samples, wherein the sample group includes healthy individuals and colorectal adenoma patients; According to the test results of the sample, screening genes or proteins with significant differences between the healthy individuals and the colorectal adenoma patients by statistical methods; Using the significantly different gene or protein contents screened out from the sample population and the colorectal adenoma prevalence in the sample population, a prediction model is constructed through a machine learning training method; The gene or protein that appears more than a threshold number of times in the prediction model is the biomarker; Preferably, the selection of the threshold in the screening method needs to be subjected to random segmentation, the number of random segmentations is multiple, the number of times the prediction model is constructed is multiple, and the prediction model needs to be cross-validated multiple times; Preferably, the statistical method comprises any one or more of a T test, a generalized linear model, a rank test, a logistic regression, a multiple difference, a correction for multiple hypothesis testing, or a Kruskal-Wallis test; Preferably, the sample includes tissue, blood, plasma, serum, blood cells; Preferably, the sample includes tissue, blood, or plasma; Preferably, the sample is tissue or blood; Preferably, the sample is a tissue; Preferably, the process of constructing the prediction model includes dividing the sample population into a training set and a test set, establishing a mapping relationship with the colorectal adenoma prevalence condition, obtaining a training data set or a test data set respectively, and obtaining the prediction model through a machine learning training method and cross-validation.

6. A method for constructing a diagnostic model for colorectal adenoma, the method comprising obtaining a biomarker using the screening method of claim 5 and constructing a diagnostic model using a machine learning training method; Preferably, the biomarker comprises UGT2B15.

7. A processor, wherein the processor is used to run a program, and when the program is run, the screening method according to claim 5, the construction method according to claim 6, or the diagnostic model according to claim 6 is executed.

8. A computer medium, comprising a storage medium, wherein the storage medium is used to store a program, and when the program is run, the device connected to the storage medium is controlled to execute the screening method of claim 5, the construction method of claim 6, or the diagnostic model of claim 6.

9. Application of biomarkers in constructing a system for automated diagnosis or prediction of colorectal adenoma, wherein the biomarkers include UGT2B15.

10. A device for screening biomarkers for diagnosing or predicting colorectal adenoma, the device comprising: An acquisition module is configured to acquire the colorectal adenoma disease conditions and samples of a sample group, detect the samples, and obtain the gene or protein expression levels of the samples, wherein the sample group includes healthy individuals and colorectal adenoma patients; A statistical module is configured to use the test results of the sample to screen genes or proteins that are significantly different between the healthy individual and the colorectal adenoma patient by statistical methods; A model building module is configured to build a prediction model by using the contents of genes or proteins with significant differences screened out from the sample population and the colorectal adenoma morbidity of the sample population through a machine learning training method; A biomarker determination module is configured to identify genes or proteins that appear in the prediction model more than a threshold value as the biomarker; Preferably, the gene or protein expression level detected in the sample includes UGT2B15.

Images