Gastric cancer risk assessment and diagnosis method based on gastric mucosa micro-ecological change

Abstract

The invention discloses a gastric cancer risk assessment and diagnosis method based on gastric mucosa micro-ecological change, relates to the technical field of medical diagnosis and biology, and provides the gastric cancer risk assessment and diagnosis method based on the gastric mucosa micro-ecological change. The method comprises the following steps: collecting gastric mucosa tissues or gastric juice samples, extracting total DNA (deoxyribonucleic acid) of microorganisms, performing high-throughput sequencing by using an Illumina MiSeq platform, analyzing the diversity of the microorganisms, and identifying key flora. And inputting a random forest model by combining the concentration of metabolites such as short-chain fatty acid and clinical information, outputting a gastric cancer risk score, and dividing risk grades. The model is trained based on a large number of samples, and high accuracy is ensured. High-risk persons suggest gastroscope precise examination within three months, and personalized intervention suggestions are provided, so that early warning and precise health management are realized.

Description

Technical Field

[0001] This invention belongs to the field of medical diagnostics and biotechnology, specifically relating to a gastric cancer risk assessment and diagnostic method based on changes in gastric mucosal microecology. By analyzing changes in the structure, functional gene expression, and metabolites of the microbial community in gastric mucosal tissue or gastric juice, and combining them with a machine learning model, an early and accurate assessment of an individual's risk of developing gastric cancer can be achieved. Background Technology

[0002] Gastric cancer is one of the most common malignant tumors worldwide, and its occurrence and development are closely related to Helicobacter pylori infection, chronic inflammation, environmental factors, and genetic susceptibility. Recent studies have shown that gastric mucosal microecological imbalance plays a crucial role in the development of gastric cancer. Traditional diagnostic methods, such as gastroscopy biopsy and serological testing, suffer from problems such as high invasiveness, insufficient sensitivity, or latency. Existing microbiome-based research largely focuses on gut microbiota, and a standardized, widely applicable diagnostic method for systematically assessing the relationship between local gastric mucosal microecology and gastric cancer risk is still lacking. Therefore, there is an urgent need to establish a non-invasive or minimally invasive gastric cancer risk assessment technique based on gastric mucosal microecological characteristics. Summary of the Invention

[0003] The purpose of this invention is to provide a gastric cancer risk assessment and diagnosis method based on changes in the gastric mucosal microecology. By analyzing the composition, functional pathways and metabolites of the gastric mucosal microorganisms from multiple dimensions and combining them with clinical parameters, a risk assessment model is constructed to achieve early warning and stratified diagnosis of gastric cancer.

[0004] The technical solution of this invention is implemented as follows: a method for assessing and diagnosing gastric cancer risk based on changes in the gastric mucosal microecology, comprising the following steps: S1. Obtain gastric mucosal tissue or gastric fluid samples from the subjects, and preserve and transport them at 4°C using aseptic techniques. S2. Extract total microbial DNA from the sample and purify it using a DNA extraction kit according to standard procedures; S3. High-throughput sequencing was performed using the Illumina MiSeq platform to obtain microbial community composition data, analyze microbial α-diversity and β-diversity, and identify key differential bacterial groups. S4. High performance liquid chromatography was used to detect the concentration of key metabolites related to gastric cancer in the sample, such as short-chain fatty acids and amino acids. S5. Obtain clinical information of the subjects, including Helicobacter pylori infection status, gastritis type and pathological grade, and confirm through endoscopy and biopsy. S6. Input the microbial community data, metabolite data and clinical information into a pre-trained random forest model; S7. Output gastric cancer risk score and classify it into high, medium and low risk levels based on the score.

[0005] In a preferred embodiment, the key differentially expressed bacterial communities include those with significantly varying abundance. Helicobacter pylori, Prevotella, Fusobacterium, Bacteroides and Streptococcus At least one of the genera, and its relative abundance change serves as an important biomarker for assessing gastric cancer risk.

[0006] As a preferred embodiment, the short-chain fatty acids include acetic acid, propionic acid, and butyric acid. When their concentration is more than 20% lower than that of the normal control group, it indicates that there is an imbalance of gastric mucosal microecology and a high risk of gastric cancer.

[0007] As a preferred embodiment, the clinical information also includes the degree of gastric mucosal atrophy, the state of intestinal metaplasia, and a family history of gastric cancer. This information is collected through pathological reports and questionnaires and standardized into a structured data input model.

[0008] In a preferred embodiment, the training process of the random forest model includes: (1) Collect retrospective sample data from no less than 1,000 patients with pathologically confirmed gastric cancer and healthy controls; (2) Feature screening was performed on microbiome, metabolites and clinical variables, and the input feature set was optimized using recursive feature elimination (RFE); (3) 10-fold cross-validation is performed during model training to ensure that the AUC value is not lower than 0.90.

[0009] As a preferred embodiment, the gastric cancer risk score ranges from 0 to 100 points, where 0 to 30 points indicate low risk, 31 to 70 points indicate medium risk, and 71 to 100 points indicate high risk. Individuals with high risk undergo detailed gastroscopy within 3 months.

[0010] In a preferred embodiment, the high-throughput sequencing has a sequencing depth of not less than 50,000 reads / sample, and uses QIIME2 or Mothur software for OTU clustering and species annotation, with Greengenes or SILVA as the database.

[0011] As a preferred implementation, the method also includes generating a personalized intervention recommendation report after outputting the risk level. The recommendations include a Helicobacter pylori eradication treatment plan, dietary adjustment suggestions, probiotic supplementation recommendations, and a follow-up schedule.

[0012] After adopting the above technical solution, the beneficial effects of the present invention are: 1. By integrating gastric mucosal microecology, metabolites, and multidimensional clinical data, and combining them with machine learning models, we can overcome the limitations of traditional diagnostic methods in terms of lag and invasiveness, and significantly improve the sensitivity and specificity of gastric cancer risk warning.

[0013] 2. The random forest model was trained and validated based on a large sample of retrospective data. Recursive feature elimination and cross-validation were used for optimization to ensure stable model performance (AUC≥0.90), demonstrating good clinical applicability and potential for standardized promotion.

[0014] 3. By outputting risk stratification results, personalized intervention suggestions can also be generated, covering Helicobacter pylori eradication, dietary adjustments, probiotic supplementation, and follow-up plans, helping to achieve precise health management of "early detection and early intervention". Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below.

[0016] Figure 1 This is a flowchart of the risk assessment and diagnosis method of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Example

[0018] A method for assessing and diagnosing gastric cancer risk based on changes in the gastric mucosal microecology includes the following steps: S1. Obtain gastric mucosal tissue or gastric fluid samples from the subjects, and preserve and transport them at 4°C using aseptic techniques. S2. Extract total microbial DNA from the sample and purify it using a DNA extraction kit according to standard procedures; S3. High-throughput sequencing was performed using the Illumina MiSeq platform to obtain microbial community composition data, analyze microbial α-diversity and β-diversity, and identify key differential bacterial groups. S4. High-performance liquid chromatography was used to detect the concentration of key metabolites associated with gastric cancer in the sample, such as short-chain fatty acids and amino acids. S5. Obtain clinical information of the subjects, including Helicobacter pylori infection status, gastritis type and pathological grade, and confirm through endoscopy and biopsy. S6. Input the microbial community data, metabolite data and clinical information into a pre-trained random forest model; S7. Output gastric cancer risk score and classify it into high, medium and low risk levels based on the score.

[0019] Furthermore, key differentially expressed microbiota include those with significant abundance variations. Helicobacter pylori, Prevotella, Fusobacterium, Bacteroides and Streptococcus At least one of the genera, and its relative abundance change serves as an important biomarker for assessing gastric cancer risk.

[0020] In practical applications, changes in the abundance of gut microbiota may appear before clinical symptoms. Detecting these changes can provide early warning of gastric cancer risks and offer a window of opportunity for timely intervention. Furthermore, gut microbiota characteristics vary among individuals. Combining these biomarkers can enable more precise and personalized gastric cancer risk assessments, which can help develop targeted screening and prevention strategies.

[0021] Furthermore, when the concentration of short-chain fatty acids, including acetic acid, propionic acid, and butyric acid, is more than 20% lower than that of the normal control group, it indicates an imbalance in the gastric mucosal microecology and a high risk of gastric cancer.

[0022] In practical applications, changes in short-chain fatty acid (SCFA) concentration serve as biomarkers for assessing gastric cancer risk. Changes in SCFA concentration can reflect the progression of gastric cancer and contribute to a deeper understanding of its pathogenesis.

[0023] Furthermore, clinical information also includes the degree of gastric mucosal atrophy, the state of intestinal metaplasia, and a family history of gastric cancer. This information is collected through pathological reports and questionnaires and standardized into a structured data input model.

[0024] In practical applications, by using multi-dimensional clinical information such as the degree of gastric mucosal atrophy, the state of intestinal metaplasia, and family history of gastric cancer, an individual's gastric cancer risk status can be more comprehensively reflected, thus improving the accuracy of risk assessment.

[0025] Furthermore, the training process for the random forest model includes: (1) Collect retrospective sample data from no less than 1,000 patients with pathologically confirmed gastric cancer and healthy controls; (2) Feature screening was performed on microbiome, metabolites and clinical variables, and the input feature set was optimized using recursive feature elimination (RFE); (3) 10-fold cross-validation is performed during model training to ensure that the AUC value is not lower than 0.90.

[0026] In practical applications, random forest models trained through rigorous sample data collection, feature selection and optimization, and cross-validation can provide highly accurate and robust predictions of gastric cancer risk, which helps to achieve early screening and intervention for gastric cancer.

[0027] Furthermore, the gastric cancer risk score ranges from 0 to 100, with 0 to 30 indicating low risk, 31 to 70 indicating medium risk, and 71 to 100 indicating high risk. Individuals at high risk should undergo a detailed gastroscopy within 3 months.

[0028] Furthermore, the high-throughput sequencing depth is no less than 50,000 reads / sample, and OTU clustering and species annotation are performed using QIIME2 or Mothur software, with Greengenes or SILVA as the database.

[0029] In practical applications, it provides more comprehensive coverage of the microbial community in samples, reduces the underestimation of diversity, improves detection sensitivity, standardizes bioinformatics processes, reduces analytical bias, and enhances the comparability and reproducibility between different studies.

[0030] Furthermore, after outputting the risk level, it also includes generating a personalized intervention recommendation report, which includes Helicobacter pylori eradication treatment plan, dietary adjustment suggestions, probiotic supplementation recommendations, and follow-up schedule.

[0031] This invention provides a method for assessing and diagnosing gastric cancer risk based on changes in the gastric mucosal microecology. It involves collecting gastric mucosal or gastric juice samples, extracting microbial DNA and performing high-throughput sequencing to analyze microbial diversity and changes in key bacterial communities; simultaneously detecting the concentration of metabolites such as short-chain fatty acids, and integrating clinical information. The multidimensional data is input into a random forest model trained on a large sample size, outputting a gastric cancer risk score of 0–100 points, classifying it into high, medium, and low risk levels, achieving early warning of gastric cancer, and providing personalized intervention suggestions.

[0032] The above description is only a preferred embodiment of the present invention. It should be noted that those skilled in the art can make several improvements and additions without departing from the method of the present invention, and these improvements and additions should also be considered within the scope of protection of the present invention.

Claims

1. A method for assessing and diagnosing gastric cancer risk based on changes in the gastric mucosal microecology, characterized in that, Includes the following steps: S1. Obtain gastric mucosal tissue or gastric fluid samples from the subjects, and preserve and transport them at 4°C using aseptic techniques. S2. Extract total microbial DNA from the sample and purify it using a DNA extraction kit according to standard procedures; S3. High-throughput sequencing was performed using the Illumina MiSeq platform to obtain microbial community composition data, analyze microbial α-diversity and β-diversity, and identify key differential bacterial groups. S4. High performance liquid chromatography was used to detect the concentration of key metabolites related to gastric cancer in the sample, such as short-chain fatty acids and amino acids. S5. Obtain clinical information of the subjects, including Helicobacter pylori infection status, gastritis type and pathological grade, and confirm through endoscopy and biopsy. S6. Input the microbial community data, metabolite data and clinical information into a pre-trained random forest model; S7. Output gastric cancer risk score and classify it into high, medium and low risk levels based on the score.

2. The method for assessing and diagnosing gastric cancer risk based on changes in gastric mucosal microecology according to claim 1, characterized in that, The key differentially expressed microbial communities include those with significantly varying abundance. Helicobacter pylori, Prevotella, Fusobacterium, Bacteroides and Streptococcus At least one of the genera, and its relative abundance change serves as an important biomarker for assessing gastric cancer risk.

3. The method for assessing and diagnosing gastric cancer risk based on changes in gastric mucosal microecology according to claim 1, characterized in that, The short-chain fatty acids, including acetic acid, propionic acid, and butyric acid, are indicated when their concentrations are more than 20% lower than those of the normal control group, suggesting an imbalance in the gastric mucosal microecology and a high risk of gastric cancer.

4. The method for assessing and diagnosing gastric cancer risk based on changes in gastric mucosal microecology according to claim 1, characterized in that, The clinical information also includes the degree of gastric mucosal atrophy, the state of intestinal metaplasia, and a family history of gastric cancer. This information is collected through pathological reports and questionnaires and standardized into a structured data input model.

5. The method for assessing and diagnosing gastric cancer risk based on changes in gastric mucosal microecology according to claim 1, characterized in that, The training process of the random forest model includes: (1) Collect retrospective sample data from no less than 1,000 patients with pathologically confirmed gastric cancer and healthy controls; (2) Feature screening was performed on microbiome, metabolites and clinical variables, and the input feature set was optimized using recursive feature elimination (RFE); (3) 10-fold cross-validation is performed during model training to ensure that the AUC value is not lower than 0.

90.

6. The gastric cancer risk assessment and diagnostic method based on changes in gastric mucosal microecology according to claim 1, characterized in that, The gastric cancer risk score ranges from 0 to 100, with 0 to 30 indicating low risk, 31 to 70 indicating medium risk, and 71 to 100 indicating high risk. Individuals at high risk should undergo a detailed gastroscopy within 3 months.

7. The method for assessing and diagnosing gastric cancer risk based on changes in gastric mucosal microecology according to claim 1, characterized in that, The high-throughput sequencing has a sequencing depth of no less than 50,000 reads / sample, and uses QIIME2 or Mothur software for OTU clustering and species annotation, with Greengenes or SILVA as the database.

8. The method for assessing and diagnosing gastric cancer risk based on changes in gastric mucosal microecology according to claim 1, characterized in that, It also includes generating a personalized intervention recommendation report after outputting the risk level. The recommendations include Helicobacter pylori eradication treatment plan, dietary adjustment suggestions, probiotic supplementation recommendations, and follow-up schedule.

Images