Use of reagents for detecting the abundance of gut microbiota markers in the preparation of an in vitro diagnostic kit for cryptosporidium infection in non-human primates

Abstract

The application discloses application of a reagent for detecting abundance of an intestinal flora marker in preparation of an in-vitro diagnostic kit for cryptosporidium infection of non-human primates, obtains bacterial abundance by performing metagenomic sequencing on bacterial components in feces of healthy non-human primates and cryptosporidium infection patients, finds 23 specific species of flora as microbial markers by screening and calculation, constructs a cryptosporidium infection risk prediction model, and optimizes the model to improve the accuracy and sensitivity of the model. Experimental results prove that the prediction model has high prediction sensitivity and good specificity, can effectively distinguish cryptosporidium infection patients from healthy individuals in various samples, and has an accuracy of more than 95%, so that the flora can be used as a cryptosporidium infection detection marker, and the reagent for detecting the abundance of the intestinal flora marker can be used in preparation of an in-vitro diagnostic kit for cryptosporidium infection of non-human primates, thereby laying a foundation for cryptosporidium infection disease research.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of medical detection, and particularly relates to application of an intestinal flora marker in preparation of an in vitro diagnostic reagent for non-human primate cryptosporidium infection. BACKGROUND

[0002] For individuals with normal immune function, cryptosporidium infection has no clinical symptoms in the early stage of infection, and obvious symptoms such as nausea, vomiting, abdominal pain and weight loss usually appear after the oocysts mature. At present, cryptosporidium infection usually adopts etiology detection, molecular detection and immunological detection.

[0003] Among them, the etiology detection has low sensitivity; the molecular detection method has high sensitivity, but has the shortcomings of time-consuming, expensive reagents and instruments; the immunological detection may have cross-reaction with other parasitic infections, which may lead to a high false positive rate, therefore, a new method with sensitivity, specificity and diagnostic value is urgently needed to be developed.

[0004] The molecular biology method based on specific intestinal microbial markers has been proved to be a sensitive, specific and efficient detection method in a variety of diseases, therefore, it is of great significance to develop microbial markers with high specificity and good sensitivity for early diagnosis of cryptosporidium infection. However, there is no related report on using microbial markers for early diagnosis of cryptosporidium infection. SUMMARY

[0005] The application aims to solve the technical problems of low detection sensitivity, false positive and high cost in the existing early diagnosis method of cryptosporidium infection, and provides application of a reagent for detecting intestinal flora marker abundance in preparation of an in vitro diagnostic kit for non-human primate cryptosporidium infection, so as to provide a new direction for the development of cryptosporidium infection diagnosis technology.

[0006] To achieve the purpose, the application adopts the following technical solutions:

[0007] The application provides application of a reagent for detecting intestinal flora marker abundance in preparation of an in vitro diagnostic kit for non-human primate cryptosporidium infection, and the intestinal flora marker comprises:

[0008] s_Clostridiaceae_bacterium_OM08_6BH,

[0009] s Clostridiaceae unciassifed SGB4769, s Clostridium fessum, s Lacrimispora amygdalina, s GGB2983 SGB3965, s GGB1244 SGB1664, s Prevotella copri clade A, s Anaerobutyricum hallii, s Prevotella SGB1680, s GGB1243 SGB1663, s GGB4780 SGB6615, s Allisonella histaminiformans, s Bacteroidaceae bacterium, s Butyricicoccus sp AM29_23AC, s Blautia wexlerae, s GGB13181 SGB20397, s Prevotella copri clade C, s Ruminococcus bicirculans, s GGB3189 SGB4212, s Blautia obeum, s Ruminococcaceae bacterium AM07_15, s Prevotella SGB1675, s Clostridium sp AF20_17LB;

[0010] As a further preferred technical solution of the present application, the screening method of the intestinal flora marker comprises the following steps:

[0011] (1) Extracting microbial DNA from fecal samples from healthy individuals and individuals infected with cryptosporidiosis;

[0012] (2) Amplifying and building a library for metagenomic sequencing using the two microbial DNAs in step (1) as templates, respectively, to calculate the abundance value of the intestinal flora at the species level;

[0013] (3) Constructing a random forest model using the abundance value of the species in step (2);

[0014] (4) Selecting the abundance value of the specific species that best predicts the number of microorganisms to input into the random forest model in step (3) for training to obtain a trained random forest model;

[0015] (5) Testing the trained random forest model in step (4), and selecting the specific species with an accuracy of not less than 85% as the microbial marker for cryptosporidiosis.

[0016] Preferably, in step (2), the abundance value calculation method is: counting the number of sequencing reads corresponding to each sequencing obtained species level genome sequence to obtain the abundance value of the species, and using the BLAST method to search for matching to the bacterial database for classification annotation.

[0017] Preferably, in step (4), the determination method of the optimal predicted number of microorganisms is a ten-fold cross-validation method, that is, the average error rate of gradually increasing the number of bacterial variables is calculated, and the number of species with the minimum error rate is the optimal predicted number of microorganisms.

[0018] Preferably, in step (4), the specific species abundance value of the optimal predicted number of microorganisms is sorted from high to low according to the importance score of the variable characteristics, and the abundance values of 23 specific species with higher importance scores are selected.

[0019] Preferably, in step (4), the training also includes the input of the optimal mtry parameter.

[0020] Preferably, the screening method of the optimal mtry parameter is a grid search using a ten-fold cross-validation method.

[0021] The above cryptosporidiosis microbial marker is used to construct a prediction model for detecting cryptosporidiosis.

[0022] Compared with the prior art, the cryptosporidiosis microbial marker has the following beneficial effects:

[0023] The present application obtains the abundance of bacteria by performing metagenomic sequencing on the bacterial components in the feces of healthy non-human primates and cryptosporidium infected individuals, finds 23 specific species of flora as microbial markers through screening operation, constructs a cryptosporidium infection risk prediction model, and optimizes the model by using data with labels to improve the accuracy and sensitivity of the model. Experiments prove that the cryptosporidium infection risk prediction model of the present application has high prediction sensitivity and good specificity, can effectively distinguish cryptosporidium infected patients from healthy individuals in multiple samples, and has an accuracy of more than 95%, so the flora can be used as a cryptosporidium infection detection marker, and a reagent for detecting the abundance of the intestinal flora marker can be used in the preparation of a non-human primate cryptosporidium infection in vitro diagnostic kit, thereby laying a foundation for cryptosporidium infection disease research. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 ANCOM analysis results of intestinal flora abundance values;

[0025] Figure 2 Cryptosporidium intestinal microbial diagnostic model based on species level constructed using random forest;

[0026] Figure 3 MDS plot of the control group and the infection group;

[0027] Figure 4 AUC curves of the training set and the test set for implementing the random forest model. DETAILED DESCRIPTION

[0028] The application will be described in detail below with reference to the accompanying drawings.

[0029] From 2021 to 2022, the applicant collected 348 fecal samples of cynomolgus monkeys from cages in Yuanjiang County, Yunnan Province. The middle layer of the feces was carefully extracted using a disposable sterile spoon, about 5 to 10 grams of fresh feces was taken, and it was placed into a sterile sampling tube containing a DNA preservation solution to obtain the sample. For each sample, detailed information such as monkey identification number, gender, age, species, sampling location, and health status was recorded in detail. After collection, the samples were transported to the laboratory with dry ice and stored at -80°C until further analysis. The fecal DNA was extracted using the E.Z.N.A. ® Fecal DNA Kit (D4015-00, OMEGA, USA). According to the primer sequence and amplification scheme of SSU rRNA, the nested PCR method was used to amplify the cryptosporidium DNA from all fecal DNA samples. In addition, in order to ensure that the difference in the flora results is caused by cryptosporidium, the applicant implemented a PCR amplification-based method to exclude the infection of other common parasites. Five kinds of parasites (Trichomonas, Enterocytozoon bieneusi, Cyclospora, Giardia and nematode) were screened out using qPCR or PCR method. Samples containing these listed pathogens from two groups were excluded for further analysis.

[0030] In addition, the applicant statistically analyzed the gender, age, and BMI of the same area cynomolgus monkey cryptosporidium infection group and non-infection group, and the results are as follows.

[0031]

[0032] The results show that there is no significant difference in gender, age, and BMI between the same area cynomolgus monkey cryptosporidium infection group and non-infection group, which indicates that the difference in the flora of the two groups that follows will not be the result of the influence of these factors.

[0033] In the present application, metagenomic sequencing is performed using an Illumina NovaSeq X high-throughput sequencing platform.

[0034] The reagent for detecting the abundance of the intestinal flora marker provided by the present application is used for preparing a non-human primate cryptosporidium infection in vitro diagnostic kit, and the intestinal flora marker comprises:

[0035] s_Clostridiaceae_bacterium_OM08_6BH,

[0036] s_Clostridiaceae_unciassifed_SGB4769, s_Clostridium_fessum, s_Lacrimispora_amygdalina, s_GGB2983_SGB3965, s_GGB1244_SGB1664, s_Prevotella_copri_clade_A, s_Anaerobutyricum_hallii, s_Prevotella_SGB1680, s_GGB1243_SGB1663, s_GGB4780_SGB6615, s_Allisonella_histaminiformans, s_Bacteroidaceae_bacterium, s_Butyricicoccus_sp_AM29_23AC, s_Blautia_wexlerae, s_GGB13181_SGB20397, s_Prevotella_copri_clade_C, s_Ruminococcus_bicirculans, s_GGB3189_SGB4212, s_Blautia_obeum, s_Ruminococcaceae_bacterium_AM07_15, s_Prevotella_SGB1675, s_Clostridium_sp_AF20_17LB.

[0037] The screening method of upper intestinal flora markers comprises the following steps:

[0038] (1) Extracting microbial DNA from fecal samples from healthy individuals and individuals infected with cryptosporidiosis.

[0039] (2) Using the two kinds of microbial DNA in step (1) as templates, respectively, amplifying and building a library for metagenomic sequencing, counting the number of sequencing reads corresponding to the species-level genomic sequences obtained by each sequencing, obtaining the abundance value of the species-level intestinal flora, and using the BLAST method to search for matching to the bacterial database for classification annotation.

[0040] Figure 1 The results of using ANCOM for analysis show that, at the species level, there are 18 bacteria that are significantly different between the infection group and the control group, of which 16 have increased abundance after infection and 2 have decreased abundance after infection. Among them, the bacteria with increased abundance at the annotation level are mostly Clostridium.

[0041] (3) constructing a random forest model with the abundance values of the species in step (2).

[0042] In the present application, it is preferred to use machine learning to select a model algorithm. The random forest model uses a plurality of decision tree classifiers, and the output category is determined by the mode of the categories output by the individual trees.

[0043] Figure 2 A species-level cryptosporidium intestinal microorganism diagnostic model based on a random forest was constructed, which contained 23 bacteria with different relative abundances between the control group and the infected group, and it was shown that the increase of clostridium was indeed an important feature after cryptosporidium infection.

[0044] (4) According to the ten-fold cross-validation method, the average error rate of gradually increasing the number of bacterial variables is calculated, and the number of species with the minimum error rate is 23, which is the optimal prediction microorganism number; according to the importance score of the variable characteristics from high to low, the abundance values of the 23 specific species of the specific species of bacteria with a Mean Decrease Accuracy value greater than 4 are input into the random forest model in step (3), and then the ten-fold cross-validation method is used for grid search to obtain the optimal mtry parameter, in the present application, the optimal mtry parameter is preferably 2, and then the optimal mtry parameter is input into the model for training to obtain the trained random forest model.

[0045] (5) The trained random forest model in step (4) is used for testing, and the specific species of the optimal prediction microorganism number with an accuracy of not less than 85% is selected as the microorganism marker of cryptosporidium infection.

[0046] Figure 3 The MDS plot of the control group and the infected group shows that the random forest model can distinguish the control group and the infected group on the test set, which indicates that the random forest model has a certain effectiveness in distinguishing the two groups of samples.

[0047] Figure 4 The AUC curve of the training set and the test set of the random forest model is shown. The AUC curve analysis result shows that the AUC of the model in the training set (n = 20) is 1, and the AUC of the model in the test set (n = 20) is 0.95, indicating that the model has good test accuracy, which shows that the intestinal flora marker provided by the present application provides an effective non-invasive method for monitoring cryptosporidium infection. The reagent for detecting the abundance of the intestinal flora marker can be used in the preparation of a non-human primate cryptosporidium infection in vitro diagnostic kit, thereby laying a foundation for the study of cryptosporidium infection diseases.

Claims

1. The application of reagents for detecting the abundance of intestinal flora markers in the preparation of an in vitro diagnostic kit for cryptosporidiosis in cynomolgus monkeys, characterized in that, Manufacturer specific products_Clostridiaceae_bacterium_OM08_6BHs_Clostridiaceae_unciassifed_SGB4769. s_Clostridium_fessum、s_Lacrimispora_amygdalina、s_GGB2983_SGB3965、s_GGB1244_SGB1664、s_Prevotella_copri_clade_A、s_Anaerobutyr icum_hallii、s_Prevotella_SGB1680、s_GGB1243_SGB1663、s_GGB4780_SGB6615、s_Allisonella_histaminiformans、s_Bacteroidaceae_bacter ium、s_Butyricicoccus_sp_AM29_23AC、s_Blautia_wexlerae、s_GGB13181_SGB20397、s_Prevotella_copri_clade_C、s_Ruminococcus_bicircle ans_s_GGB3189_SGB4212 s_Blautia_obeum s_Ruminococcaceae_bacterium_AM07_15 s_Prevotella_SGB1675 s_Clostridium_sp_AF20_17LB strains.

Images