A combination and screening method of microbial markers for predicting BMI in saliva samples

Abstract

The present invention discloses a microbial marker combination and screening method for predicting BMI of saliva samples, and relates to the field of biotechnology. The microbial marker combination includes 16 genus-level microorganisms. The present invention takes saliva microorganisms as the research object, and based on the 16S rRNA metagenomic sequencing data of individuals with different BMI, combines a machine learning algorithm to systematically identify microbial markers related to BMI; based on the selected microbial markers related to BMI, a machine learning model is constructed for the characterization of BMI. Compared with DNA genetic markers, the microbial marker combination used in the present invention has the advantage of rich content, which is more conducive to the detection and analysis of trace samples. In addition, the cell wall structure of the microorganism also makes it have strong resistance to the external environment, which is more suitable for the analysis of old and degraded samples.

Description

Technical Field

[0001] The present invention relates to the field of biotechnology, and in particular to a microbial marker combination and screening method for predicting BMI of saliva samples. Background Art

[0002] Individual identification is a major research task in forensic medicine. DNA, as the "blueprint" of biological development and operation, stores all the genetic information of organisms. Forensic scientists usually use DNA genetic markers to track and find people involved in the case. Conventional DNA genetic marker testing mainly involves extracting DNA samples from biological samples (such as hair, blood, semen, saliva, etc.) left at the scene, and performing DNA genetic marker typing to obtain map information; by comparing with reference DNA maps or forensic DNA databases, the identity information of relevant persons can be determined.

[0003] Forensic phenotyping is the process of testing biological samples for DNA genetic markers, and inferring possible phenotypic characteristics of the individual to be tested, such as appearance, skin color, and height, based on the obtained DNA typing. In addition to some common phenotypic characteristics, body shape, as the most intuitive phenotype, has also been the focus of forensic scientists in recent years. For body shape, researchers usually use body mass index (BMI) for assessment. Previous whole-genome association studies have identified a large number of gene loci associated with BMI. However, biological samples extracted from the crime scene have characteristics such as trace amounts and degradation, making it difficult to use a large number of DNA genetic markers to infer BMI. Therefore, for BMI, it is urgent to select new molecular genetic markers for inferring BMI.

[0004] Microorganisms have the advantages of rich species, wide distribution and huge numbers. Even trace samples can contain a large number of microorganisms, which has very good application advantages in forensic research, especially for those trace samples. In addition, the structure of the microbial cell wall makes it highly resistant to the external environment and not easily degraded, which is more suitable for the analysis and detection of old and degraded samples. Previous studies have found that there are certain differences in the composition of microbial communities between individuals with different BMI. Therefore, based on microbiome data, microbial markers related to BMI can be systematically identified for the characterization of individual BMI.

[0005] Saliva is a common type of biological specimen, so it is very important to characterize its molecular characteristics. More importantly, humans can secrete about 1 liter of saliva every day, which contains a large number of microorganisms. In recent years, forensic scientists have begun to conduct a series of studies based on salivary microorganisms, such as speculating on their possible biogeographical origins and exposure time in the external environment. Although salivary microorganisms have a high application prospect in forensic research, forensic scientists are currently mainly concerned with body fluid identification and retention time, and there are relatively few studies on BMI. Summary of the invention

[0006] The purpose of the present invention is to provide a microbial marker combination and screening method for predicting BMI of saliva samples to solve the problems existing in the above-mentioned prior art. The microbial marker combination provided by the present invention can realize the characterization research of BMI with high accuracy.

[0007] To achieve the above object, the present invention provides the following solutions:

[0008] The present invention provides a microbial marker combination for predicting BMI of saliva samples, comprising the following genus-level microorganisms: g__Achromobacter, g__Aerococcus, g__Alloprevotella, g__Amnipila, g__Blastomonas, g__Campylobacter, g__Catonella, g__Enhydrobacter, g__Mobiluncus, g__Oribacterium, g__Prevotella, g__Pseudolabrys, g__Selenomonas, g__unclassified_f__Pasteurellaceae, g__unclassified_o__Bacteroidales and g__Weissella.

[0009] The present invention also provides the use of the above-mentioned microbial marker combination in predicting BMI based on saliva samples.

[0010] The present invention also provides the use of a reagent for detecting the above-mentioned microbial marker combination in preparing a product for predicting BMI based on a saliva sample.

[0011] Furthermore, the product is a reagent, a kit or a chip.

[0012] The present invention also provides a product for predicting BMI based on saliva samples, including a reagent for detecting the above-mentioned microbial marker combination.

[0013] The present invention also provides the use of the above product in predicting BMI based on saliva samples.

[0014] The present invention also provides a method for constructing a machine learning model for predicting BMI based on saliva samples, comprising the following steps:

[0015] Extracting genomic DNA from a saliva sample, and constructing a DNA library using the primers shown in SEQ ID NO.1-2;

[0016] The DNA library is sequenced to obtain sequencing data, and the sequences are divided into operational classification units using UPARSE7.1 software with a sequence similarity of 97% as a standard; the microorganisms in the SILVAv138 database are compared using an RDP classifier to obtain microbial annotation information of each sample at different classification levels;

[0017] The LEfSe software system was used to identify microbial marker combinations with significant differences in distribution among different BMI groups A;

[0018] The microeco package in R software was used to identify the microbial marker combination B associated with BMI;

[0019] Using the overlapping microbial markers in the microbial marker combination A and the microbial marker combination B as the microbial marker combination for predicting BMI of saliva samples;

[0020] Adaptive synthesis algorithm SMOTE-NC was used to simulate samples in a ratio of 1:1:1 between groups to solve the problem of unbalanced sample sizes between different groups and obtain a data set; the data set was randomly divided into training set samples and test set samples in a ratio of 7:3;

[0021] Using the training set samples to construct a machine learning model for BMI prediction;

[0022] The caret package in the R software is used to optimize the parameters of the machine learning model using a ten-fold cross validation method to determine an optimal prediction model;

[0023] The test set samples are used to verify the performance of the prediction model, and the evaluation model for predicting BMI based on saliva samples is obtained.

[0024] The present invention also provides a machine learning model for predicting BMI based on saliva samples, which is constructed according to the above-mentioned construction method.

[0025] The present invention also provides a method for predicting BMI based on a saliva sample, comprising the following steps:

[0026] The genomic DNA of the saliva sample to be tested was extracted. First, PCR amplification was performed using primers in the 16S rRNA V3-V4 region, and the amplified products were sequenced to obtain the 16S rRNA metagenomic data of the sample to be tested. Next, the obtained data was subjected to OTU classification, and the above-mentioned 16 microbial markers in the OTU classification results were input into the above-mentioned machine learning model to obtain the prediction results of the BMI of the sample to be tested.

[0027] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program is used to execute the above-mentioned method for predicting BMI based on a saliva sample.

[0028] The present invention discloses the following technical effects:

[0029] The present invention uses salivary microorganisms as the research object, and based on the 16S rRNA metagenomic sequencing data of individuals with different BMI, combines machine learning algorithms to systematically identify microbial markers related to BMI; based on the selected microbial markers related to BMI, a machine learning model is constructed for the characterization of BMI.

[0030] Compared with DNA genetic markers, the microbial marker combination used in the present invention has the advantage of rich content, which is more conducive to the detection and analysis of trace samples. In addition, the cell wall structure of microorganisms also makes them more resistant to the external environment, which is more suitable for the analysis of old and degraded samples.

[0031] The present invention can achieve the characterization study of BMI with only 16 microbial markers and has high accuracy. Compared with large-scale whole genome sequencing research, the present invention is easier to operate and has lower cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] 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 will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0033] Figure 1 It is the technical roadmap of the present invention;

[0034] Figure 2 Heat map of microbial community distribution at the phylum level for different BMI groups;

[0035] Figure 3 Heat map of microbial community distribution at the genus level in different BMI groups;

[0036] Figure 4 It is the linear discriminant analysis diagram of species partial least squares method at the phylum level;

[0037] Figure 5 It is the linear discriminant analysis diagram of species partial least squares method at the genus level;

[0038] Figure 6 These are 16 microbial markers at the genus level selected using the LEfSe software system;

[0039] Figure 7 The microeco package in R software was used to select 17 microbial markers at the genus level;

[0040] Figure 8 This is a heat map showing the distribution of 17 microbial markers among different BMI groups based on the microeco package in R software;

[0041] Fig. 9 The prediction results of random forest linear regression of 16 microbial markers selected by LEfSe software system and microeco package;

[0042] Fig.10 The random forest classification prediction results of 16 microbial markers selected by the LEfSe software system and the microeco package; DETAILED DESCRIPTION

[0043] 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.

[0044] It should be understood that the terms described in the present invention are only for describing a particular embodiment 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. The intermediate value in any stated value or stated range, and each smaller range between 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.

[0045] 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.

[0046] 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 those skilled in the art. The present invention description and examples are exemplary only.

[0047] The words “include,” “including,” “have,” “contain,” etc. used in this document are open-ended terms, meaning including but not limited to.

[0048] The reagents used in the present invention include MagAtrract PowerSoil Pro DNA extraction kit, first-round PCR reagents (5×TransStart FastPfu buffer, 2.5 mM dNTPs, 16S rRNA V3-V4 primer pair (5 μM), TransStart FastPfu DNA polymerase and ddH2O), AxyPrep DNA Gel Extraction kit and NEXTFLEX Rapid DNA-Seq kit.

[0049] The software involved in the present invention includes UPARSE 7.1, RDP classifier, LEfSe and microeco, randomforest, caret and themis packages in R software.

[0050] The technical roadmap of the present invention is as follows Figure 1 As shown, the specific operations are as follows:

[0051] Example 1

[0052] 1. Metagenomic sequencing of 16S rRNA in saliva samples of individuals with different BMI

[0053] Saliva samples were collected from 53 unrelated healthy individuals and divided into three groups according to their BMI. The basic information of each group is shown in Table 1.

[0054] Table 1 Basic information of 3 saliva samples

[0055]

[0056]

[0057] Note: Underweight: low group; Normalweight: normal group; Overweight: overweight group.

[0058] A DNA library of 53 saliva samples was constructed as follows:

[0059] Taking an individual's saliva sample as an example, the MagAtrract PowerSoil Pro DNA extraction kit was used to extract DNA from the saliva sample. Then, the first round of PCR amplification was performed on the extracted DNA sample using a primer pair in the V3-V4 region (upstream primer: ACTCCTACGGGAGGCAGCAG, SEQ ID NO.1; downstream primer: GGACTACHVGGGTWTCTAAT, SEQ ID NO.2). The reaction system is shown in Table 2, and the reaction conditions are shown in Table 3. In SEQ ID NO.2, H is A, C or T; V is A, C or G; and W is A or T.

[0060] Table 2 First round PCR reaction system

[0061] Components volume 5×TransStartFastPfu Buffer 4μL 2.5 mM dNTPs 2μL Upstream primer (5 μM) 0.8μL Downstream primer (5 μM) 0.8μL <![CDATA[ddH2O]]> 7μL Sample DNA 10μL TransStartFastPfuDNA Polymerase 0.4μL

[0062] Table 3 First round PCR reaction conditions

[0063]

[0064] After the first round of PCR, 2% agarose gel was used for electrophoresis, and then the electrophoresis product was purified using the AxyPrep DNA gel extraction kit. The purified product was amplified by the second round of PCR using the NEXTFLEX Rapid DNA-Seq kit (with its own primers). The reaction system is shown in Table 4, and the reaction conditions are shown in Table 5.

[0065] Table 4 Second round PCR reaction system

[0066]

[0067] Table 5 Second round PCR reaction conditions

[0068]

[0069] 2. DNA Library Purification and Quantification

[0070] After the DNA library is constructed, the PCR product is subjected to agarose gel electrophoresis, the target fragment is selected for gel excision recovery, and then subjected to magnetic bead purification. The purified DNA library is quantified using the QuantusTM fluorescence instrument of Promega (USA). Based on the quantitative results, the DNA library is homogenized.

[0071] 3. Sequencing and microbial taxonomic annotation. The constructed library was sequenced using the Illumina NovaSeq6000 sequencer. For the obtained sequencing data, the UPARSE7.1 software was used to divide the sequences into operational taxonomic units (OTUs) with a sequence similarity of 97%. The RDP classifier was used to compare with the microorganisms in the SILVAv138 database to obtain the microbial annotation information of each sample at different taxonomic levels.

[0072] 4. Microbial community distribution characteristics of individuals with different BMI

[0073] Figure 2-3 Heat maps of microbial community distribution at the phylum and genus levels for different BMI groups are shown. The results showed that at the phylum level, p__Firmicutes, p__Proteobacteria, and p__Bacteroidota were the three phyla with the highest content. In addition, p__Spirochaetota showed a relatively high abundance in the normal-weight group compared with other groups; p__Campilobacter ota had a relatively low abundance in the overweight group. At the genus level, a total of 46 genera were identified (species with a relative abundance of less than 1% were classified as other species) ( Figure 3 Among them, the five microorganisms with the highest genus abundance were g__Streptococcus, g__Neisseria, g__Prevotella, g__Porphyromonas, and g__Haemophilus. Compared with other groups, g__Alloprevotella and g__Prevotella showed relatively high abundance in the overweight group; g__Neisseria and g__Moraxell a showed relatively high abundance in the underweight group; g__Achromobacter, g__Selenomonas, and g__Filifactor showed relatively high abundance in the normal weight group.

[0074] Subsequently, the present invention used the partial least squares linear discriminant analysis method to analyze these saliva samples based on species at the phylum and genus levels. Figure 4-5 .from Figure 4-5 It can be seen that at the genus level, individuals in different BMI groups have better discrimination efficiency, indicating that the microbial community at the genus level has better BMI discrimination efficiency. Therefore, in the subsequent microbial selection and machine learning model construction, the present invention is mainly based on the microbial community at the genus level.

[0075] 5. Selection of BMI-related microbial markers

[0076] Based on the distribution characteristics of microbial communities at the genus level in individuals with different BMI, the present invention first uses the LEfSe software system to identify microbial markers with significantly different distributions between different groups. The selection principle is as follows: LDA of microbial markers between different BMI groups ≥ 2 and P < 0.05. Based on this principle, a total of 16 microbial markers were identified. In addition, the present invention also uses the microeco package in the R software to evaluate the importance of microbial communities in BMI identification, and further identifies microbial markers related to BMI. The selection principle is: Mean Decrease Gini of microorganisms between different BMI groups ≥ 0.1 and P < 0.05. Based on this principle, a total of 17 microbial markers were identified. The results obtained by the two screening methods are shown in Figure 6-8 Through the above methods, 16 microbial markers with overlap were finally selected. The species information of the 16 microbial markers and their average relative abundance in different BMI groups are shown in Table 6.

[0077] Table 6 Relative abundance of 16 selected microorganisms and numerical information of two selection methods

[0078]

[0079] 6. Construction of machine learning model

[0080] Based on the selected 16 microbial markers, the present invention uses the randomforest package in the R software package to construct an intelligent model for predicting BMI of saliva samples. Before model construction, the present invention first uses the adaptive synthesis algorithm SMOTE-NC to simulate the samples according to the ratio of 1:1:1 between the sample sizes of the groups to solve the problem of unbalanced sample sizes between different groups. Next, the constructed data set is randomly divided into training set and test set samples according to the ratio of 7:3. Then, the present invention uses the training set samples to construct a machine learning model for BMI prediction; at the same time, the caret package in the R software is used to optimize the parameters of the constructed model using a ten-fold cross validation method to determine an optimal prediction model for subsequent prediction of BMI of saliva samples.

[0081] Finally, the present invention uses the test set samples to evaluate the performance of the model. The scatter plot of the predicted BMI values ​​and the actual BMI values ​​in the test set samples is shown in Fig. 9 The results show that the constructed random forest model has a relatively high correlation between the predicted BMI values ​​and the actual BMI values ​​of the test set samples (r 2 =0.8825), the MAE and RMSE of the predicted BMI value and the actual BMI value are 1.3094 and 1.6933 respectively. In addition, the prediction results of the BMI grouping of the test set samples are as follows Fig.10As shown, the results show that only 7 of the 32 samples in the test set were misclassified, among which all the samples in the Underweight group were correctly classified, and the classification accuracy of all samples was 78.12%, indicating that most individuals in different BMI groups can be divided into the correct BMI grouping.

[0082] In summary, the microbial markers selected by the present invention and the machine learning model constructed on this basis can better realize the inference of BMI of saliva samples. The results of the present invention can provide new solutions and research methods for BMI characterization based on saliva-related samples in the future.

[0083] The embodiments described above are only descriptions of the preferred embodiments of the present invention, and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. Use of a reagent for detecting the abundance of a combination of microbial markers in the preparation of a product for predicting adult BMI based on a saliva sample for non-diagnostic purposes, characterized in that: The microbial marker combination is the following microorganisms at the genus level: Achromobacter , Aerococcus , Alloprevotella , Amnipila , Blastomonas , Campylobacter , Catonella , Enhydrobacter , Mobiluncus , Oribacterium , Prevotella , Pseudolabrys , Selenomonas , unclassified_f__Pasteurellaceae , unclassified_o__Bacteroidales and Weissella .

2. The use according to claim 1, characterized in that: The product is a reagent or a kit.

3. The use according to claim 1, characterized in that: The product is a chip.

Images