Application of ACOD1 in the severity grading of chronic atrophic gastritis

Abstract

This invention discloses the application of ACOD1 in the severity grading of chronic atrophic gastritis. The assessment method includes classifying patients with chronic atrophic gastritis as mild, moderate, or severe based on the ACOD1 level in samples. This invention establishes an objective and quantifiable grading method for chronic atrophic gastritis by quantitatively detecting the expression level of ACOD1 in gastric tissue. This invention reduces reliance on the subjective experience of endoscopists and multiple biopsies, improves the accuracy, consistency, and reproducibility of grading results, simplifies the procedure, and reduces patient trauma risk. This invention overcomes the shortcomings of existing evaluation methods in terms of accuracy, safety, and clinical applicability, providing a reliable technical means for the early identification, accurate grading, and cancer risk prediction of chronic atrophic gastritis.

Description

Technical Field

[0001] This invention relates to the field of biomedical technology, and in particular to the application of ACOD1 in the severity grading of chronic atrophic gastritis. Background Technology

[0002] Chronic atrophic gastritis (CAG) is a common and refractory digestive system disease characterized by long-term, repeated damage to the gastric mucosal epithelium, significant reduction or atrophy of gastric glands, and thinning of the gastric mucosa, with or without intestinal metaplasia and pseudopyloric gland metaplasia. The clinical manifestations of CAG are mainly upper abdominal distension and pain, belching, acid reflux, and nausea. Some patients experience weight loss, fatigue, and anemia, while others have no obvious discomfort. This disease has an insidious onset, slow progression, long course, and is difficult to cure. It is now recognized as a precancerous condition of the stomach and occupies an intermediate stage in the development of gastric cancer. With the continuous progression of gastric mucosal lesions, the risk of gastric cancer gradually increases. Therefore, early diagnosis and accurate grading of CAG are crucial for preventing gastric cancer and improving patient prognosis.

[0003] Currently, the clinical grading of CAG (mild, moderate, severe) mainly relies on the Kimura-Takemoto classification, OLGA (Operative Link for Gastritis Assessment), and OLGIM (Operative Link on Gastritis Intestinal Metaplasia) scoring systems. The Kimura-Takemoto classification determines the severity based on the endoscopic extent of atrophy. This method heavily depends on the endoscopist's visual observation and experience, resulting in high subjectivity, significant variations in interpretation, and a lack of objective quantitative standards, affecting the consistency and reliability of the grading results. The OLGA and OLGIM scoring systems grade gastric mucosal atrophy and intestinal metaplasia by assessing the distribution and extent of these conditions. They require high-quality biopsy specimens, needing one each from the greater curvature, lesser curvature, angle, greater curvature, and lesser curvature of the gastric body, for a total of five biopsy specimens. This procedure is complex, time-consuming, and demands high levels of skill and adherence to standardized procedures from the endoscopist. Furthermore, multiple biopsies increase the risk of mucosal damage, bleeding, and perforation in patients. Some patients cannot complete standard sampling due to poor tolerance or contraindications to biopsy, further affecting the accuracy and reproducibility of the grading results and limiting the widespread clinical application of the aforementioned scoring system. Existing CAG clinical grading evaluation methods mainly rely on the endoscopist's visual observation and experience, which are easily affected by subjective factors, have poor consistency, and lack objective quantitative standards; or require multiple biopsies, which are cumbersome, high-risk, and have low patient compliance, thus limiting their clinical application.

[0004] Therefore, finding an objective, simple, non-invasive or minimally invasive biomarker for CAG grading assessment is of great significance and application value for the clinical diagnosis and treatment of CAG. Summary of the Invention

[0005] In view of the shortcomings of the prior art described above, the purpose of this invention is to provide the application of ACOD1 in the severity grading of chronic atrophic gastritis.

[0006] The first aspect of this invention protects a method for grading the severity of chronic atrophic gastritis, comprising:

[0007] Based on the ACOD1 level of the samples of patients with chronic atrophic gastritis, patients with chronic atrophic gastritis were classified as mild, moderate, or severe.

[0008] Another aspect of the present invention protects a severity grading assessment system for chronic atrophic gastritis, comprising:

[0009] The judgment unit is used to compare the ACOD1 level of the patient sample with the first threshold T1 and / or the second threshold T2.

[0010] When the ACOD1 level of the sample to be tested is less than or equal to the first threshold T1, it is determined to be mild.

[0011] When the ACOD1 level of the sample to be tested is greater than the first threshold T1 and less than or equal to the second threshold T2, it is determined to be moderate.

[0012] When the ACOD1 level in the sample to be tested is greater than the second threshold T2, it is determined to be severe.

[0013] Another aspect of the present invention protects a device including a processor and a memory, the memory for storing a computer program, and the processor for executing the computer program stored in the memory to cause the device to perform the evaluation method described above.

[0014] Another aspect of the present invention protects a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the evaluation method described above.

[0015] Another aspect of the present invention protects the use of substances that detect ACOD1 in the preparation of products for grading the severity of chronic atrophic gastritis.

[0016] Compared with the prior art, the present invention has the following beneficial effects:

[0017] This invention establishes an objective and quantifiable severity grading system for chronic atrophic gastritis (CAG) by quantitatively detecting the expression level of ACOD1 in gastric tissue. This reduces reliance on the subjective experience of endoscopists and multiple biopsies, improving the accuracy, consistency, and reproducibility of clinical grading results for CAG. Simultaneously, it simplifies the procedure, reduces patient trauma, and overcomes the shortcomings of existing evaluation methods in terms of accuracy, safety, and clinical applicability. This provides a reliable technical means for the early identification, accurate grading, and cancer risk prediction of CAG. Attached Figure Description

[0018] Figure 1A The image shown is an H&E staining image of rat gastric tissue from Example 1 of this application.

[0019] Figure 1B The image shown is a volcano plot of RNA transcription sequencing from rat gastric tissue, as shown in Example 1 of this application.

[0020] Figure 2A The image shown is an ACOD1 immunofluorescence image of gastric tissue from patients with non-chronic atrophic gastritis and chronic atrophic gastritis in Example 2 of this application.

[0021] Figure 2B The diagram shown is a distribution map of ACOD1 in mild, moderate, and severe chronic atrophic gastritis in Example 2 of this application.

[0022] Figure 3A The graph shown is an analysis result of the first ROC curve constructed in Embodiment 3 of this application.

[0023] Figure 3B The graph shown is an analysis result of the second ROC curve constructed in Embodiment 3 of this application.

[0024] Figure 3C The chart shown is a histogram of the diagnostic performance of the training queue in Embodiment 3 of this application.

[0025] Figure 4A The diagram shows the distribution of ACOD1 in mild, moderate, and severe chronic atrophic gastritis in the validation cohort in Example 3 of this application.

[0026] Figure 4B The graph shown is an analysis result of the first ROC curve constructed in Embodiment 3 of this application.

[0027] Figure 4C The graph shown is an analysis result of the second ROC curve constructed in Embodiment 3 of this application.

[0028] Figure 4D The bar chart shown is a breakdown performance histogram of the verification queue in Embodiment 3 of this application. Detailed Implementation

[0029] Aconitate decarboxylase 1 (ACOD1), also known as immune response gene 1, regulates the conversion of aconitic acid (an intermediate in the tricarboxylic acid cycle) to itaconic acid. Current research reports that ACOD1 promotes tumor growth in hepatocellular carcinoma (HCC); it can be used to assess sepsis; and it is a characteristic molecule of the acute inflammatory phase of benign airway stenosis, serving as an inflammatory marker. However, there are currently no reports of ACOD1 being associated with chronic atrophic gastritis.

[0030] The first aspect of this invention protects a method for grading the severity of chronic atrophic gastritis, comprising:

[0031] Based on the ACOD1 level of the samples of patients with chronic atrophic gastritis, patients with chronic atrophic gastritis were classified as mild, moderate, or severe.

[0032] This invention, through transcriptomics analysis, screened a total of 1437 differentially expressed genes in a chronic atrophic gastritis (CAG) model, of which 587 were upregulated and 850 were downregulated. ACOD1 was significantly upregulated in the CAG model and ranked among the top differentially expressed genes, suggesting a high correlation between ACOD1 and the occurrence and development of CAG.

[0033] Based on this, RT-qPCR and immunofluorescence staining were used to detect the mRNA transcription level of ACOD1 and the protein expression level of its encoded protein, respectively. The results showed that in CAG patient samples, both the mRNA level and protein expression level of ACOD1 (expressed as positive area ratio) were significantly positively correlated with disease severity, suggesting that ACOD1 can serve as a reliable quantitative indicator reflecting CAG progression.

[0034] Furthermore, this invention constructs a first threshold T1 and a second threshold T2 for CAG severity grading based on ACOD1. By directly comparing the ACOD1 level in the test sample with the above two thresholds, a three-level quantitative stratification of the severity of the test sample can be achieved. ROC curve analysis results show that the area under the curve (AUC) of the evaluation method of this invention is greater than 0.9, indicating its good discriminative ability. Moreover, at the first threshold T1 and the second threshold T2, the corresponding sensitivity is greater than 0.7 and the specificity is greater than 0.8, respectively, confirming that ACOD1 has high sensitivity and excellent specificity in CAG severity grading.

[0035] In this invention, mild chronic atrophic gastritis refers to a condition where, according to the Kimura-Takemoto classification, the gastric mucosa appears as fine granules under gastroscopy, with some vascular network visible. It is often focal, and the atrophy is limited to the antrum.

[0036] In this invention, moderate chronic atrophic gastritis refers to a condition where, according to the Kimura-Takemoto classification, the gastric mucosa appears moderately granular under gastroscopy, with a clearly visible and continuous vascular network. This condition is often diffuse, with the mucosal folds becoming flattened and shallower, and the atrophy extending from the gastric antrum to the gastric angle.

[0037] In this invention, severe chronic atrophic gastritis refers to a condition where, according to the Kimura-Takemoto classification, the gastric mucosa appears as coarse granules with nodules under gastroscopy, blood vessels extend to the surface, mucosal folds disappear, and atrophy extends to the upper and middle parts of the gastric body.

[0038] In some embodiments, the sample is selected from one or more of gastric tissue, gastric juice, plasma, serum, saliva, urine, and feces. Preferably, it is gastric tissue.

[0039] In some embodiments, the ACOD1 level includes ACOD1 gene transcripts, proteins, or their detection signals. Specifically, it refers to the relative expression level of ACOD1 gene mRNA, the concentration or abundance of ACOD1 protein in the sample, which represents the activity / expression status of the ACOD1 gene in a specific biological sample (such as gastric mucosa tissue). This level can be characterized in the following ways: 1) mRNA level: refers to the content of ACOD1 transcript (mRNA) detected by techniques such as real-time quantitative PCR (RT-qPCR), RNA sequencing, or nucleic acid hybridization. 2) Protein level: refers to the expression level of ACOD1-encoded protein determined by protein detection techniques such as immunohistochemistry (IHC), immunofluorescence (IF), or Western blotting.

[0040] In a specific embodiment of the present invention, the ACOD1 level is characterized by the positive area ratio determined by immunofluorescence staining. Specifically, goat anti-rabbit is labeled with primary antibody DF7723 and secondary antibody Alexa Fluor 594, and cell nuclei are counterstained with DAPI. The cells are then mounted using autofluorescence quenching and antifluorescence quenching mounting media. DAPI (blue light) excitation wavelength is 330-380 nm, emission wavelength is 420 nm; CY3 (red light) excitation wavelength is 510-560 nm, emission wavelength is 590 nm. Images are taken at specific wavelengths using Saiviewer-2.2.2 scanning software. Using AIpathwell v2 image analysis software, the red light positive area ratio (Red Light Positive Area Ratio) is automatically calculated as: total red fluorescent positive area / tissue pixel area.

[0041] In some implementations, the ACOD1 level of the sample to be tested is compared with a first threshold T1 and / or a second threshold T2, wherein the first threshold T1 is less than the second threshold T2.

[0042] When the ACOD1 level of the sample to be tested is less than or equal to the first threshold T1, it is determined to be mild.

[0043] When the ACOD1 level of the sample to be tested is greater than the first threshold T1 and less than or equal to the second threshold T2, it is determined to be moderate.

[0044] When the ACOD1 level of the sample to be tested is greater than the second threshold T2, it is determined to be severe.

[0045] In some implementations, the first threshold T1 is obtained by a method including the following steps:

[0046] Acquire the ACOD1 levels of known mild, moderate, and severe chronic atrophic gastritis patient samples, label the mild group as the first negative category, and combine the moderate and severe groups and label them as the first positive category to obtain the first binary outcome variable Binary1;

[0047] The first ROC curve was constructed using the first binary outcome variable Binary1 as the state variable and the ACOD1 level as the predictor variable.

[0048] The first Youden index is calculated based on the first ROC curve, and the ACOD1 level corresponding to the maximum of the first Youden index is determined as the first threshold T1.

[0049] The second threshold T2 is obtained using a method comprising the following steps:

[0050] The mild and moderate groups were merged and labeled as the second negative category, and the severe group was labeled as the second positive category, resulting in the second binary outcome variable Binary2;

[0051] A second ROC curve was constructed using the second binary outcome variable Binary2 as the state variable and the ACOD1 level as the predictor variable.

[0052] The second Yoden index is calculated based on the second ROC curve, and the ACOD1 level corresponding to the maximum second Yoden index is determined as the second threshold T2.

[0053] In some implementations, the conversion of the three-category grouping variables into binary outcome variables is achieved using a conditional judgment function, which assigns a corresponding binary classification code based on the value of the grouping variable.

[0054] In some implementations, the conditional decision function is the ifelse() function, used to convert the ternary grouping variable into a binary outcome variable Binary1 or Binary2. This functionality is not limited to R language implementation; any statistical software, programming language, or custom algorithm capable of conditional decision-making and classification assignment can achieve the same function, including but not limited to conditional classification functions or logical discriminant algorithms in software platforms such as Python, SPSS, SAS, MATLAB, and Stata.

[0055] In some implementations, the conditional function is selected from the ifelse() function or the np.where() function.

[0056] The conditional statement takes the ifelse() function as an example. If the "mild group" is encoded as 0 and the "moderate and severe groups" are encoded as 1, the binary outcome variable Binary1 can be generated by ifelse(group == "mild", 0, 1).

[0057] The conditional statement described uses the ifelse() function as an example. If "mild and moderate groups" are encoded as 0, and "severe group" is encoded as 1, then the binary outcome variable Binary2 can be generated using ifelse(group == "severe", 1,0,). The same functionality can be achieved in Python using the np.where() function. Those skilled in the art will understand that any programming language or statistical software with conditional statement capabilities can implement the above conversion, including but not limited to conditional statements or logical assignment statements in platforms such as R, Python, SPSS, SAS, MATLAB, and Stata.

[0058] In some implementations, the first and second ROC curves are constructed by using the binary outcome variables Binary1 and Binary2 as state variables and ACOD1 level as a predictor variable, respectively. This is achieved by iterating through each candidate classification threshold and calculating the corresponding sensitivity and specificity, plotting 1-specificity on the x-axis and sensitivity on the y-axis. These curves are used to evaluate the discriminative ability of the predictor variables for different disease groups. For example, the `roc()` function can be used to construct the first ROC curve with Binary1 as the state variable and ACOD1 level as the predictor variable; the `roc()` function can be used to construct the second ROC curve with Binary2 as the state variable and ACOD1 level as the predictor variable.

[0059] In some implementations, the first threshold T1 and the second threshold T2 are extracted using the Youden exponent maximization principle. For example, the coords() function can be used with the parameter method="best" to determine the optimal cutoff value based on the Youden exponent maximization principle, and simultaneously obtain the sensitivity and specificity corresponding to that cutoff point.

[0060] In some more specific embodiments, the first threshold T1 is 0.24–0.26, or it can be 0.24, 0.25, or 0.26; the second threshold T2 is 0.27–0.28, or it can be 0.27 or 0.28.

[0061] The ROC curve (Receiver Operating Characteristic Curve), also known as the receiver operating characteristic curve, was first used to distinguish signals from noise in radar signal detection. Later, it was widely used to evaluate the effectiveness of the latest medical diagnostic methods in practical applications. Furthermore, in statistical modeling, the ROC curve is frequently used to evaluate the accuracy of predictive models. In principle, the ROC curve is typically based on a binary classification (1 / 0) approach, pairwise mixing of all predicted and actual results to derive four cases: TP, FP, FN, and TN.

[0062] TP represents the number of true positive samples, FP represents the number of false positive samples, FN represents the number of false negative samples, and TN represents the number of true negative samples. Based on this, the true positive rate (TPR), true negative rate (TNR), false positive rate (FPR), and false negative rate (FNR) of the model classification can be calculated. TPR is also known as sensitivity, and TNR is also known as specificity. The calculation formulas are as follows:

[0063] TPR = TP / (TP + FN)

[0064] TN = TN / (TN + FP)

[0065] FPR = FP / (TN + FP)

[0066] FNR = FN / (TN + FP)

[0067] The Youden Index, also known as the correctness index, is the sum of sensitivity and specificity minus 1 (i.e., the difference between the true positive rate and the false positive rate). The formula is as follows:

[0068] Youden Index = Sensitivity + Specificity – 1 = TPR + TNR – 1 = TPR - FPR

[0069] The Youden index, ranging from 0 to 1, represents the ability to distinguish between positive and negative results at the current threshold. A higher Youden index indicates a higher probability of true positives and a lower probability of false positives, signifying better classification performance at that threshold.

[0070] Another aspect of the present invention protects a severity grading assessment system for chronic atrophic gastritis, comprising:

[0071] The judgment unit is used to compare the ACOD1 level of the patient sample with the first threshold T1 and / or the second threshold T2.

[0072] When the ACOD1 level of the sample to be tested is less than or equal to the first threshold T1, it is determined to be mild.

[0073] When the ACOD1 level of the sample to be tested is greater than the first threshold T1 and less than or equal to the second threshold T2, it is determined to be moderate.

[0074] When the ACOD1 level in the sample to be tested is greater than the second threshold T2, it is determined to be severe.

[0075] In some embodiments, the sample is selected from one or more of gastric tissue, gastric juice, plasma, serum, saliva, urine, and feces.

[0076] In some implementations, the ACOD1 level refers to the ACOD1 gene transcript, protein, or its detection signal in the sample.

[0077] In some implementations, the ACOD1 level is obtained by detecting the sample through immunological detection, nucleic acid detection, or mass spectrometry.

[0078] In some embodiments, the system further includes a data unit for obtaining ACOD1 level data of the patient sample with chronic atrophic gastritis to be tested.

[0079] In some embodiments, the system further includes an analysis unit configured to:

[0080] Acquire the ACOD1 levels of known mild, moderate, and severe chronic atrophic gastritis patient samples, label the mild group as the first negative category, and combine the moderate and severe groups and label them as the first positive category to obtain the first binary outcome variable Binary1;

[0081] Using the first binary outcome variable Binary1 as the state variable and the ACOD1 level as the predictor variable, a first ROC curve is constructed, and the ACOD1 level corresponding to the maximum Youden index on the first ROC curve is determined as the first threshold T1.

[0082] The mild and moderate groups were merged and labeled as the second negative category, and the severe group was labeled as the second positive category, resulting in the second binary outcome variable Binary2;

[0083] Using the second binary outcome variable Binary2 as the state variable and the ACOD1 level as the predictor variable, a second ROC curve is constructed, and the ACOD1 level corresponding to the maximum Youden index on the second ROC curve is determined as the second threshold T2.

[0084] In some implementations, the known mild, moderate, and severe chronic atrophic gastritis patient samples refer to patients clinically diagnosed with chronic atrophic gastritis according to the Kimura-Takemoto classification.

[0085] In some implementations, the first threshold T1 is 0.24–0.26, or it can be 0.24, 0.25, or 0.26; the second threshold T2 is 0.27–0.28, or it can be 0.27 or 0.28.

[0086] Another aspect of the present invention protects a device including a processor and a memory, the memory for storing a computer program, and the processor for executing the computer program stored in the memory to cause the device to perform the evaluation method described above.

[0087] Another aspect of the present invention protects a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the evaluation method described above.

[0088] In some implementations, the electronic terminal performs the following steps:

[0089] Based on the ACOD1 level of the samples of patients with chronic atrophic gastritis, patients with chronic atrophic gastritis were classified as mild, moderate, or severe.

[0090] In some implementations, the electronic terminal performs the following steps:

[0091] 1) ACOD1 levels in samples from patients with chronic atrophic gastritis to be tested;

[0092] 2) Based on the level in 1), patients with chronic atrophic gastritis are classified as mild, moderate or severe.

[0093] In the above embodiments, the functions of each functional unit can be implemented entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. A computer program product includes one or more computer instructions (programs). When the computer program instructions (programs) are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. A computer-readable storage medium can be any available medium that a computer can access, or a data storage device such as a server or data center that integrates one or more available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., digital video discs (DVDs)), or semiconductor media (e.g., solid-state disks (SSDs)).

[0094] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0095] This invention, through transcriptomic analysis, screened 1437 differentially expressed genes in a chronic atrophic gastritis (CAG) model, including 587 upregulated genes and 850 downregulated genes. ACOD1 was significantly upregulated in the CAG model and ranked among the top differentially expressed genes, suggesting a high correlation between ACOD1 and the occurrence and development of CAG.

[0096] To further verify the association between ACOD1 and the severity of chronic atrophic gliosis (CAG), this invention used immunofluorescence technology to detect the expression level of ACOD1, using the red-positive area ratio as a quantitative indicator. The experimental results showed that the red-positive area ratio increased progressively from the mild to the moderate to severe CAG groups, indicating that the ACOD1 protein expression level increased with increasing CAG severity, suggesting that ACOD1 can serve as a potential biomarker reflecting the severity of CAG.

[0097] Based on the above findings, this invention further constructs a method for assessing the severity of chronic atrophic gastritis (CAG) based on ACOD1. This assessment method employs a dual-threshold discrimination strategy: the first threshold (T1) is determined by performing a first ROC curve analysis on "mild" and "moderate-severe" (i.e., moderate + severe), used to distinguish between mild and moderate CAG; the second threshold (T2) is determined by performing a second ROC curve analysis on "mild-moderate" (i.e., mild + moderate) and "severe," used to distinguish between severe and mild-moderate CAG.

[0098] Through analysis of the training and validation sets, this invention establishes a stable dual-threshold grading standard: T1 = 0.24–0.26 (preferably 0.258) and T2 = 0.27–0.28 (preferably 0.286). By comparing the ACOD1 level (red light positivity area ratio) of the sample to be tested with the above two thresholds, the grading of mild, moderate, and severe CAG can be achieved.

[0099] This invention achieves three-level quantitative stratification of CAG severity using a single biomarker, reducing the reliance on the subjective experience of endoscopists in traditional grading methods and minimizing the trauma risks associated with multiple biopsies. It provides an objective, quantifiable, and clinically valuable technical means for the accurate grading of CAG.

[0100] Another aspect of the present invention protects the use of substances that detect ACOD1 in the preparation of products for grading the severity of chronic atrophic gastritis.

[0101] In some embodiments, the substance used to detect ACOD1 includes a substance used to detect the level of ACOD1 in the sample.

[0102] In some embodiments, the sample is selected from one or more of gastric tissue, gastric juice, plasma, serum, saliva, urine, and feces.

[0103] In some implementations, the ACOD1 level is detected in the sample by immunological detection, nucleic acid detection, or mass spectrometry.

[0104] In some embodiments, the ACOD1 level includes the ACOD1 gene, transcript, protein, or detection signal thereof.

[0105] In some embodiments, the product comprises a reagent kit, a protein chip, or a detection reagent.

[0106] In some implementations, the severity of the chronic atrophic gastritis is categorized as mild, moderate, and severe.

[0107] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention.

[0108] Before further describing specific embodiments of the present invention, it should be understood that the scope of protection of the present invention is not limited to the specific embodiments described below; it should also be understood that the terminology used in the embodiments of the present invention is for describing specific embodiments and not for limiting the scope of protection of the present invention; in the specification and claims of the present invention, unless otherwise expressly stated in the text, the singular forms "a", "an" and "this" include the plural forms.

[0109] When numerical ranges are given in the embodiments, it should be understood that, unless otherwise stated in the present invention, both endpoints of each numerical range and any value between the two endpoints may be selected. Unless otherwise defined, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art. In addition to the specific methods, apparatus, and materials used in the embodiments, based on the knowledge of the prior art possessed by one of ordinary skill in the art and the description of this invention, any prior art methods, apparatus, and materials similar to or equivalent to those described, apparatus, and materials in the embodiments of this invention may be used to implement the present invention.

[0110] In the following examples, N-methyl-N'-nitro-N-nitrosoguanidine (abbreviated as MNNG) was purchased from Shanghai Haoyuan Pharmaceutical Technology Co., Ltd., and was prepared into a solution of 170 μg / mL with pure water.

[0111] The rats were purchased from Shanghai Slack Laboratory Animal Co., Ltd., and all were male.

[0112] Example 1: Screening markers

[0113] 1.1 CAG Animal Experiments

[0114] Male SD rats (weighing 200±20g) were randomly divided into two groups after 1 week of acclimatization: a blank control group (n=10) and an MNNG model group (n=10).

[0115] Blank control group: Normal saline was administered by gavage every other day, with free access to food and water for 10 consecutive weeks;

[0116] MNNG model group: MNNG (170 μg / mL) was administered by gavage every other day, followed by fasting every other day, while MNNG (170 μg / mL) was given free access to water daily for 10 consecutive weeks.

[0117] Ten weeks later, rats in the blank control group and the MNNG model group were euthanized, and gastric tissue was taken for gastric mucosal pathological morphology examination.

[0118] Pathological morphological examination of the gastric mucosa was performed by H&E staining, and the results are shown in the figure. Figure 1A The left side shows the model before molding, and the right side shows the model after molding.

[0119] from Figure 1A The fact that a significant amount of mucosal layer structure detached indicates that the modeling was successful.

[0120] 1.2 Transcriptome Sequencing

[0121] The specific steps are as follows: Total RNA quality detection, mRNA capture, mRNA fragmentation and double-stranded cDNA synthesis, end repair with A and adapter addition, fragment sorting, PCR amplification, library quality control, and Illumina sequencing.

[0122] 1 μg of total RNA was used for subsequent library preparation. For mRNA enrichment, poly(A) mRNA was isolated using Oligo(dT) magnetic beads. mRNA fragmentation was performed using divalent cations and high temperature. Primer annealing was performed using random primers. First-strand cDNA was synthesized using reverse transcriptase, followed by second-strand cDNA synthesis. For strand-specific library construction, dUTP was used instead of standard dTTP during second-strand synthesis; subsequently, the second strand was selectively degraded using USER enzyme before PCR amplification to preserve strand information. The purified double-stranded cDNA underwent end repair and dA tailing was added in a single reaction, followed by adapter ligation at both ends. Fragment size screening of the adapter-ligated DNA was performed using DNA purification magnetic beads. PCR amplification was then performed on each sample using P5 and P7 primers, and the PCR products were quality-verified. Finally, the libraries with different indices were mixed and sequenced using a 2×150 bp paired-end sequencing configuration on an Illumina HiSeq / Illumina Novaseq / MGI2000 platform according to the manufacturer's instructions. The specific work was entrusted to Genewiz.

[0123] 1.3 Results

[0124] Transcriptome sequencing results are shown below Figure 1B .

[0125] from Figure 1B It can be seen that the gene expression profiles of the CAG model group and the blank control group are significantly different. Compared with the blank control group, the CAG model group upregulated 587 genes and downregulated 850 genes. Among them, ACOD1 was significantly upregulated in the CAG model and ranked among the differentially expressed genes.

[0126] Example 2

[0127] Based on Example 1, gastric tissue samples from non-CAG patients (28 cases) and CAG patients (83 cases) were analyzed through relevant molecular biology and pharmacology experiments to clarify the expression characteristics of ACOD1 in gastric tissue of CAG patients and reveal the correlation between its expression level and CAG clinical grading.

[0128] 2.1 Endoscopic biopsy (taking the antrum of the stomach as an example)

[0129] ① Ensure the patient has fasted for at least 6 hours and abstained from water for at least 2 hours before the procedure to rule out gastric food retention and ensure a clear surgical field. Verify the patient's basic information and medical history, paying particular attention to any coagulation disorders, history of anticoagulant / antiplatelet drug use, and whether the patient has signed informed consent forms for gastroscopy and biopsy. Inform the patient of the purpose of the biopsy, possible discomfort, and risks such as bleeding and infection. Administer local anesthesia to the pharynx (oral anesthetic) before the procedure; reassure the patient and instruct them to maintain a left lateral decubitus position with a mouth pad placed and secured.

[0130] ② Debug the electronic gastroscopy equipment to ensure that the endoscopic image is clear, the air and water injection and suction functions are normal; prepare disposable sterile biopsy forceps and check that the forceps head closure and extension functions are intact.

[0131] ③ The endoscope is slowly inserted through the patient's mouth pad, passing sequentially through the oral cavity, pharynx, esophagus, and cardia, into the stomach cavity. Appropriate air is injected to expand the stomach cavity, and gastric mucus and foam are aspirated. The fundus and body of the stomach are observed progressively until the antrum is reached. The antral mucosa is clearly exposed, with particular attention paid to the condition of the greater curvature, lesser curvature, anterior wall, and posterior wall mucosa. Atrophic lesions are confirmed to pinpoint the precise biopsy target. The endoscope angle is adjusted to keep the biopsy target centered in the field of view, maintaining the endoscope perpendicular to the antral mucosa and avoiding obstruction by mucosal folds to ensure a stable operating field.

[0132] ④ Under direct endoscopic visualization, slowly insert the biopsy forceps through the biopsy channel of the endoscope until the forceps tip extends beyond the tip of the endoscope and approaches the target gastric antrum mucosa for biopsy, avoiding contact with surrounding normal mucosa to prevent damage. Fine-tune the position of the biopsy forceps so that the forceps tip is vertically aligned with the target gastric antrum mucosa, gently press down on the forceps tip to conform to the mucosal surface, quickly close the forceps tip, and bite off an appropriate amount of gastric antrum mucosal tissue (avoid excessive force to prevent mucosal tearing, and also prevent the tissue from being bitten too superficially, affecting pathological examination). Keeping the forceps tip closed, slowly retract the biopsy forceps along the endoscopic biopsy channel and remove it from the endoscope, avoiding shaking during the process to prevent the bitten tissue from falling off and being lost. Gently open the forceps tip and place the removed gastric antrum mucosal tissue intact into the pre-prepared sample tube.

[0133] Referencing the "Guidelines for the Diagnosis and Treatment of Chronic Gastritis in China" (Chinese Society of Gastroenterology, 2022, Shanghai) and the "Consensus Opinion on Chronic Gastritis in China" (Chinese Society of Gastroenterology, 2017), the grading criteria for gastric mucosal manifestations under gastroscopy are shown in Table 1 below.

[0134] Table 1

[0135] normal No obvious atrophic lesions in the gastric mucosa Mild The gastric mucosa appears granular with partially visible vascular networks, often focal, and the atrophy is limited to the gastric antrum. moderate The gastric mucosa is moderately granular with a clearly visible and continuous vascular network, often diffuse. The mucosal folds are flattened and shallowed, and the atrophy extends from the gastric antrum to the gastric angle. Severe The gastric mucosa appears coarse and granular with nodules; blood vessels extend to the surface; mucosal folds have disappeared; and the atrophy extends to the upper and middle parts of the gastric body.

[0136] Clinically, it is classified into open and closed types according to the Kimura-Takemoto classification. The open type includes O1-O3, and the closed type includes C1-C3. Clinically, C1-C2 is considered mild, C3-O1 is moderate, and O2-O3 is severe.

[0137] 2.2 Paraffin embedding for pathological sample preparation

[0138] ① Sample processing: Fix the gastric antral mucosal tissue sample from step 2.1 with fixative (Wuhan Saiweier Biotechnology Co., Ltd.) for at least 24 hours. Remove the sample from the fixative and trim the tissue at the target site using a scalpel in a fume hood. Place the trimmed tissue into the corresponding labeled dehydration box.

[0139] ② Dehydration and wax impregnation: Place the dehydration box into the dehydrator and perform dehydration with alcohol in a gradient: 75% alcohol 1.5h -- 85% alcohol 1.5h -- 90% alcohol 1.5h -- 95% alcohol 1-1.5h -- anhydrous ethanol I 1-1.5h -- anhydrous ethanol II 1-1.5h -- benzene 5-10min -- xylene I 10-20min -- xylene II 10-20min -- 65℃ melt paraffin I 0.5h -- 65℃ melt paraffin II 1h -- 65℃ melt paraffin III 1.5h.

[0140] ③ Embedding: Samples impregnated with paraffin are transferred to an embedding machine for embedding. First, molten paraffin is placed into the embedding frame. Before the paraffin solidifies, the tissue is removed from the dehydration box, placed into the embedding frame according to the embedding surface requirements, and the corresponding label is attached. The sample is cooled on a -20°C freezing stage. After the paraffin solidifies, the paraffin block is removed from the embedding frame and trimmed.

[0141] 2.3 Immunofluorescence

[0142] ① Paraffin sections to water: Place the paraffin sections from step 2.2 into environmentally friendly dewaxing solution (Wuhan Saiweier Biotechnology Co., Ltd.) for 10 min - environmentally friendly dewaxing solution for 10 min - environmentally friendly dewaxing solution for 10 min - anhydrous ethanol for 5 min - anhydrous ethanol for 5 min - anhydrous ethanol for 5 min - and then wash with distilled water.

[0143] ② Antigen retrieval: Retrieve according to the retrieval conditions. During the retrieval process, prevent excessive evaporation of the buffer solution and avoid drying the slide. After retrieval, allow it to cool naturally. Place the slide in PBS and wash it three times on a destaining shaker for 5 minutes each time.

[0144] ③ Circle blocking with serum: After slightly drying the section, draw a circle around the tissue with a histochemical pen, add BSA, and block for 30 minutes.

[0145] ④ Add primary antibody: Add primary antibody ACOD1 (DF7723, Jiangsu Qinke Biotechnology Research Center Co., Ltd.), place the slice flat in a humidified box and incubate overnight at 4°C.

[0146] ⑤ Add secondary antibody: Place the slide in PBS and wash three times on a decolorizing shaker for 5 minutes each time. Add the corresponding secondary antibody Alexa Fluor 594-labeled goat anti-rabbit (G828301, Wuhan Saiweier Biotechnology Co., Ltd.) and incubate at room temperature in the dark for 50 minutes.

[0147] ⑥ Counterstaining cell nuclei with DAPI: Place the slide in PBS. Wash three times on a destaining shaker, 5 min each time. Add DAPI staining solution and incubate at room temperature in the dark for 10 min.

[0148] ⑦ Autofluorescence quenching: After slightly drying the sections, place them flat in a light-proof humidified chamber. Add the autofluorescence quencher to the circle for 5 minutes, then rinse with running water for 10 minutes. The autofluorescence quencher is 0.1-0.3% Sudan Black B dissolved in 70% ethanol.

[0149] ⑧ Mounting: Mounting with anti-fluorescence quenching mounting medium. Anti-fluorescence quenching mounting medium (glycerol-PVA (polyvinyl alcohol)).

[0150] ⑨ Image acquisition: DAPI (blue light) excitation wavelength 330-380nm, emission wavelength 420nm; CY3 (red light) excitation wavelength 510-560nm, emission wavelength 590nm. Use Saiviewer-2.2.2 scanning software to take pictures at the specific wavelengths.

[0151] Using AIpathwell v2 image analysis software, the following can be automatically calculated in the image:

[0152] Red Light Positive Area Ratio = Total positive area of ​​red fluorescence / Area of ​​tissue pixels.

[0153] Immunofluorescence images of ACOD1 in gastric tissues from non-CAG and CAG patients are shown below. Figure 2A .

[0154] from Figure 2A It can be seen that the immunofluorescence of gastric tissue from CAG patients is different from that of non-CAG patients.

[0155] The Kimura-Takemoto classification method divided 83 CAG patients into 46 mild cases, 27 moderate cases, and 10 severe cases.

[0156] To assess the variation characteristics of ACOD1 expression levels in CAG grading, a combination of descriptive statistics and visualization analysis was employed. Boxplots were plotted using the `geom_boxplot` function from the `ggplot2` package to display the median, quartile range, and outliers for each group. Simultaneously, a scatter plot was overlaid using the `geom_jitter` function to illustrate the actual values ​​and discrete distribution of each sample. The distribution of the red-light positive area ratio in the grading of chronic atrophic gastritis was analyzed using boxplots overlaid with scatter plots. Figure 2B In the figure, the central horizontal line represents the median of each group of samples, reflecting the central tendency of ACOD1 expression levels in that group; the upper and lower boundaries of the box represent the first quartile (Q1) and the third quartile (Q3), respectively, and the interval (Q1–Q3) represents the range of major data distribution within the group; the upper and lower bands represent the range of non-outlier values, and the scatter plots represent individual sample observations.

[0157] from Figure 2BIt can be seen that the red light positivity area ratio increases progressively across different disease severity groups. Specifically, the median is lowest in the mild group, followed by the moderate group, and highest in the severe group (i.e., the median increases with severity, mild < moderate < severe). The interquartile range (IQR) of each group generally shifts upward with disease severity, meaning the moderate group is generally higher than the mild group, and the severe group is generally higher than the moderate group. Furthermore, the scatter plot distribution of the samples in each group also shows an overall trend towards higher values.

[0158] The above results suggest that the red light positive area ratio gradually increases with the severity of chronic atrophic gastritis (CAG) (in a positive correlation), indicating that there is a consistent increasing trend between ACOD1 expression level and disease severity, thus providing a basis for the construction of subsequent discriminant models.

[0159] Through the above visualization analysis, the differences in the red light positive area ratio among the mild, moderate and severe groups can be observed intuitively, and its increasing trend with the severity of CAG can be assessed, providing a basis for the construction of subsequent discrimination models.

[0160] Example 3: Severity Grading Assessment Method for Chronic Atrophic Gastritis

[0161] 3.1 Evaluation Methods

[0162] 3.1.1 Data Sources and Preprocessing

[0163] The original data came from the test results of 252 clinical samples of chronic atrophic gastritis (including 10 severe, 10 moderate, and 10 mild cases from Example 2). Chronic atrophic gastritis (CAG) was divided into three groups according to the Kimura-Takemoto classification: mild (C1-C2), moderate (C3-O1), and severe (O2-O3). Each sample included the red light-positive area ratio as a quantitative indicator.

[0164] First, 74 samples (38 mild, 18 moderate, and 18 severe; including 10 severe, 10 moderate, and 10 mild cases from Example 2) were obtained as a training queue to generate three-level classification rules and determine the first and second thresholds.

[0165] Subsequently, 178 independent clinical samples (62 mild, 54 moderate, and 62 severe) were added as a validation cohort to independently validate the generated three-level grading rules, in order to further evaluate the distribution characteristics of ACOD1 expression levels in different grades of chronic atrophic gastritis severity and its diagnostic ability.

[0166] The validation cohort was analyzed using the same data processing and ROC analysis workflow as the training cohort to ensure methodological consistency and comparability of results. All statistical analyses and graphical representations were performed in the R software (4.5.3) environment.

[0167] First, three sets of data are read separately, and the data structure is processed uniformly, including:

[0168] ① The ratio of red light positive area was extracted as a continuous predictor variable for subsequent group comparisons and ROC analysis construction;

[0169] ② The percentage of red light positive area (expressed as a percentage) is converted by dividing the percentage value by 100 and converting it into a decimal form in the range of 0 to 1 (inclusive). For example, 50% is converted to 0.50 to unify the data representation and ensure the consistency of subsequent statistical analysis.

[0170] ③ Add group labels to each sample, namely Mild, Moderate, and Severe;

[0171] ④ Merge the three sets of data to construct a complete dataset, and set the grouping variable as an ordered categorical variable (factor). The grouping order is defined as Mild < Moderate < Severe, which is used to reflect the progressive relationship of disease severity and the order of disease progression.

[0172] The input data matrix (phenotype matrix, pheno) consists of three parts: sample ID, red light positive area ratio (a continuous variable), and group information. In the ROC analysis, the group information is transformed into a binary outcome variable (0 / 1) to distinguish different disease states. In the first ROC curve (Mild vs. Moderate + Severe), the mild group is defined as 0 (normal), and the moderate and severe groups are combined and defined as 1 (disease). In the second ROC curve (Mild + Moderate vs. Severe), the mild and moderate groups are combined and defined as 0 (normal), and the severe group is defined as 1 (disease). Therefore, the input matrix for the ROC analysis consists of sample ID, continuous predictor variables, and the corresponding group information (0 / 1).

[0173] Specifically, the operation command is:

[0174] First, import the sample data from the mild, moderate, and severe groups into the R software environment (version 4.5.3) and construct the training dataset:

[0175] qd <- read.csv("Mild (C1-C2).csv", header = TRUE, row.names = 1)

[0176] zd <- read.csv("Moderate(C3-O1).csv", header = TRUE, row.names = 1)

[0177] zzd <- read.csv("Severe (O2-O3).csv", header = TRUE, row.names = 1)

[0178] We obtained the following respectively:

[0179] qd = mild group data

[0180] zd = moderate group data

[0181] zzd = severe group data

[0182] qd$Group <- "Mild"

[0183] zd$Group <- "Moderate"

[0184] zzd$Group <- "Severe"

[0185] data_all <- rbind(qd, zd, zzd)

[0186] data_all$Group <- factor(data_all$Group, levels = c("Mild", "Moderate", "Severe"))

[0187] Where qd, zd, and zzd represent the sample data of the mild, moderate, and severe groups, respectively;

[0188] data_all is the merged training data matrix, which contains the red light positive area ratio and corresponding grouping information;

[0189] read.csv(): Reads a comma-separated value (CSV) file, used for importing external data;

[0190] header=TRUE: This indicates that the first row contains variable names (column names).

[0191] row.names=1: Set the first column as the row name (row names, sample number);

[0192] $: Used to extract or add predictor variables (column selection / assignment) in a data frame;

[0193] Group: categorical variable;

[0194] Mild / Moderate / Severe: Disease severity labels used to indicate the disease level to which each sample belongs;

[0195] rbind(): row bind(row-to-row dataframe concatenation) combines multiple dataframes vertically to obtain a combined dataset.

[0196] factor(): Converts the predictor variable into a factor (categorical variable);

[0197] levels: Specifies the order of categories;

[0198] Mild → Moderate → Severe: This indicates a progressive relationship of disease severity, and its function is to set the grouping variable as an ordered categorical variable.

[0199] 3.1.2 Constructing the first ROC curve analysis (mild vs. moderate + severe, i.e., Mild vs. Moderate + Severe)

[0200] To determine the first threshold: a training cohort of 74 cases (38 mild, 18 moderate, and 18 severe) was used, and univariate receiver operating characteristic (ROC) curve analysis was performed based on the red-light positive area ratio. The first ROC curve was used to evaluate the diagnostic efficacy of the red-light positive area ratio in distinguishing between mild and moderate-to-severe chronic atrophic gastritis.

[0201] First, the three-class classification grouping variable is converted into a two-class outcome variable Binary1 using the ifelse() function. The mild group (Mild) is defined as 0, and the moderate and severe groups (Moderate and Severe) are combined and defined as 1, thus transforming the three-class classification problem into a two-class classification problem of "mild vs. moderate and severe".

[0202] Subsequently, the roc() function was used to construct the first ROC curve with the binary outcome variable Binary1 as the true state variable and the red light positive area ratio as the predictor variable, in order to evaluate the ability of ACOD1 to distinguish between the mild and moderate-to-severe groups.

[0203] Next, the area under the ROC curve (AUC) is calculated using the auc() function to quantify the overall discriminative performance. The closer the AUC value is to 1, the stronger the discriminative ability. The AUC area is usually around 0.5 to 1. The larger the value, the higher the classification accuracy of the classification model. The accuracy of the classification model is judged according to the following criteria: AUC = 1, perfect classification; 0.85 ≤ AUC ≤ 0.95, very good classification effect; 0.7 ≤ AUC < 0.85, average classification effect; 0.5 < AUC < 0.7, low classification effect; AUC = 0.5, the same as random guessing (e.g., flipping a coin), the model has no predictive value; AUC < 0.5, worse than random guessing, but as long as it always does the opposite of predicting, it is better than random guessing.

[0204] Finally, using the `coords()` function (parameter `method="best"`), the optimal cutoff value (threshold, also known as the first threshold T1) is determined based on the Youden index principle. Simultaneously, the sensitivity and specificity corresponding to this cutoff value are obtained to determine the optimal cutoff value and its diagnostic efficacy index. The Youden index is defined as: Youden Index = Sensitivity + Specificity − 1; where sensitivity represents the ability to correctly identify moderate to severe chronic atrophic gastritis; and specificity represents the ability to correctly identify mild chronic atrophic gastritis. The optimal cutoff value is the predictor variable value that maximizes the Youden index.

[0205] Specifically, the operation command is:

[0206] library(pROC)

[0207] data_all$Binary1 <- ifelse(data_all$Group == "Mild", 0, 1)

[0208] roc1 <- roc(data_all$Binary1, data_all$Red.light.positive.area.ratio)

[0209] auc1 <- auc(roc1)

[0210] coords1 <- coords(roc1, "best", ret = c("threshold", "sensitivity", "specificity"))

[0211] Binary1 is the first binary outcome variable, defined as 0 for the mild group and 1 for the combined moderate and severe groups;

[0212] ifelse(): a conditional function (if-else vectorized function, a vectorized conditional statement);

[0213] data_all$Group == "Mild": Determines whether the group is Mild, transforming the three-class classification problem into a two-class ROC problem;

[0214] roc1 is the first ROC curve;

[0215] `roc()`: Constructs the ROC curve function (Receiver Operating Characteristic curve function); the first parameter is the true label (0 / 1); the second parameter is the predictor variable (continuous value); it calculates sensitivity and specificity at different thresholds.

[0216] auc1 is the area under the first ROC curve;

[0217] auc(): Calculates the area under the curve, used to evaluate whether the ratio of red-positive area can distinguish Mild vs others;

[0218] coords1 is used to extract the best cutoff value, sensitivity, and specificity; it extracts points (coordinates) on the ROC curve; "best" selects the best cutoff value, usually based on: Youden index (sensitivity + specificity - 1 maximum) ret=c (...) indicates that three metrics are returned.

[0219] 3.1.3 Constructing a second ROC curve for analysis (mild + moderate vs. severe, i.e., Mild + Moderate vs. Severe)

[0220] A sample of 74 cases (38 mild, 18 moderate, and 18 severe, same as step 3.1.2) was included in the training cohort, and univariate receiver operating characteristic (ROC) curves based on the red light positivity area ratio were used for analysis. A second ROC curve was used to evaluate the diagnostic efficacy of the red light positivity area ratio in differentiating between severe and mild-to-moderate chronic atrophic gastritis.

[0221] First, the three-class grouping variable is converted into a two-class outcome variable Binary2 using the ifelse() function, where the severe group (Severe) is defined as 1, and the mild group (Mild) and the moderate group (Moderate) are combined and defined as 0, thus transforming the three-class classification problem into a two-class classification problem.

[0222] Subsequently, the roc() function was used to construct a second ROC curve with the binary outcome variable Binary2 as the true state variable and the red light positive area ratio as the predictor variable, to evaluate the ability of ACOD1 to distinguish between the severe group and the mild to moderate group.

[0223] Next, the area under the curve (AUC) is calculated using the auc() function to quantify the overall discrimination performance; the closer the AUC value is to 1, the stronger the discrimination ability.

[0224] Finally, using the `coords()` function, the optimal cutoff value (threshold, also known as the second threshold T2) is determined based on the Youden index principle. Simultaneously, the sensitivity and specificity corresponding to this cutoff value are obtained to determine the maximum optimal cutoff value and its diagnostic efficacy index. The Youden index is defined as: Youden Index = Sensitivity + Specificity − 1; where sensitivity represents the ability to correctly identify severe chronic atrophic gastritis; and specificity represents the ability to correctly identify mild to moderate chronic atrophic gastritis. The optimal cutoff value is the predictor variable value that maximizes the Youden index.

[0225] Specifically, the operation command is:

[0226] data_all$Binary2 <- ifelse(data_all$Group == "Severe", 1, 0)

[0227] roc2 <- roc(data_all$Binary2, data_all$Red.light.positive.area.ratio)

[0228] auc2 <- auc(roc2)

[0229] coords2 <- coords(roc2, "best", ret = c("threshold", "sensitivity", "specificity"))

[0230] Binary2 is the second binary outcome variable, defined as 1 for the severe group and 0 for the mild and moderate groups combined.

[0231] roc2 is the second ROC curve;

[0232] auc2 is the area under the second ROC curve;

[0233] coords2 is used to extract the optimal cutoff value, sensitivity, and specificity.

[0234] 3.1.4 Generating three-level classification rules

[0235] Through the calculations in steps 3.1.2 and 3.1.3, the first threshold T1 and the second threshold T2 (Threshold 1 and Threshold 2, respectively) can be obtained and used to generate the three-level classification judgment rules.

[0236] Let X be the ACOD1 level (expressed as the ratio of red light-positive area) of the sample from the patient with chronic atrophic gastritis to be tested, and let the first threshold T1 and the second threshold T2 be as follows:

[0237] Threshold 1 = T1

[0238] Threshold 2 = T2

[0239] The following formula is used to determine whether a patient sample with chronic atrophic gastritis is classified as mild, moderate, or severe.

[0240]

[0241] In the steps of generating the first threshold T1 and the second threshold T2, the ROC curve is constructed using the basic plotting function `plot`, and the area below the curve is filled using the `polygon` function to enhance the readability and display effect of the graph. The AUC value is also labeled in the graph for a visual representation of performance. See the results below. Figure 3A and Figure 3B .

[0242] To further enhance the intuitiveness of the results, bar charts were constructed to visually compare the sensitivity and specificity under the first and second ROC curves. The `geom_bar` function from the `ggplot2` package was used to create grouped bar charts, and the `position_dodge` parameter was used to display different indicators side-by-side. The `geom_text` function was used to annotate specific values ​​to improve the clarity of information presentation. Furthermore, the `annotate` function was used to annotate the AUC values ​​and cut-off thresholds of each model in the graphs, giving the graphs both statistical information and visual representation. The results are shown below. Figure 3C .

[0243] 3.2 Results

[0244] from Figure 3A and 3C As can be seen, in the training cohort, the area under the curve (AUC) for the first ROC curve (mild vs. moderate + severe) is 0.950, indicating that ACOD1 has high discriminative power. The optimal cutoff value (first threshold T1) determined based on the maximum Yoden index is 0.239. At this threshold, the sensitivity is 0.889 and the specificity is 0.895, indicating that ACOD1 can effectively identify mild CAG patients while accurately excluding moderate + severe patients.

[0245] As shown in 3B and 3C, in the training cohort, the area under the curve (AUC) for the second ROC curve (mild + moderate vs. severe) is 0.989, further demonstrating that ACOD1 has extremely high discriminative ability. The optimal cutoff value (second threshold T2) determined based on the maximum Yoden index is 0.286. At this threshold, the sensitivity is 1.000 and the specificity is 0.946, indicating that ACOD1 has extremely high sensitivity in identifying severe CAG patients, while accurately excluding mild to moderate patients.

[0246] from Figures 3A-3B As can be seen, both ROC curves are significantly higher than the gray dashed reference (AUC=0.5), proving that ACOD1 has good diagnostic value. The Youden index corresponding to the first threshold T1 and the second threshold T2 reaches its maximum and can be used as the criterion for determining the clinical grading of CAG.

[0247] 3.3 Verification

[0248] 3.3.1 Verify queue status

[0249] Based on data from nearly 178 clinical samples (62 cases of mild CAG, 54 cases of moderate CAG, and 62 cases of severe CAG), the expression level of ACOD1 in gastric antral mucosal tissue samples was determined using the immunofluorescence staining method in step 2.3 (expressed as the ratio of red-positive area). Then, the distribution of the red-positive area ratio in the severity grading of chronic atrophic gastritis was analyzed using box plots. Results are shown below. Figure 4A .

[0250] from Figure 4A It was found that the red-positive area ratio showed significant distributional differences among the mild, moderate, and severe groups, generally increasing with the severity of CAG. Specifically, the index level was generally lower in the mild group, while the red-positive area ratio was higher in the moderate group than in the mild group, and further increased in the severe group with a wider distribution. Although there were some individual differences and overlaps among the groups, the overall trend was clear, suggesting that the red-positive area ratio of ACOD1 is well correlated with disease severity.

[0251] 3.3.2 Verification

[0252] The ACOD1 levels (expressed as the ratio of red light positive area) of nearly 178 clinical samples were used as continuous variables, and the grading rules obtained in the training phase were independently validated using validation samples.

[0253] The specific process is as follows:

[0254] 1) Import validation data and build the validation dataset:

[0255] qd <- read.csv("Mild (C1-C2).csv", header = TRUE, row.names = 1)

[0256] zd <- read.csv("Moderate(C3-O1).csv", header = TRUE, row.names = 1)

[0257] zzd <- read.csv("Severe (O2-O3).csv", header = TRUE, row.names = 1)

[0258] qd$Group <- "Mild"

[0259] zd$Group <- "Moderate"

[0260] zzd$Group <- "Severe"

[0261] data_all <- rbind(qd, zd, zzd)

[0262] data_all$Group <- factor(data_all$Group, levels = c("Mild", "Moderate", "Severe"))

[0263] The validation dataset is constructed in the same way as the training dataset to ensure a consistent input format.

[0264] 2) Repeatedly construct two binary classification ROC curves on the validation set and obtain the first threshold T1# and the second threshold T2# for the validation samples. First, validate the first threshold T1# (mild vs. moderate + severe):

[0265] data_all$Binary1 <- ifelse(data_all$Group == "Mild", 0, 1)

[0266] roc1 <- roc(data_all$Binary1, data_all$Red.light.positive.area.ratio)

[0267] auc1 <- auc(roc1)

[0268] coords1 <- coords(roc1, "best", ret = c("threshold", "sensitivity", "specificity"))

[0269] Further verification of the second threshold T2# (mild + moderate vs. severe):

[0270] data_all$Binary2 <- ifelse(data_all$Group == "Severe", 1, 0)

[0271] roc2 <- roc(data_all$Binary2, data_all$Red.light.positive.area.ratio)

[0272] auc2 <- auc(roc2)

[0273] coords2 <- coords(roc2, "best", ret = c("threshold", "sensitivity", "specificity"))

[0274] 3) Finally, output the AUC, threshold, sensitivity, and specificity results of the validation samples to evaluate the stability and discriminative performance of the first threshold T1# and the second threshold T2# in clinical samples:

[0275] roc_data <- data.frame(

[0276] Comparison = c("Mild vs Moderate+Severe", "Mild+Moderate vsSevere"),

[0277] AUC = c(auc1, auc2),

[0278] Cutoff = c(coords1["threshold"], coords2["threshold"]),

[0279] Sensitivity = c(coords1["sensitivity"], coords2["sensitivity"]),

[0280] Specificity = c(coords1["specificity"], coords2["specificity"]) )

[0282] write.csv(roc_data, "ROC_Cutoff_Sensitivity_Specificity.csv", row.names = FALSE)

[0283] The ability to distinguish severity levels of CAG was assessed by constructing receiver operating characteristic (ROC) curves, and the area under the curve (AUC) was calculated to quantify the overall discriminative power. Results are shown in [Figure number missing]. Figure 4B and Figure 4C .

[0284] In addition to ROC curves, the validation results also included a bar chart to visually compare the sensitivity and specificity of different levels of CAG. The corresponding AUC values ​​and the first threshold T1# and second threshold T2# were labeled in the chart to more intuitively demonstrate the diagnostic performance of ACOD1. The construction of the bar chart was the same as in step 3.2, and the results are shown below. Figure 4D .

[0285] In an independent validation cohort, the results that the first and second thresholds can be used as criteria for determining the clinical classification of CAG were further validated.

[0286] from Figure 4B and 4D The validation cohort showed that, for the binary classification of "mild" and "moderate to severe" (mild vs. moderate + severe), ROC analysis revealed an area under the curve (AUC) of 0.926, indicating that ACOD1 has high discriminative power. The first threshold T1#, determined based on the maximum Yoden index, was 0.258. At this threshold, the sensitivity was 0.759 and the specificity was 0.968, demonstrating that ACOD1 exhibits good sensitivity in identifying mild CAG patients while maintaining excellent specificity.

[0287] from Figure 4C and 4D The validation cohort showed that, for the binary classification of "severe" and "mild-moderate" (mild + moderate vs. severe), ROC analysis yielded an AUC of 0.963, further demonstrating ACOD1's excellent discriminative ability. Simultaneously, the corresponding second threshold T2#, determined based on the maximum Yoden index, was 0.285. At this threshold, the sensitivity was 0.984 and the specificity was 0.897, confirming that ACOD1 possesses extremely high sensitivity in identifying severe CAG patients while maintaining excellent specificity.

[0288] Based on the combined results of the training and validation cohorts, ACOD1 demonstrates good stability and reproducibility as a biomarker for grading the severity of chronic atrophic gastritis (CAG). By comparing the ACOD1 levels in the test samples with the first threshold T1 and the second threshold T2, patients with mild, moderate, and severe CAG can be effectively distinguished, thus achieving precise three-tier grading of CAG.

[0289] The above embodiments are for illustrating the implementation schemes disclosed in this invention and should not be construed as limiting the invention. Furthermore, various modifications listed herein, as well as variations in the methods and compositions of the invention, will be apparent to those skilled in the art without departing from the scope and spirit of the invention. Although the invention has been specifically described in conjunction with various specific preferred embodiments, it should be understood that the invention should not be limited to these specific embodiments. In fact, various modifications as described above that are obvious to those skilled in the art to obtain the invention should be included within the scope of this invention.

Claims

1. A method for grading and assessing the severity of chronic atrophic gastritis, characterized in that, include: Based on the ACOD1 level of the samples of patients with chronic atrophic gastritis, patients with chronic atrophic gastritis were classified as mild, moderate, or severe.

2. The evaluation method as described in claim 1, characterized in that, The ACOD1 level of the sample to be tested is compared with a first threshold T1 and / or a second threshold T2, wherein the first threshold T1 is less than the second threshold T2. When the ACOD1 level of the sample to be tested is less than or equal to the first threshold T1, it is determined to be mild. When the ACOD1 level of the sample to be tested is greater than the first threshold T1 and less than or equal to the second threshold T2, it is determined to be moderate. When the ACOD1 level of the sample to be tested is greater than the second threshold T2, it is determined to be severe.

3. The evaluation method as described in claim 1, characterized in that: The first threshold T1 is obtained using a method comprising the following steps: Acquire the ACOD1 levels of known mild, moderate, and severe chronic atrophic gastritis patient samples, label the mild group as the first negative category, and combine the moderate and severe groups and label them as the first positive category to obtain the first binary outcome variable Binary1; The first ROC curve was constructed using the first binary outcome variable Binary1 as the state variable and the ACOD1 level as the predictor variable. The first Youden index is calculated based on the first ROC curve, and the ACOD1 level corresponding to the maximum of the first Youden index is determined as the first threshold T1. The second threshold T2 is obtained using a method comprising the following steps: The mild and moderate groups were merged and labeled as the second negative category, and the severe group was labeled as the second positive category, resulting in the second binary outcome variable Binary2; A second ROC curve was constructed using the second binary outcome variable Binary2 as the state variable and the ACOD1 level as the predictor variable. The second Yoden index is calculated based on the second ROC curve, and the ACOD1 level corresponding to the maximum second Yoden index is determined as the second threshold T2.

4. The evaluation method as described in claim 1, characterized in that, The samples were selected from one or more of the following: gastric tissue, gastric juice, plasma, serum, saliva, urine, and feces. And / or, the ACOD1 level refers to the ACOD1 gene transcript, protein, or its detection signal in the sample; And / or, the ACOD1 level is obtained by detecting the sample through immunological detection, nucleic acid detection or mass spectrometry detection; And / or, the first threshold is 0.24–0.26; And / or, the second threshold is 0.27-0.

28.

5. A severity grading and assessment system for chronic atrophic gastritis, characterized in that, The system includes: The judgment unit is used to compare the ACOD1 level of the patient sample with the first threshold T1 and / or the second threshold T2. When the ACOD1 level of the sample to be tested is less than or equal to the first threshold T1, it is determined to be mild. When the ACOD1 level of the sample to be tested is greater than the first threshold T1 and less than or equal to the second threshold T2, it is determined to be moderate. When the ACOD1 level of the sample to be tested is greater than the second threshold T2, it is determined to be severe.

6. The evaluation system as described in claim 5, characterized in that, The samples were selected from one or more of the following: gastric tissue, gastric juice, plasma, serum, saliva, urine, and feces. And / or, the ACOD1 level refers to the ACOD1 gene transcript, protein, or its detection signal in the sample; And / or, the ACOD1 level is obtained by detecting the sample through immunological detection, nucleic acid detection or mass spectrometry detection; And / or, the system further includes a data unit for obtaining ACOD1 level data of the patient sample with chronic atrophic gastritis to be tested.

7. The evaluation system as described in claim 5, characterized in that, The system further includes an analysis unit, which is configured to: Acquire the ACOD1 levels of known mild, moderate, and severe chronic atrophic gastritis patient samples, label the mild group as the first negative category, and combine the moderate and severe groups and label them as the first positive category to obtain the first binary outcome variable Binary1; Using the first binary outcome variable Binary1 as the state variable and the ACOD1 level as the predictor variable, a first ROC curve is constructed, and the ACOD1 level corresponding to the maximum Youden index on the first ROC curve is determined as the first threshold T1. The mild and moderate groups were merged and labeled as the second negative category, and the severe group was labeled as the second positive category, resulting in the second binary outcome variable Binary2; Using the second binary outcome variable Binary2 as the state variable and the ACOD1 level as the predictor variable, a second ROC curve is constructed, and the ACOD1 level corresponding to the maximum Youden index on the second ROC curve is determined as the second threshold T2.

8. A device comprising a processor and a memory, the memory being used to store a computer program, characterized in that, The processor is configured to execute a computer program stored in the memory to cause the device to perform the evaluation method as described in any one of claims 1-4.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the evaluation method as described in any one of claims 1-4.

10. Application of substances that detect ACOD1 in the preparation of products for grading the severity of chronic atrophic gastritis.

Images