Oral microbial gene marker for noninvasive diagnosis of esophageal cancer patient and application of oral microbial gene marker
By applying machine learning and oral microbiome data to esophageal cancer patients, a non-invasive esophageal cancer diagnostic model was established to detect specific microbial gene markers, solving the problem of early diagnosis and achieving highly specific and sensitive esophageal cancer detection.
Patent Information
- Application Number
- CN202511404471.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2025-09-08
- Filing Date
- 2025-09-29
- Publication Date
- 2026-01-02
AI Technical Summary
Existing technologies lack highly specific and sensitive non-invasive detection methods for the early diagnosis of esophageal cancer, and there are challenges in processing and interpreting oral microbiome data, leading to esophageal cancer patients being diagnosed at an intermediate or late stage, which affects treatment outcomes.
Using machine learning technology combined with oral microbiome data, a microbial gene biomarker model was established to distinguish between esophageal cancer patients and healthy individuals by detecting six specific microbial genes (Alloprevotella, Prevotella, Campylobacter, Neisseria, etc.). Specific primers were used for detection, and the distinguishing model was established through high-throughput sequencing analysis.
It achieves highly efficient differentiation between esophageal cancer patients and healthy individuals, with a differentiation capability of 93.96%-95.18%, significantly improving the accuracy and feasibility of early diagnosis and providing a non-invasive detection method.
Smart Images
Figure CN121249920A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of biomedical technology, specifically relating to an oral microbial gene marker for non-invasive diagnosis of esophageal cancer patients and its application. Background Technology
[0002] Esophageal cancer is the seventh most common cancer worldwide and the sixth leading cause of cancer-related death, posing a significant threat to public health. The main pathological types of esophageal cancer include squamous cell carcinoma and adenocarcinoma, with squamous cell carcinoma being more common. In recent years, the incidence of esophageal cancer has been steadily rising. Due to the lack of early screening and diagnostic methods, most patients are diagnosed at an advanced stage, resulting in an overall 5-year survival rate of only 13%-18%. In fact, 40% of patients already have distant metastases at initial diagnosis, which is a major cause of treatment failure and death in esophageal cancer patients. Early diagnosis allows for timely intervention and is crucial for improving clinical prognosis. However, the gold standard for diagnosing esophageal cancer is endoscopy, an invasive and expensive procedure that limits its widespread use. Therefore, there is an urgent need to develop non-invasive detection methods with high specificity and sensitivity.
[0003] The human microbiome plays a crucial role in the occurrence, development, and treatment response of diseases. Recent studies have revealed the association between the oral microbiome and diseases, including local lesions (such as dental caries and periodontal disease) and systemic diseases (such as cancer). Therefore, the oral microbiome is increasingly being used to identify biomarkers and potential therapeutic targets for various diseases. Over the past five years, progress has been made in research on the role of the oral microbiome in the development and progression of esophageal diseases, including precancerous lesions, esophageal squamous cell carcinoma, and esophageal adenocarcinoma. Oral microbiome dysbiosis may participate in the development and progression of esophageal diseases by regulating tumor immune responses, promoting inflammation, and influencing the expression of aberrant signaling molecules. However, many challenges remain, including small sample sizes and a lack of independent validation. Therefore, the development of reliable microbial biomarkers suitable for clinical applications is crucial.
[0004] Although the microbiome encompasses a wide range of microorganisms found in clinical samples, processing and interpreting such complex data remains challenging. Machine learning, as a widely used artificial intelligence technique, has demonstrated its value in automating data analysis across multiple medical fields. It offers unique advantages in interpreting omics data, developing predictive models, identifying biomarkers, and stratifying patients for precision medicine. However, research applying machine learning to analyze microbiome data and develop potential biomarkers for esophageal cancer is still in its early stages, highlighting the potential for further research in this area. Currently, no oral microbiome models have been reported to differentiate between esophageal cancer patients and healthy individuals. Summary of the Invention
[0005] This invention provides an oral microbial gene marker for distinguishing esophageal cancer patients from healthy individuals, composed of six microbial genes shown in SEQ ID NO:1-6, which are enriched in the human oral cavity.
[0006] The gene sequences of the six microorganisms shown in SEQ ID NO:1-6 are as follows: Alloprevotella (>ASV18) TGAGGAATATTGGTCAATGGGCGAGAGCCTGAACCAGCCAAGTAGCGTGCAGGATGACGGCCCTCCGGGTTGTAAACTGCTTTTAGTTGGGAATAAAAAAAGGGACTTGTCCCTTCTTGTATGTACCTTCAGAAAAAGGACCGGCTAATTCCGTGCCAGCAGCCGCGGTAATACGGAAGGTCCAGGCGTTATCCGGATTTATTGGGTTTA AAGGGAGCGTAGGCGGATTGTTAAGTCAGCGGTAAAGGGTGTGGCTCAACCATGCATTGCCGTTGAAACTGGCGATTCTTGAGTGCAGACAGGGATGCCGGAATTCGTGGTGTAGCGGTGAAATGCTTAGATATCACGAAGAACTCCGATCGCGAAGGCAGGTGTCCGGGCTGCAACTGACGCTGAGGCTCGAAAGTGTGGGTATCAAACA Prevotella (>ASV21) TGAGGAATATTGGTCAATGGACGGAAGTCTGAACCAGCCAAGTAGCGTGCAGGATGACGGCCCTATGGGTTGTAAACTGCTTTTGTATGGGGATAAAGTTAGGGACGTGTCCCTATTTGCAGGTACCATACGAATAAGGACCGGCTAATTCCGTGCCAGCAGCCGCGGTAATACGGAAGGTCCAGGCGTTATCCGGATTTATTGGGTTTAAAGGGAGCGTAGGCTGGAGATTAAGTGTGTTGTGAAATGTAGACGCTCAACGTCTGAATTGCAGCGCATACTGGTTTCCTTGAGTACGCACAACGTTGGCGGAATTCGTCGTGTAGCGGTGAAATGCTTAGATATGACGAAGAACTCCGATTGCGAAGGCAGCTGACGGGAGCGCAACTGACGCTTAAGCTCGAAGGTGCGGGTATCGAACA Campylobacter (>ASV50) TAGGGAATATTGCTCAATGGGGGAAACCCTGAAGCAGCAACGCCGCGTGGAGGATGACACTTTTCGGAGCGTAAACTCCTTTTGTTAGGGAAGAACAATGACGGTACCTAACGAATAAGCACCGGCTAACTCCGTGCCAGCAGCCGCGGTAATACGGAGGGTGCAAGCGTTACTCGGAATCACTGGGCGTAAAGGACGCGTAGGCGGATTATCAAGTCTCTTGTGAAATCCTATGGCTTAACCATAGAACTGCTTGGGAAACTGATAATCTAGAGTGAGGGAGAGGCAGATGGAATTGGTGGTGTAGGGGTAAAATCCGTAGAGATCACCAGGAATACCCATTGCGAAGGCGATCTGCTGGAACTCAACTGACGCTAATGCGTGAAAGCGTGGGGAGCAAACA Neisseria (>ASV67) TGGGGAATTTTGGACAATGGGCGCAAGCCTGATCCAGCCATGCCGCGTGTCTGAAGAAGGCCTTCGGGTTGTAAAGGACTTTTGTCAGGGAAGAAAAGGGCGGGGTTAATACCCCTGTCTGATGACGGTACCTGAAGAATAAGCACCGGCTAACTACGTGCCAGCAGCCGCGGTAATACGTAGGGTGCGAGCGTTAATCGGAATTACTGGGCGTAAAGCGGGCGCAGACGGTTACTTAAGCAGGATGTGAAATCCCCGGGCTCAACCTGGGAACTGCGTTCTGAACTGGGTGACTAGAGTGTGTCAGAGGGAGGTAGAATTCCACGTGTAGCAGTGAAATGCGTAGAGATGTGGAGGAATACCGATGGCGAAGGCAGCCTCCTGGGATAACACTGACGTTCATGCCCGAAAGCGTGGGTAGCAAACA Campylobacter (>ASV215) TAGGGAATATTGCGCAATGGGGGAAACCCTGACGCAGCAACGCCGCGTGGAGGATGACACTTTTCGGAGCGTAAACTCCTTTTGTTAGGGAAGAATAATGACGGTACCTAACGAATAAGCACCGGCTAACTCCGTGCCAGCAGCCGCGGTAATACGGAGGGTGCAAGCGTTACTCGGAATCACTGGGCGTAAAGGACGCGTAGGCGGATTATCAAGTCTCTTGTGAAATCTAGTGGCTTAACCACTAAACTGCTTGGGAAACTGATAATCTAGAGTAAGGGAGAGGCAGATGGAATTCTTGGTGTAGGGGTAAAATCCGTAGAGATCAAGAAGAATACTTATTGCGAAGGCGATCTGCTAGAACTTAACTGACGCTAATGCGTGAAAGCGTGGGGAGCAAACA Neisseria (>ASV329) TGGGGAATATTGCACAATGGGGGAAACCCTGATGCAGCGACGCCGCGTGAGTGAAGAAGTATTTCGGTATGTAAAGCTCTATCAGCAGGGAAGATAATGACAGTACCTGACTAAGAAGCCCCGGCTAACTACGTGCCAGCAGCCGCGGTAATACGTAGGGTGCGAGCGTTAATCGGAATTACTGGGCGTAAAGCGAGCGCA GACGGTTACTTAAGCAGGATGTGAAATCCCCGGGCTCAACCTGGGAACTGCGTTCTGAACTGGGTGACTAGAGTGTGTCAGAGGGAGGTAGAATTCCACGTTGTAGCAGTGAAATGCGTAGAGATGTGGAGGAATACCGATGGCGAAGGCAGCCTCCTGGGATAACACTGACGTTCATGCTCGAAAGCGTGGGTAGCAAACA In addition, the present invention provides a detection reagent comprising primers for detecting the six microbial genes shown in SEQ ID NO:1-6, wherein the primer sequences are SEQ ID NO:7-8.
[0007] Primers Sequencing region V3+V4: 341F-805R Upstream primer: 5'-CCTACGGGNGGCWGCAG-3' Downstream primer: 5'-GACTACHVGGGTATCTAATCC-3' The present invention also provides the application of the detection reagent in establishing a tongue coating microbial model (detection kit) that distinguishes between esophageal cancer patients and healthy individuals. The detection reagent is suitable for detecting the six microbial genes shown in SEQ ID NO:1-6.
[0008] The microbial differentiation model is applicable to distinguishing between esophageal cancer patients and healthy individuals, where the esophageal cancer patients are pathologically diagnosed.
[0009] The tongue swabs of the subjects were tested to determine whether the sample contained the microbial genes and whether a microbial gene model could be established to distinguish between esophageal cancer patients and healthy individuals.
[0010] Tongue swab samples were collected from the subjects, total microbial DNA was extracted, and 16S rDNA Miseq sequencing of the microbial DNA was performed to detect the presence of the six microbial genes described in claim 1.
[0011] Furthermore, total microbial DNA was extracted from tongue swab samples collected from the enrolled subjects, and 16S rDNA Miseq sequencing of the oral microbiota was performed. Based on the high-throughput sequencing data, a microbial differentiation model between esophageal cancer patients and healthy individuals was established in the discovery cohort, and a probability of disease (POD) index was established for esophageal cancer patients. The POD index was used to calculate and validate its discriminative ability in the validation cohort, thus realizing the universality of the microbial gene differentiation model in esophageal cancer patients.
[0012] Specifically, it includes: (1) Collect tongue swab samples from the subjects (esophageal cancer patients and healthy people), extract total microbial DNA from the tongue samples according to the standard DNA extraction method, and perform 16S rDNA high-throughput sequencing of tongue flora on the Illumina MiSeq platform. (2) Based on high-throughput sequencing data, in the discovery cohort of the microbial differentiation model, among 153 esophageal cancer patients and 167 healthy controls, the six best microbial gene markers for the model were identified using a random forest model and a five-fold cross-validation algorithm.
[0013] (3) Based on six microbial gene markers, the probability of disease (POD) index of esophageal cancer patients was calculated by using the ratio of randomly generated decision trees.
[0014] (4) The microbial differentiation model achieved a differentiation ability of 93.96% between 107 esophageal cancer patients and 117 healthy controls. The POD index was significantly increased in esophageal cancer patients, and there was a significant difference between the two groups (p<0.0001).
[0015] (5) In the validation cohort, the microbial differentiation model achieved a differentiation ability of 95.18% between 50 esophageal cancer patients and 46 healthy controls; the POD index was significantly increased in esophageal cancer patients, and there was a significant difference between the two groups (p<0.0001).
[0016] Therefore, the microbial gene differentiation model of the present invention has achieved good differentiation ability between esophageal cancer patients and healthy controls.
[0017] Additionally, a kit for differentiating oral microbial models of esophageal cancer patients and healthy controls is provided, comprising primers for detecting the six microbial genes shown in SEQ ID NO:1-6 as described in claim 1.
[0018] The specific operation steps of this invention are as follows: (1) In accordance with the design principles of prospective clinical trials, the research design of this invention is as follows: Figure 1 As shown. The study protocol was approved by the Ethics Committee of Zhengzhou University Cancer Hospital. All enrolled patients signed informed consent forms for the study protocol and clinical sample collection.
[0019] (2) To ensure a sufficient number of effective samples, two sets of tongue coating samples were collected from each subject. To minimize the impact of factors such as diet and medication on changes in the tongue coating microecology, each subject was required to fast for at least 2 hours before sample collection and rinse their mouth with saline 2-3 times. A specially trained operator scraped the surface of the tongue coating 2-3 times with a throat swab, and then immersed the tip of the throat swab into a test tube. All collected samples were inactivated in a 56°C water bath for at least 30 minutes and finally stored in a freezer at -80°C. The extraction method for total DNA from oral bacteria was performed according to the kit instructions.
[0020] (3) Amplification of total oral bacterial DNA samples and construction of DNA libraries were completed, and 16S rDNA sequencing was performed on the Illumina Miseq sequencing platform. All output sequences underwent basic preprocessing and basic bioinformatics analysis.
[0021] (4) The DADA2 (Divisive Amplicon Denoising Algorithm 2) algorithm was used to reduce noise in the raw sequencing data to obtain an amplicon sequence variant (ASV) table. Representative sequences of each ASV were annotated based on the SILVA reference database (SSU138.2), and then taxonomic analysis was performed to obtain information on the species composition of the samples. The ASV gene sequences of all samples were collected and organized according to the microbial gene marker discovery cohort and the microbial gene marker validation cohort.
[0022] (5) Based on representative sequences generated from high-throughput sequencing data, frequency files of ASVs in the discovery set and the validation set of microbial genetic markers were calculated. These ASVs were used in a correlation study to identify the abundance of ASVs that showed significant differences between esophageal cancer patients and healthy individuals. The Wilcoxon test was used to statistically analyze the microbial genetic markers that showed differences between esophageal cancer patients and healthy individuals. Six selected microbial genetic markers were further analyzed.
[0023] (6) In the discovery cohort of the microbial differential model, including 107 esophageal cancer patients and 117 healthy controls, microbial gene biomarkers were screened using the selected ASV abundance files in a random forest model (R software 3.4.1 and random forest package 4.6–12) with a five-fold cross-validation algorithm (software parameters were default except for setting "importance=TRUE"). Five trials of five-fold cross-validation were conducted, and the cross-validation error curve was obtained, with the smallest cross-validation error point used as the cut-off value. The minimum cross-validation error value plus the standard deviation of the corresponding value was used as the cut-off value. A set of ASV biomarkers with error rates less than the cut-off value was selected, and the set with the smallest number of ASVs was selected as the optimal set of microbial gene biomarkers. Finally, the six best microbial gene biomarkers for this model were identified. Figure 2 , Figure 3 The gene sequences of the six selected microbial OTUs markers are shown in SEQ ID NO:1-6.
[0024] (7) The probability of disease (POD) index is calculated using the ratio of randomly generated decision trees. The decision tree predicts the sample as "CP", and the predicted parameters are set as follows: probability=T, norm.votes=T, predict.all=TRUE. The random forest model built in the LOO mode is used to predict the POD index of each sample in the validation set, and finally the average predicted POD index of each sample is calculated.
[0025] (8) Use the pROC tool in the R 3.3.0 package to calculate the receiver operating curve (ROC) to evaluate the microbial differentiation model. The area under the curve (AUC) is used to specify the effect value of the ROC.
[0026] (9) The differential diagnostic model achieved a differential efficacy of 93.96% between 107 esophageal cancer patients and 117 healthy controls. Figure 5 The POD index was significantly elevated in esophageal cancer patients, with a statistically significant difference between the two groups (p<0.0001). Figure 4 ).
[0027] (10) In the validation cohort, the POD index was significantly elevated in 50 patients with esophageal cancer, with a significant difference between the two groups (p<0.0001). Figure 6 This microbial differential model achieved a differential efficacy of 95.18% between 50 esophageal cancer patients and 46 healthy controls. Figure 7 ).
[0028] Therefore, the microbial gene differentiation model of the present invention has achieved good differentiation ability in esophageal cancer patients and healthy controls, proving the feasibility, applicability and universality of the microbial gene differentiation model in esophageal cancer patients. Attached Figure Description
[0029] Figure 1 A research design and clinical application of an oral microbiome model for differentiating esophageal cancer patients from healthy controls.
[0030] Figure 2 and Figure 3 The optimal oral microbial genetic markers were identified using a five-fold cross-validation method based on a random forest model.
[0031] Figure 4 In a cohort of 107 esophageal cancer patients and 117 healthy controls, the prevalence (POD) index showed a difference in expression between the two groups.
[0032] Figure 5. Differential efficacy of the microbial gene differentiation model in a cohort of 107 esophageal cancer patients and 117 healthy controls.
[0033] Figure 6. Differences in the expression of prevalence (POD) index in 50 esophageal cancer patients in the validation cohort compared with 46 healthy controls.
[0034] Figure 7. The ability of the prevalence (POD) index to differentiate between 50 patients with esophageal cancer and 46 healthy controls in the validation cohort. Detailed Implementation
[0035] The present invention will be further described below with reference to the embodiments, but the scope of protection of the present invention is not limited to these embodiments.
[0036] Unless otherwise specified, the methods used in the following examples are conventional methods. Unless otherwise specified, the materials or reagents required in the following examples are all commercially available.
[0037] This invention involves collecting tongue swab samples from enrolled subjects, extracting total microbial DNA, and performing 16S rDNA Miseq sequencing of the oral microbiota. Based on high-throughput sequencing data, a microbial differentiation model between esophageal cancer patients and healthy individuals is established in the discovery cohort, and a probability of disease (POD) index is established for esophageal cancer patients. The POD index is then used to calculate and validate its discriminative power in the validation cohort, thus achieving the universality of the microbial gene differentiation model in esophageal cancer patients.
[0038] The operating steps are as follows: (1) In accordance with the design principles of prospective clinical trials, the research design of this invention is as follows: Figure 1 As shown. The study protocol was approved by the Ethics Committee of Zhengzhou University Cancer Hospital. All enrolled patients signed informed consent forms for the study protocol and clinical sample collection.
[0039] (2) To ensure a sufficient number of effective samples, two sets of tongue coating samples were collected from each subject. To minimize the impact of factors such as diet and medication on changes in the tongue coating microecology, each subject was required to fast for at least 2 hours before sample collection and rinse their mouth with saline 2-3 times. A specially trained operator scraped the surface of the tongue coating 2-3 times with a throat swab, then immersed the tip of the swab into a test tube. All collected samples were inactivated in a 56°C water bath for at least 30 minutes and finally stored in a -80°C refrigerator. The extraction method for total oral bacterial DNA was performed according to the kit instructions.
[0040] (3) Amplification of total oral bacterial DNA samples and construction of DNA libraries were completed, and 16S rDNA sequencing was performed on the Illumina Miseq sequencing platform. All output sequences underwent basic preprocessing and basic bioinformatics analysis.
[0041] (4) The DADA2 (Divisive Amplicon Denoising Algorithm 2) algorithm was used to reduce noise in the raw sequencing data to obtain an amplicon sequence variant (ASV) table. Representative sequences of each ASV were annotated based on the SILVA reference database (SSU138.2), and then taxonomic analysis was performed to obtain information on the species composition of the samples. The ASV gene sequences of all samples were collected and organized according to the microbial gene marker discovery cohort and the microbial gene marker validation cohort.
[0042] (5) Based on representative sequences generated from high-throughput sequencing data, frequency files of ASVs in the discovery set and the validation set of microbial genetic markers were calculated. These ASVs were used in a correlation study to identify the abundance of ASVs that showed significant differences between esophageal cancer patients and healthy individuals. The Wilcoxon test was used to statistically analyze the microbial genetic markers that showed differences between esophageal cancer patients and healthy individuals. Six selected microbial genetic markers were further analyzed.
[0043] (6) In the discovery cohort of the microbial differential model, including 107 esophageal cancer patients and 117 healthy controls, microbial gene biomarkers were screened using the selected ASV abundance files in a random forest model (R software 3.4.1 and random forest package 4.6–12) with a five-fold cross-validation algorithm (software parameters were default except for setting "importance=TRUE"). Five trials of five-fold cross-validation were conducted, and the cross-validation error curve was obtained, with the smallest cross-validation error point used as the cut-off value. The minimum cross-validation error value plus the standard deviation of the corresponding value was used as the cut-off value. A set of ASV biomarkers with error rates less than the cut-off value was selected, and the set with the smallest number of ASVs was selected as the optimal set of microbial gene biomarkers. Finally, the six best microbial gene biomarkers for this model were identified. Figure 2 , Figure 3 The gene sequences of the six selected microbial OTUs markers are shown in SEQ ID NO:1-6.
[0044] (7) The probability of disease (POD) index is calculated using the ratio of randomly generated decision trees. The decision tree predicts the sample as "CP", and the predicted parameters are set as follows: probability=T, norm.votes=T, predict.all=TRUE. The random forest model built in the LOO mode is used to predict the POD index of each sample in the validation set, and finally the average predicted POD index of each sample is calculated.
[0045] (8) Use the pROC tool in the R 3.3.0 package to calculate the receiver operating curve (ROC) to evaluate the microbial differentiation model. The area under the curve (AUC) is used to specify the effect value of the ROC.
[0046] (9) The differential diagnostic model achieved a differential efficacy of 93.96% between 107 esophageal cancer patients and 117 healthy controls. Figure 5 The POD index was significantly elevated in esophageal cancer patients, with a statistically significant difference between the two groups (p<0.0001). Figure 4 ).
[0047] (10) In the validation cohort, the POD index was significantly elevated in 50 patients with esophageal cancer, with a significant difference between the two groups (p<0.0001). Figure 6 This microbial differential model achieved a differential efficacy of 95.18% between 50 esophageal cancer patients and 46 healthy controls. Figure 7 ).
[0048] Therefore, the microbial gene differentiation model of the present invention has achieved good differentiation ability in esophageal cancer patients and healthy controls, proving the feasibility, applicability and universality of the microbial gene differentiation model in esophageal cancer patients.
Claims
1. An oral microbial gene marker for distinguishing esophageal cancer patients from healthy controls, characterized in that: Composed of six microbial genomes shown in SEQ ID NO:1-6, which are enriched in the oral cavity.
2. A detection reagent for detecting the oral microbial model of claim 1, comprising primers for detecting the six microbial genes shown in SEQ ID NO:1-6 of claim 1.
3. The detection reagent according to claim 2, characterized in that: The primer sequence is SEQ ID NO:7-8.
4. The use of the detection reagent of claim 2 or 3 in the preparation of a diagnostic kit for distinguishing between esophageal cancer patients and healthy controls, wherein the detection reagent is suitable for detecting the oral microbial genes of claim 1.
5. The application according to claim 4, characterized in that: The distinguishing test kit is suitable for differentiating esophageal cancer patients from healthy controls, wherein the esophageal cancer patients are pathologically diagnosed.
6. The application according to claim 2, characterized in that: Tongue swabs from the subjects were tested to determine whether the sample contained the microbial genes and whether a microbial gene model could be established to distinguish between esophageal cancer patients and healthy controls.
7. The application according to claim 6, characterized in that: By collecting tongue swab samples from the subjects, total microbial DNA was extracted, and 16S rDNA Miseq sequencing of the microbial DNA was performed to detect the presence of the six microbial genes described in claim 1.
8. The application according to claim 7, characterized in that: By collecting tongue swab samples from enrolled subjects, total microbial DNA was extracted and 16S rDNA Miseq sequencing of the tongue flora was performed. Based on high-throughput sequencing data, a microbial differentiation model between esophageal cancer patients and healthy controls was established in the discovery cohort, and the prevalence POD index of esophageal cancer patients was established. The POD index was used to calculate and validate its discriminative ability in the validation cohort, realizing the universality of the microbial gene differentiation model in esophageal cancer patients.
9. The application according to claim 4, characterized in that, Specifically, it includes: (1) Collect tongue swab samples from the subjects. The subjects included 153 esophageal cancer patients and 167 healthy controls. The total DNA of microorganisms in the tongue swab samples was extracted according to the standard DNA extraction method. The 16S rDNA high-throughput sequencing of the tongue flora was completed on the Illumina MiSeq platform. (2) Based on high-throughput sequencing data, in the discovery cohort of the microbial differentiation model, among 107 esophageal cancer patients and 117 healthy controls, the six best microbial gene markers for the model were identified using a random forest model and a five-fold cross-validation algorithm. (3) Based on six microbial gene markers, the POD index of esophageal cancer patients was calculated by using the ratio of randomly generated decision trees; (4) In the discovery cohort, the microbial differentiation model achieved a differentiation ability of 93.96% between 107 esophageal cancer patients and 117 healthy controls. The POD index was significantly higher in esophageal cancer patients, and there was a significant difference between the two groups. (5) In the validation cohort, the microbial differentiation model achieved a differentiation ability of 95.18% between 50 esophageal cancer patients and 46 healthy controls; the POD index was significantly elevated in the diagnosed patients, and there was a significant difference between the two groups.
10. A kit for differentiating oral microbial models of esophageal cancer patients and healthy controls, comprising primers for detecting the six microbial genes shown in SEQ ID NO:1-6 as claimed in claim 1.