Non-alcoholic fatty liver intestinal microbial marker, application thereof and method for constructing non-alcoholic fatty liver detection model

Abstract

The invention discloses a non-alcoholic fatty liver intestinal microbial marker, application of the non-alcoholic fatty liver intestinal microbial marker and a method for constructing a non-alcoholic fatty liver detection model, and belongs to the technical field of biomedicine. The method comprises the following steps: extracting intestinal microorganisms related to the non-alcoholic fatty liver disease, establishing a random forest model for diagnosing the non-alcoholic fatty liver disease through a non-alcoholic fatty liver disease intestinal microorganism marker, and evaluating to obtain a non-alcoholic fatty liver disease detection model; according to the method provided by the invention, differential bacteria of the non-alcoholic fatty liver patient and a healthy control can be identified more accurately, so that the accuracy of early auxiliary diagnosis of the non-alcoholic fatty liver is improved, personalized treatment can be carried out on the non-alcoholic fatty liver patient through the ROC curve and the marker bacteria, and the accuracy of early auxiliary diagnosis of the non-alcoholic fatty liver is improved. And expensive medical tests and unnecessary medical intervention can be reduced, so that medical resources are saved.

Description

Technical Field

[0001] The present invention relates to the field of biomedical technology, and more specifically to a non-alcoholic fatty liver disease intestinal microbial marker and its application and a method for constructing a non-alcoholic fatty liver disease detection model. Background Art

[0002] The incidence of nonalcoholic fatty liver disease (NAFLD) is on the rise worldwide, and the prevalence varies in different regions and populations. In developed countries and regions, such as Europe and the United States, the prevalence is high; in developing countries, with changes in lifestyle and the prevalence of obesity, the prevalence is also gradually increasing. Nonalcoholic fatty liver disease (NAFLD) can be seen in all age groups, but it is more common in middle-aged people, and the prevalence in men is slightly higher than that in women. The current diagnosis mainly relies on imaging examinations (such as ultrasound, magnetic resonance imaging, etc.) and blood biochemical indicators (such as alanine aminotransferase, aspartate aminotransferase, blood lipids, etc.), but these methods are not sensitive and specific enough in the early stages of the disease; although liver biopsy is the gold standard for diagnosis, it is an invasive examination with certain risks and low patient acceptance.

[0003] Studies have found that the intestinal microbial diversity of patients with non-alcoholic fatty liver disease may be reduced, the number of certain beneficial bacteria is reduced, and the number of harmful bacteria or conditional pathogens is increased; specific intestinal microbial species, flora ratios or their metabolites are expected to become new diagnostic markers for non-alcoholic fatty liver disease. The present invention aims to identify the differential microbial communities in fecal samples of patients with alcoholic fatty liver disease and normal people through metagenomic sequencing technology, and verify these differential microorganisms using a random forest model, in order to discover new biomarkers or therapeutic targets, and improve the early diagnosis rate and treatment effect of NAFLD. Summary of the invention

[0004] The purpose of the present invention is to provide a non-alcoholic fatty liver disease intestinal microbial marker and its application and a method for constructing a non-alcoholic fatty liver disease detection model, by analyzing the composition and functional changes of the intestinal microorganisms of non-alcoholic fatty liver disease patients, the condition of non-alcoholic fatty liver disease can be discovered earlier and the treatment effect can be monitored.

[0005] To achieve the above object, the present invention provides the following technical solutions: The present invention provides a non-alcoholic fatty liver disease intestinal microbial marker, wherein the intestinal microbial marker comprises the following bacterial species: Actinomyces_dentalis, Bifidobacterium_longum, Faecalibacillus_intestinalis, Anaerobutyricum_hallii, Blautia_wexlerae, Anaerostipes_hadrus, Bifidobacterium_pseudocatenulatum, and Dorea_longicatena.

[0006] Furthermore, the application includes products for diagnosing whether a patient has non-alcoholic fatty liver disease or non-alcoholic fatty liver disease, as a target for the assessment and treatment of non-alcoholic fatty liver disease, and for constructing a non-alcoholic fatty liver disease detection model; the product includes any one of a detection reagent, a detection kit, and a detection chip.

[0007] The present invention also provides a method for constructing a non-alcoholic fatty liver disease detection model using non-alcoholic fatty liver disease intestinal microbial markers, comprising the following steps: (1) Total DNA was extracted from the feces of patients with non-alcoholic fatty liver disease and healthy people, and then sequenced after pretreatment. The sequencing results were subjected to metagenomic assembly, gene prediction and gene set construction, species composition and abundance, gene function analysis, diversity analysis, and multi-sample comparative analysis to screen out intestinal microbial markers for non-alcoholic fatty liver disease; (2) Establish a random forest model to identify NAFLD and healthy samples using NAFLD intestinal microbial markers, evaluate and obtain a NAFLD detection model; (3) The metagenome and abundance information of the patient to be tested are input into the non-alcoholic fatty liver disease detection model for identification.

[0008] Furthermore, in the step (1), the pretreatment includes: library preparation, and the library preparation is specifically: The DNA fragments are randomly broken into specific lengths, the ends of the fragmented DNA are repaired, the two ends of the repaired DNA fragments are connected to specific adapter sequences, and the excess adapters and primers are removed, and the specific adapter sequence is amplified by PCR.

[0009] Furthermore, in step (1), the sequencing includes: removing adapters using fastp, removing host sequences using bowtie2, performing marker-based species abundance analysis using MetaPhlAn4, and performing functional analysis using HUMAnN33.

[0010] Furthermore, in step (2), a random forest model for identifying samples of non-alcoholic fatty liver disease and healthy people is established by using non-alcoholic fatty liver disease intestinal microbial markers, and the non-alcoholic fatty liver disease detection model is obtained by evaluation, specifically: The non-alcoholic fatty liver patients and healthy people were divided into a training set and a test set, and the random forest model was trained on the training set using the sklearn.ensemble.RandomForestClassifier module in python; The evaluation method is: evaluating the fitting degree of the random forest model by drawing a learning curve, and verifying the random forest model by ten-fold cross validation.

[0011] Furthermore, the training parameters of the random forest model in step (2) are: gradient n_estimators={100,500,1000}, max_depth={1, 2, 3, 7, 9}, max_features=["log2", "sqrt"], and the python software GridSearchCV module is used to search for the optimal parameters. The random forest model is rebuilt using the optimal parameters and prediction is performed on the test set.

[0012] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects: (1) The present invention extracts biomarkers related to the pathogenesis and progression of non-alcoholic fatty liver disease from metagenomic data; the random forest algorithm is used to construct a non-alcoholic fatty liver disease detection model, which can increase the generalization ability and anti-overfitting ability of the model, and can also more accurately predict and evaluate the condition of diabetes, which helps doctors develop more effective treatment plans and improve patients' survival rate and quality of life.

[0013] (2) The present invention obtains the error curve CVerror through the random forest model and five tests of 10-fold cross validation, which can more accurately identify the differential bacteria between patients with non-alcoholic fatty liver disease and healthy controls, thereby improving the accuracy of early auxiliary diagnosis of non-alcoholic fatty liver disease. In addition, through the ROC curve and marker bacteria, personalized treatment can be performed for patients with non-alcoholic fatty liver disease to improve the treatment effect. The model of the present invention has been verified and applied in clinical practice, which can reduce expensive medical tests and unnecessary medical interventions, thereby saving medical resources.

[0014] (3) The present invention proposes a technical solution for identifying specific intestinal microbial markers of non-alcoholic fatty liver disease using methods such as a random forest model, which can be used to reveal changes in host-microbial metabolic pathways associated with non-alcoholic fatty liver disease and may help discover new biomarkers, and may discover new biomarkers or therapeutic targets. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0016] The following is a further description of the non-alcoholic fatty liver disease intestinal microbial markers and their applications and the method for constructing a non-alcoholic fatty liver disease detection model in the present invention in conjunction with the accompanying drawings; Figure 1 Obtaining a metagenomic flow chart for Example 1; Figure 2 This is a graph of the error results of the cross validation performed ten times in Example 1; Figure 3 It is the "Mean Decrease Accuracy" indicator in the random forest model of Example 1; Figure 4 This is the ROC curve of the random forest model established based on the non-alcoholic fatty liver disease intestinal microbial markers in Example 1. DETAILED DESCRIPTION

[0017] Various exemplary embodiments of the present invention will now be described in detail. This detailed description should not be considered as limiting the present invention, but should be understood as a more detailed description of certain aspects, features, and embodiments of the present invention.

[0018] It should be understood that the terms described in the present invention are only for describing special embodiments and are not intended to limit the present invention. In addition, for the numerical range in the present invention, it should be understood that each intermediate value between the upper and lower limits of the scope is also specifically disclosed. Each smaller range between the intermediate value in any stated value or stated range and any other stated value or intermediate value in the described range is also included in the present invention. The upper and lower limits of these smaller ranges can be independently included or excluded in the scope.

[0019] Unless otherwise indicated, all technical and scientific terms used herein have the same meanings as those generally understood by those skilled in the art. Although the present invention describes only preferred methods and materials, any methods and materials similar or equivalent to those described herein may also be used in the implementation or testing of the present invention. All documents mentioned in this specification are incorporated by reference to disclose and describe the methods and / or materials associated with the documents. In the event of a conflict with any incorporated document, the content of this specification shall prevail.

[0020] It will be apparent to those skilled in the art that various modifications and variations may be made to the specific embodiments of the present invention description without departing from the scope or spirit of the present invention. Other embodiments derived from the present invention description will be apparent to the skilled artisan. The present invention description and examples are exemplary only.

[0021] The words “include,” “including,” “have,” “contain,” etc. used in this document are open-ended terms, meaning including but not limited to.

[0022] The present invention provides a method for constructing a non-alcoholic fatty liver disease detection model using non-alcoholic fatty liver disease intestinal microbial markers, comprising the following steps: (1) Total DNA was extracted from the feces of patients with non-alcoholic fatty liver disease and healthy people, and then sequenced after pretreatment. The sequencing results were subjected to metagenomic assembly, gene prediction and gene set construction, species composition and abundance, gene function analysis, diversity analysis, and multi-sample comparative analysis to screen out intestinal microbial markers for non-alcoholic fatty liver disease; (2) Establish a random forest model to identify NAFLD and healthy samples using NAFLD intestinal microbial markers, evaluate and obtain a NAFLD detection model; (3) The metagenome and abundance information of the patient to be tested are input into the non-alcoholic fatty liver disease detection model for identification.

[0023] In the step (1), the pretreatment includes: library preparation, and the library preparation specifically includes: The DNA fragments are randomly broken into specific lengths, the ends of the fragmented DNA are repaired, the two ends of the repaired DNA fragments are connected to specific adapter sequences, and the excess adapters and primers are removed, and the specific adapter sequence is amplified by PCR.

[0024] In the step (1), the sequencing includes: removing the adapter using fastp, removing the host sequence using bowtie2, performing marker-based species abundance analysis using MetaPhlAn4, and performing functional analysis using HUMAnN33.

[0025] In the step (2), a random forest model for identifying samples of non-alcoholic fatty liver disease and healthy people is established by using non-alcoholic fatty liver disease intestinal microbial markers, and an evaluation is performed to obtain a non-alcoholic fatty liver disease detection model, specifically: The non-alcoholic fatty liver patients and healthy people were divided into a training set and a test set, and the random forest model was trained on the training set using the sklearn.ensemble.RandomForestClassifier module in python.

[0026] The evaluation method is: evaluating the fitting degree of the random forest model by drawing a learning curve, and verifying the random forest model by ten-fold cross validation.

[0027] The training parameters of the random forest model in step (2) are: gradient n_estimators={100, 500, 1000}, max_depth={1, 2, 3, 7, 9}, max_features=["log2", "sqrt"]. The python software GridSearchCV module is used to search for the optimal parameters, and the random forest model is rebuilt using the optimal parameters and predicted on the test set.

[0028] The present invention provides a non-alcoholic fatty liver disease intestinal microbial marker, wherein the intestinal microbial marker comprises the following bacterial species: Actinomyces_dentalis, Bifidobacterium_longum, Faecalibacillus_intestinalis, Anaerobutyricum_hallii, Blautia_wexlerae, Anaerostipes_hadrus, Bifidobacterium_pseudocatenulatum, and Dorea_longicatena.

[0029] Example 1 Screening of intestinal microbial markers for nonalcoholic fatty liver disease like Figure 1As shown, stool samples from 13 patients with non-alcoholic fatty liver disease and 227 healthy controls were collected from the clinic. The sampling process ensured that the patients were not receiving antibiotic treatment or other related treatments that may affect the intestinal flora. Use sterile containers to collect stool samples to avoid cross-contamination; when taking stool samples from patients with alcoholic fatty liver disease and normal controls, empty the urine first to avoid urine contamination. Use a sterile spoon to take the internal stool sample, and then store the sample in a special stool box at -80°C; then use a DNA extraction kit to extract total DNA from the stool sample, remove proteins and other impurities, and obtain pure DNA. After adding adapter sequences to both ends of the interrupted DNA fragments, the Illumina platform performs high-throughput sequencing, including fastp to remove the linker, bowtie2 to remove the host sequence, MetaPhlAn4 to perform marker-based species abundance analysis, and HUMAnN33 and other tools for functional analysis. Finally, the differential analysis step is performed by Lefse analysis, comparison of metabolic species differences, multivariate analysis of variance (such as adonis and MAdLIN), and random forest methods (such as Figure 1 ), biomarkers of 8 bacterial genera that changed significantly between patients with non-alcoholic fatty liver disease and healthy people, including the following genera: Actinomyces_dentalis, Bifidobacterium_longum, Faecalibacillus_intestinalis, Anaerobutyricum_hallii, Blautia_wexlerae, Anaerostipes_hadrus, Bifidobacterium_pseudocatenulatum, and Dorea_longicatena.

[0030] In the above process, adapter sequences are added to both ends of the interrupted DNA fragments, specifically: in order to obtain high-quality and representative microbial community DNA, Qiagen's DNeasy PowerSoil Kit is used to extract DNA; after extraction, the high molecular weight DNA is randomly interrupted into fragments of a specific length, generally about 300 bp, using an ultrasonic disruptor; the ends of the fragmented DNA are repaired to make them blunt-ended; specific adapter sequences are connected to both ends of the repaired DNA fragments, and excess adapters, primers and other impurities are removed by methods such as magnetic beads or column purification; specific primers are used to perform PCR amplification on the DNA fragments with adapters to enrich a sufficient amount of library fragments.

[0031] In the above, sequencing was performed by Illumina, and the raw data obtained were subjected to quality control and data filtering, metagenomic assembly, gene prediction, construction of reference gene sets, and subsequent species, gene, and functional analysis, specifically: fastp software was used to perform quality control on the data from the sequencing machine, including the removal of adapters and low-quality sequences; Bowtie2 was used to filter out reads that may originate from the host; MetaPhlAn4 software was used to perform species annotation on the metagenomic sequencing data, and the composition of the microbial community was analyzed for the metagenomic data from the sequencing machine to obtain relative abundance information; HUMAnN3 was used to identify the species-level functional composition of known species in the population by comparing the pan-genome through a tiered retrieval strategy, complete the quantitative gene family and pathway abundance, and quickly obtain diverse species and functional composition; Bowtie2 and other statistical analysis software models (GMM) were used to compare, identify, and analyze metagenomic data.

[0032] Example 2 The ability of establishing a non-alcoholic fatty liver disease detection model using intestinal microbial markers to distinguish between non-alcoholic fatty liver disease patients and healthy controls Based on the metagenomic data of clinical non-alcoholic fatty liver disease patients and healthy people samples, features related to non-alcoholic fatty liver disease were extracted from the data, including differential microbial species abundance, metabolic pathway activity, etc., and the random forest algorithm was used to build a model. In the process of building a decision tree, the samples of non-alcoholic fatty liver disease patients and healthy people were divided into training sets and test sets. The training set was trained with a random forest model using the python software sklearn. Ensemble. RandomForestClassifier module. The fitting of the model was evaluated by drawing a learning curve. The stability and reliability of the model were further verified by 10-fold cross validation. The model training parameters were set as follows: gradient n_estimators = [100, 500, 1000]; max_depth = [1, 2, 3, 7, 9]; max_features = ["log2", "sqrt"]; the method for training the random forest model with ten cross validations was to use the python software GridSearchCV module to search for the best parameters, use the best parameters to reconstruct the random forest model, obtain the non-alcoholic fatty liver disease detection model, and make predictions on the test set.

[0033] The cross-validation test with ten runs provides a comprehensive view of the model stability and reliability. Figure 2 As shown in the figure, the vertical axis reflects the value of the cross-validation error, which ranges roughly from 0.10 to 0.20. There are several curves in the line chart, which represent different variable selection methods. Figure 2 The black curve in the middle clearly shows the trend of cross-validation error as the number of variables increases. This curve initially shows a downward trend, indicating that the prediction error of the model is gradually decreasing in the process of adding variables. Both the model training set and the test set have high accuracy, indicating that the model has good results. Among them, Actinomyces_dentalis, Bifidobacterium_longum, Faecalibacillus_intestinalis, Anaerobutyricum_hallii, Blautia_wexlerae, Anaerostipes_hadrus, Bifidobacterium_pseudocatenulatum, and Dorea_longicatena have the best retention performance. As shown in Table 1, "Mean Decrease Accuracy" and "Mean Decrease Gini" are two important indicators in the random forest model. "Mean Decrease Gini" calculates the impact of each variable on the heterogeneity of the observed values ​​at each node of the classification tree, thereby comparing the importance of the variable. The larger the value, the greater the importance of the variable. "Mean Decrease Accuracy" indicates the degree of decrease in the prediction accuracy of random forest. The larger the value, the greater the importance of the variable ( Figure 3 ). The horizontal axis (X-axis) of the ROC curve represents 1-Specificity, that is, the false positive rate, which ranges from 0 to 1, reflecting the proportion of negative examples that the model mistakenly predicts as positive examples, and the vertical axis (Y-axis) represents Sensitivity, that is, the true positive rate, which measures the model's ability to correctly identify positive examples. Figure 4 In the figure, the red solid line clearly depicts the ROC curve of the model, and the shaded part represents the confidence interval (CI). The ROC curve in the present invention is close to the upper left corner, and has a higher true positive rate at each threshold. The area under the curve (AUC) is close to 1, indicating that the classification performance of the model is better. The AUC value of the present invention reaches 0.9632, indicating that the classifier performs well in distinguishing positive and negative samples. The CI ranges from 0.9487 to 0.9777, which provides a reliable evaluation of the AUC value, indicating the stability of the model performance.

[0034]

[0035] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A non-alcoholic fatty liver disease intestinal microbial marker, characterized in that: The intestinal microbial markers include the following species: Actinomyces_dentalis, Bifidobacterium_longum, Faecalibacillus_intestinalis, Anaerobutyricum_hallii, Blautia_wexlerae, Anaerostipes_hadrus, Bifidobacterium_pseudocatenulatum, and Dorea_longicatena.

2. A use of the intestinal microbial marker for non-alcoholic fatty liver disease according to claim 1, characterized in that: The application includes products for diagnosing whether a patient has non-alcoholic fatty liver disease or non-alcoholic fatty liver disease, as a target for the assessment and treatment of non-alcoholic fatty liver disease, and for constructing a non-alcoholic fatty liver disease detection model; the product includes any one of a detection reagent, a detection kit, and a detection chip.

3. A method for constructing a non-alcoholic fatty liver disease detection model using the non-alcoholic fatty liver disease intestinal microbial markers according to claim 1, characterized in that: The following steps are involved: (1) Total DNA was extracted from the feces of patients with non-alcoholic fatty liver disease and healthy people, and then sequenced after pretreatment. The sequencing results were subjected to metagenomic assembly, gene prediction and gene set construction, species composition and abundance, gene function analysis, diversity analysis, and multi-sample comparative analysis to screen out intestinal microbial markers for non-alcoholic fatty liver disease; (2) Establish a random forest model to identify NAFLD and healthy samples using NAFLD intestinal microbial markers, evaluate and obtain a NAFLD detection model; (3) The metagenome and abundance information of the patient to be tested are input into the non-alcoholic fatty liver disease detection model for identification.

4. The method for constructing a non-alcoholic fatty liver disease detection model according to claim 3, characterized in that: In the step (1), the pretreatment includes: library preparation, and the library preparation specifically includes: The DNA fragments are randomly broken into specific lengths, the ends of the fragmented DNA are repaired, the two ends of the repaired DNA fragments are connected to specific adapter sequences, and the excess adapters and primers are removed, and the specific adapter sequence is amplified by PCR.

5. The method for constructing a non-alcoholic fatty liver disease detection model according to claim 3, characterized in that: In the step (1), the sequencing includes: removing the adapter using fastp, removing the host sequence using bowtie2, performing marker-based species abundance analysis using MetaPhlAn4, and performing functional analysis using HUMAnN33.

6. The method for constructing a non-alcoholic fatty liver disease detection model according to claim 3, characterized in that: In the step (2), a random forest model for identifying samples of non-alcoholic fatty liver disease and healthy people is established by using non-alcoholic fatty liver disease intestinal microbial markers, and an evaluation is performed to obtain a non-alcoholic fatty liver disease detection model, specifically: The non-alcoholic fatty liver patients and healthy people were divided into a training set and a test set, and the random forest model was trained on the training set using the sklearn.ensemble.RandomForestClassifier module in python; The evaluation method is: evaluating the fitting degree of the random forest model by drawing a learning curve, and verifying the random forest model by ten-fold cross validation.

7. The method for constructing a Crohn's disease detection model according to claim 6, characterized in that: The training parameters of the random forest model in step (2) are: gradient n_estimators={100, 500, 1000}, max_depth={1, 2,3, 7, 9}, max_features=["log2", "sqrt"]. The python software GridSearchCV module is used to search for the optimal parameters, and the random forest model is rebuilt using the optimal parameters and predicted on the test set.

Images