JP2021009040A - Periodontal disease progression risk measuring method and kit - Google Patents

Abstract

To provide means for measuring a progress risk to periodontitis at the stage of gingivitis.SOLUTION: A kit for detecting bacteria in the oral cavity comprises 1, 2, 3, 4 or 5 probes for detecting bacteria selected from the group consisting of (1) a probe for detecting Megasphaera genus bacteria, (2) a probe for detecting Bergeyella genus bacteria, (3) a probe for detecting Atopobium genus bacteria, (4) a probe for detecting Actinomyces genus bacteria and (5) a probe for detecting Haemophilus genus bacteria.SELECTED DRAWING: None

Description

The present disclosure relates to a method for measuring the risk of periodontal disease progression, a kit for the measurement, and the like.

The contents of all documents described in the present specification are incorporated in the present specification by reference.

In periodontal disease, gingival inflammation first occurs and then progresses to periodontitis. Gingival inflammation is localized to the gingiva, and periodontitis is a symptom of inflammation and tissue destruction that extends to other periodontal tissues.

Japanese Unexamined Patent Publication No. 2017-023093

At the stage of gingival inflammation, it is possible to recover to the original healthy state by sterilizing periodontal disease bacteria, but if it progresses to periodontitis, it will return to the original healthy state. It's hard to go back. Therefore, in order to prevent periodontal disease, it is important to treat at the stage of gingival inflammation and prevent the progression to periodontitis.

Many studies have been made on the risk, prevention, treatment, etc. of periodontal disease (for example, Patent Document 1). However, no one has examined the risk of progression by distinguishing between gingival inflammation and periodontitis.

As mentioned above, prevention and treatment at the stage of gingival inflammation are important. Therefore, if the risk of progression to periodontitis can be measured at the stage of gingival inflammation, measures to suppress the progression to periodontitis can be taken. It is preferable because it is considered that it can be performed efficiently. Therefore, the present inventors have attempted to develop a means capable of measuring the risk of progression to periodontitis at the stage of gingival inflammation.

The present inventors can use the following information on the oral abundance ratio of bacteria (1) to (5) or (1) to (6) to accurately progress the progression from gingival inflammation to periodontitis. We found a predictable possibility and conducted further studies. (1) Megasphaera bacterium, (2) Bergeyella bacterium, (3) Atopobium bacterium, (4) Actinomyces bacterium, (5) Haemophilus bacterium, (6) Lachnospira bacterium.

The present disclosure includes, for example, the subjects described in the following sections.
Item 1.
A kit for detecting bacteria in the oral cavity
(1) Probe for detecting bacteria belonging to the genus Megasphaera,
(2) Probe for detecting bacteria of the genus Bergeyella,
(3) Probe for detecting Atopobium bacteria,
A kit comprising a probe for detecting bacteria 1, 2, 3, 4, or 5 selected from the group consisting of (4) a probe for detecting Actinomyces bacteria and (5) a probe for detecting Haemophilus bacteria.
Item 2.
The probes for detecting bacteria in (1) to (5) are the same or different,
Includes an antibody or fragment thereof that recognizes the bacterium, a nucleic acid that recognizes the nucleic acid possessed by the bacterium, a substrate catalyzed by the bacterium, or a combination thereof.
Item 1. The kit according to item 1.
Item 3.
Item 2. The kit according to Item 2, wherein the bacterium detection probe of (1) to (5) contains a substrate to be catalyzed and an indicator for recognizing the catalytic reaction.
Item 4.
Item 3. The kit according to Item 3, wherein the bacterium detection probe of (1) to (5) contains a primer that specifically amplifies the nucleic acid possessed by the bacterium.
Item 5.
Item 4. The kit according to any one of Items 1 to 4, which comprises a probe for detecting all the bacteria (1) to (5).
Item 6.
(1) Bacteria of the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
For sequencing 16S ribosomal DNA and / or 16S ribosomal RNA of 1, 2, 3, 4, or 5 oral bacteria selected from the group consisting of (4) Actinomyces bacteria and (5) Haemophilus bacteria. is there,
A kit comprising a primer set for 16S ribosomal DNA and / or 16S ribosomal RNA amplification, and a reagent for sequencing the amplified nucleic acid by the primer set.
Item 7.
Item 4. The kit according to any one of Items 1 to 6, for measuring the abundance ratio of the bacteria (1) to (5) in the oral cavity.
Item 8.
A kit for measuring the abundance ratio of bacteria in the oral cavity.
A kit comprising a primer set for 16S ribosomal DNA and / or 16S ribosomal RNA amplification, and a reagent for sequencing the amplified nucleic acid by the primer set.
Item 9.
Item 6. The kit according to any one of Items 1 to 8, wherein the oral bacterium is an oral bacterium of a gingival inflammation patient.
Item 10.
The kit according to claim 9, which is used for data acquisition for measuring the risk of progression from gingival inflammation to periodontitis.
Item 11.
In the oral cavity of patients with periodontitis (1) Bacteria belonging to the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
A group of data reflecting the abundance of (4) Actinomyces bacteria and (5) Haemophilus bacteria, and
A step A in which a data group indicating whether or not the patient's gingitis has progressed to periodontitis is input to the artificial intelligence model as learning data and the artificial intelligence model is trained.
In the oral cavity of a periodontitis patient different from the periodontitis patient (1) Megasphaera bacteria,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
Step B to acquire data reflecting the abundance of (4) Actinomyces bacteria and (5) Haemophilus bacteria, and
In step C, in which the data acquired in step B is input to the trained artificial intelligence model,
The trained artificial intelligence model includes step D for calculating whether the patient's gingitis that has acquired data in step B progresses to periodontitis.
A method of predicting the risk of gingival inflammation progressing to periodontitis.
Item 12.
In the oral cavity of patients with periodontitis (1) Bacteria belonging to the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
Based on data that reflect the abundance of (4) Actinomyces and (5) Haemophilus.
A trained model for operating a computer to calculate whether the patient's gingival inflammation progresses to periodontitis.
In the oral cavity of patients with periodontitis (1) Bacteria belonging to the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
A group of data reflecting the abundance of (4) Actinomyces bacteria and (5) Haemophilus bacteria, and
A trained model in which a data group indicating whether or not the patient's gingitis has progressed to periodontitis is input to an artificial intelligence model (preferably a random forest) as learning data and trained.
Item 13.
In the oral cavity of patients with periodontitis (1) Bacteria belonging to the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
Step B to acquire data reflecting the abundance of (4) Actinomyces bacteria and (5) Haemophilus bacteria, and
In step C, in which the data acquired in step B is input to the trained model according to item 12,
The trained model includes step D for calculating whether the patient's gingitis, for which data was acquired in step B, progresses to periodontitis.
A method of predicting the risk of gingival inflammation progressing to periodontitis.
Item 14.
In the oral cavity of patients with periodontitis (1) Bacteria belonging to the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
Data reflecting the abundance of (4) Actinomyces bacteria and (5) Haemophilus bacteria
Step C to input to the trained model according to Item 12 and
The trained model includes a step D of calculating whether the gingival inflammation of the patient who acquired the data group progresses to periodontitis.
A method of predicting the risk of gingival inflammation progressing to periodontitis.
Item 15.
In the oral cavity of patients with periodontitis (1) Bacteria belonging to the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
Based on a group of data reflecting the abundance of (4) Actinomyces bacteria and (5) Haemophilus bacteria.
A method of manufacturing a trained model for operating a computer to calculate whether the patient's gingival inflammation has progressed to periodontitis.
In the oral cavity of patients with periodontitis (1) Bacteria belonging to the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
A group of data reflecting the abundance of (4) Actinomyces bacteria and (5) Haemophilus bacteria, and
A method for manufacturing a trained model, which comprises inputting a data group indicating whether or not the patient's gingitis has progressed to periodontitis into an artificial intelligence model (preferably a random forest) as training data for training. ..
Item 16.
In the oral cavity of patients with periodontitis (1) Bacteria belonging to the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
Of the data reflecting the abundance of (4) Actinomyces bacteria and (5) Haemophilus bacteria.
Use to predict the risk of gingival inflammation progressing to periodontitis.
Item 17.
In the oral cavity of patients with periodontitis (1) Bacteria belonging to the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
Data reflecting the abundance of (4) Actinomyces bacteria and (5) Haemophilus bacteria,
With data showing whether or not the patient's gingival inflammation progressed to periodontitis,
Use for the production of a trained model for operating a computer to calculate whether gingival inflammation of a gingival patient other than the gingival inflammation patient has progressed to periodontitis.
Item 18.
A computer that implements the trained model according to item 12.

Means for predicting the progression from gingival inflammation to periodontitis, kits useful for the prediction, and the like are provided.

Hereinafter, each embodiment included in the present disclosure will be described in more detail. The present disclosure preferably includes, but is not limited to, means for predicting the progression from gingival inflammation to periodontitis, kits useful for the prediction, and the like, and the present disclosure is disclosed herein. It includes everything that can be recognized by those skilled in the art.

The kit included in the present disclosure is a kit for detecting oral bacteria, which is a kit for detecting specific bacteria. The kit may be referred to as a "kit of the present disclosure". Bacterial information detected by the kits of the present disclosure (particularly abundance or abundance ratio information) is useful for predicting the progression from gingitis to periodontitis. By using the information on the oral abundance ratio of the bacteria (1) to (5) or (1) to (6) above, the progression from gingival inflammation to periodontitis can be predicted with high accuracy. Information on the oral abundance ratio of the particular bacterium is important, and any of the kits included in the present disclosure can be preferably used to obtain the information.

In one form of the kit of the present disclosure, the kit comprises at least one probe for detecting the following bacteria (1)-(5). (1) Megasphaera bacterium, (2) Bergeyella bacterium, (3) Atopobium bacterium, (4) Actinomyces bacterium, (5) Haemophilus bacterium.

In other words, the kit is a kit for detecting bacteria in the oral cavity, which includes (1) a probe for detecting Haemophila bacteria, (2) a probe for detecting Bacteria in Bergeyella, and (3) a probe for detecting bacteria in Atopobium, (4). It can be said that the kit includes 1, 2, 3, 4, or 5 bacterial detection probes selected from the group consisting of (5) Haemophilus bacterium detection probe and (5) Haemophilus bacterium detection probe. Above all, a kit including all of these five types of bacterial detection probes is preferable. The present disclosure also includes a kit set including two or more kits including 1, 2, 3, or 4 types of bacterial detection probes. It is more preferable that the kit set includes two or more of the above kits, and as a result, the kit set includes all the probes for detecting bacteria (1) to (5).

In addition, the kit may further include (6) a probe for detecting a bacterium belonging to the genus Lachnospira. In this case, the kit is an oral bacterium detection kit, which includes (1) a Haemophila bacterium detection probe, (2) a Bergeyella bacterium detection probe, and (3) an Atopobium bacterium detection probe. 4) Bacteria 1, 2, 3, 4, 5, or 6 selected from the group consisting of a probe for detecting Actinomyces bacteria, (5) a probe for detecting Haemophilus bacteria, and (6) a probe for detecting Lachnospira bacteria. It can be said that it is a kit equipped with a detection probe. Of these, a kit including all of these six types of bacterial detection probes is preferable. The present disclosure also includes a kit set including two or more kits including 1, 2, 3, 4, or 5 types of bacterial detection probes. It is more preferable that the kit set includes two or more of the kits, and as a result, the kit set includes all the probes for detecting bacteria (1) to (6).

The probe for detecting bacteria is not particularly limited, and for example, an antibody or fragment thereof that recognizes (preferably specifically) the bacterium, or a nucleic acid that (preferably specifically) recognizes the nucleic acid possessed by the bacterium. Can be mentioned. The antibody that recognizes the bacterium may be a polyclonal antibody or a monoclonal antibody. Further, for example, the antibody may be not only an antibody that recognizes the bacterium itself, but also an antibody that recognizes a metabolite peculiar to the bacterium, a protein peculiar to the bacterium (for example, an enzyme), and the like.

Furthermore, the probe for detecting bacteria may be, for example, a substrate that the bacteria catalyzes (preferably specifically). As the substrate, it is preferable that the bacterium itself or a protein (particularly an enzyme) possessed by the bacterium (preferably specifically) acts as a catalyst and causes coloration or fluorescence by changing the substrate. Moreover, not only the substrate but also the indicator can be used together. That is, the bacterium can be detected by using a compound that recognizes the change in the substrate and produces coloration or fluorescence as an indicator. In this case, a substrate and / or an indicator can be preferably used as the probe for detecting bacteria.

As each component that can be used as such a probe for detecting bacteria, a known component or a component that can be easily prepared from a known component can be used. Further, it can be prepared by a known method or a method that can be easily conceived from a known method.

The bacterium detection probe may be a combination of two or more (for example, any natural number of 2 to 100) of each component exemplified above. For example, for example, it may be a combination of a substrate and an indicator for exhibiting the color reaction, or a combination of a plurality of "combinations of a substrate and an indicator" may be used. By using such a combination of two or more components (for example, a combination of a substrate and an indicator) or a combination of the combination itself, bacteria can be identified more accurately, which is preferable. This is because even if each component alone has a low bacterial detection specificity, the bacterial detection specificity can be increased by measuring the presence or absence of detection in a plurality of components.

For example, as a kit for identifying bacteria by using a combination of a plurality of "combinations of substrates and indicators" as described above, an anaerobic bacterium biochemical identification kit "BD BBLCRYSTAL ANR identification test reagent" (Becton Dickinson, Japan) Company). The kit can detect a plurality of bacteria, and can detect the bacteria (3) and (4) among the bacteria (1) to (5) above. In addition, in the kit, bacteria are detected by the presence or absence of a plurality of color reactions, and it is considered that the abundance (ratio) of bacteria can be measured by the strength of the color reaction, and the kit (particularly, a plurality of "substrates"). It is also possible to prepare a method and a kit capable of measuring the abundance (ratio) of bacteria according to the strength of coloration with reference to the combination of "and the combination of indicators").

Antibodies that recognize the bacterium (preferably specifically) can be prepared by known methods. That is, for example, a polyclonal antibody can be prepared by immunizing a mammal using a fragment of each bacterium or a metabolite or protein (for example, an enzyme) peculiar to each bacterium as an immune source, and for example, a hybridoma is prepared. By doing so, a monoclonal antibody can be prepared. Further, the antibody fragment can also be prepared by a known method. For example, by treating the obtained antibody with an enzyme, antibody fragments such as Fab and F (ab') 2 can be obtained. These are examples, and the antibody or a fragment thereof is not particularly limited.

Further, such a nucleic acid that recognizes (preferably specifically) the nucleic acid possessed by the bacterium can also be obtained by a known method and technique. For example, the genome sequences of these bacteria can be obtained from public data disks (eg NCBI) and used with public (eg NCBI) alignment tools and PCT primer design tools (eg BLAST) to be as specific as possible to each bacterium. It is possible to search for a specific sequence site and prepare a nucleic acid (for example, a probe or a primer) for annealing or amplifying the specific sequence site. Such a nucleic acid can be suitably used as a nucleic acid that specifically recognizes the nucleic acid possessed by the bacterium. Such nucleic acids may be, for example, DNA or RNA, and are particularly preferably DNA. The sequence length of such a nucleic acid is not particularly limited as long as specific recognition is possible, and examples thereof include about 15 to 300 bp or 20 to 100 bp. In particular, when the nucleic acid is a primer, for example, about 15 to 50 bp, 15 to 40 bp, or 20 to 35 bp can be mentioned. When the nucleic acid possessed by the bacterium is amplified by using the nucleic acid (particularly a primer), a known amplification method can be used, and although there is no particular limitation, PCR is particularly preferably exemplified from the viewpoint of convenience. To. Among the PCRs, real-time PCR is preferable because the abundance ratio of each bacterium can be easily measured. When a nucleic acid that recognizes a nucleic acid possessed by the bacterium is used as a probe for detecting the bacterium, a primer that specifically amplifies the nucleic acid possessed by the bacterium is preferable, and a primer set containing such a primer is particularly preferable. preferable.

With this kit, the abundance (or abundance ratio) of each bacterium in the oral cavity can be preferably measured. The kit can be preferably used for measuring the abundance (or abundance ratio) of each bacterium in the oral cavity.

In another embodiment of the kit of the present disclosure, the kit is a kit for measuring the abundance ratio of oral bacteria, which comprises a primer set for 16S ribosomal DNA and / or 16S ribosomal RNA amplification, and an amplified nucleic acid by the primer set. A reagent for sequencing is provided.

What kind of bacteria are present in the oral cavity by sequencing (sequencing) the nucleic acids derived from the bacteria existing in the oral cavity without specifically detecting each oral bacterium. Can be measured. That is, the bacterial species and the abundance ratio thereof can be measured depending on the ratio of the nucleic acid sequences derived from which bacteria. Based on this, by using the kit, it is possible to measure the abundance ratio of the bacteria (1) to (5) or the bacteria (1) to (6) in the oral cavity.

Therefore, the kits of the present disclosure preferably include the following forms. It is selected from the group consisting of 1, 2, 3, 4, or 5 oral bacteria selected from the group consisting of the above bacteria (1) to (5), or the group consisting of the above bacteria (1) to (6). 16S ribosomal DNA and / or 16S ribosomal RNA amplification primer set for sequencing 16S ribosomal DNA and / or 16S ribosomal RNA of 1, 2, 3, 4, 5, or 6 oral bacteria, and A kit comprising a reagent for sequencing the amplified nucleic acid by the primer set. With the kit, the abundance ratio of the oral bacteria can be preferably measured.

The primer set for amplifying 16S ribosomal DNA and / or 16S ribosomal RNA is not particularly limited as long as it is a primer set capable of amplifying 16S ribosomal DNA and / or 16S ribosomal RNA, but in particular, it recognizes a storage region of 16S ribosomal RNA. It is preferably a set of primers to be annealed. For example, 16S ribosomal DNA has nine variable regions (V1 to V9), upstream of V1, between V1-V2, between V2-V3, between V3-V4, between V4-V5, and between V5-V6. , V6-V7, V7-V8, V8-V9, or a primer that recognizes the downstream region of V9 can be used as a set by selecting two types so that 16S ribosomal DNA can be amplified. It is preferable that the amplification region contains one or more variable regions (for example, 2, 3, 4, 5, 6, 7, 8, or 9). For example, a primer set that can obtain an amplified nucleic acid containing V3 and V4 regions in the amplified region is preferable. In particular, for example, a set of a forward primer that anneals between V2-V3 and a reverse primer that anneals between V4-V5 is preferable.

As such a primer set for 16S ribosomal DNA amplification, a commercially available product can be purchased and used. For example, 16S V3-V4 Primer Mix included in 16S (V3-V4) Metagenomic Library Construction Kit for NGS (Takara Bio Inc.) can be used.

Amplification of 16S ribosomal DNA and / or 16S ribosomal RNA (particularly 16S ribosomal DNA) using the primer set can be performed by a known method. Above all, it is preferable to carry out by PCR.

By sequencing the amplified nucleic acids thus obtained and collating each amplified nucleic acid with a database to determine which bacteria are derived, the abundance ratio of oral bacteria can be measured. A HOMD (Human Oral Microbial Database) database can be used as the database. If the amplified nucleic acid has 97% or more homology with any of the bacteria belonging to each genus registered in the database, the amplified nucleic acid is considered to be derived from the genus.

As the reagent for sequencing the amplified nucleic acid, a known reagent can be used and is not particularly limited. For example, reagents for using Illumina's MiSeq system are preferred. In this case, it is particularly preferable to add an index sequence and a sequence primer sequence to the end of the amplified nucleic acid, and therefore, as the reagent, for example, a nucleic acid having the index sequence and the sequence primer sequence (particularly DNA) or for carrying out an amplification reaction. Enzymes (polymerases) and the like are preferably mentioned.

As such a reagent, a commercially available product can be purchased and used, and examples thereof include (Nextera XT Index Kit (Illumina), and in particular, Nextera XT index primer provided in the kit, etc.). As the reaction enzyme, for example, KAPA HiFi HotStart Ready Mix (Kapa Biosystems) and the like are preferably mentioned.

The kits of the present disclosure can be preferably used to measure the risk of progression to periodontitis at the stage of gingival inflammation. Therefore, the kit of the present disclosure is preferably used for detection of oral bacteria (preferably abundance ratio measurement) in patients with gingival inflammation. In addition, the kit of the present disclosure is preferably used for data acquisition for measuring the risk of progression from gingival inflammation to periodontitis (particularly whether or not it progresses).

The present disclosure preferably includes a method for measuring the risk of gingival inflammation progressing to periodontitis, a trained model generated by the method, and the like.

The method comprises the following steps A-D, B-D, or C-D.
Step A: A data group that reflects the abundance of bacteria (1) to (5) in the oral cavity of a gingitis patient, and a data group that shows whether or not the patient's gingitis has progressed to periodontitis. Is input to the artificial intelligence model as training data, and the artificial intelligence model is trained.
Step B: A step of acquiring data reflecting the abundance of (1) to (5) in the oral cavity of a gingival inflammation patient different from the gingival inflammation patient.
Step C: A step of inputting the data acquired in step B into the trained artificial intelligence model.
Step D: A step D in which the trained artificial intelligence model is made to calculate whether or not the patient's gingitis that has acquired data in step B progresses to periodontitis.

The part of "bacteria of (1) to (5)" in each of the above steps and the description of each of the following steps may be replaced (or read) with "bacteria of (1) to (6)". Further, the artificial intelligence model (learner) used in the method is not particularly limited as long as it is a known artificial intelligence model and exhibits the effect of the present invention, and a deep learning model, a machine learning model, or the like is used. be able to. In addition, an ensemble (to generate one model by fusing a plurality of models) may be performed. Among the artificial intelligence models, the random forest model is particularly preferable.

As such a known artificial intelligence model, for example, various algorithms, programs, modules, applications, or software included in the known Python package Anaconda are preferably mentioned. For example, each algorithm provided in the scikit-learn module included in Anaconda can be preferably used, and among them, Random Forest Classifier can be particularly preferably used as a random forest model.

The step A can be roughly said to be a step of feeding the learning data to the artificial intelligence model for learning.

(A-1) The data reflecting the abundance of bacteria is preferably data showing the abundance ratio of bacteria. The data is preferably data in which the abundance of bacteria in (1) to (5) is represented by a relative value. Further, (A-2) data indicating whether or not the patient's gingival inflammation has progressed to periodontitis means that the gingival inflammation of each patient has "progressed" or "not progressed" to periodontitis. It is the data which shows either of.

For example, after measuring the abundance (ratio) of the bacteria of (1) to (5) for a plurality of gingival inflammation patients by the kit of the present disclosure or the like, the periodontal disease patients are later (for example, one year later or 2). By investigating whether or not periodontitis has progressed (years later), (A-1) data and (A-2) data can be obtained for each gingival inflammation patient. The (A-1) and (A-2) data can be used as training data. The number of gingival inflammation patients (and thus the number of learning data) for collecting these learning data is preferably 10 or more, for example. Further, it is more preferably 15 or more, 20 or more, 25 or more, 30 or more, 35 or more, 40 or more, 45 or more, 50 or more, or 55 or more. The upper limit is not particularly limited, and examples thereof include 100, 1000, 10000, and 100000.

Step B roughly wants to predict the risk of gingival inflammation progressing to periodontitis (more specifically, predicting whether or not gingival inflammation progresses to periodontitis) in the oral cavity of a gingival patient. , (1) to (5) can be said to be a step of acquiring abundance (ratio) data of bacteria.

The data can be obtained, for example, using the kits of the present disclosure described above.

Step C is a step of inputting the data obtained in step B into the trained model manufactured in step A, and step D acquires input data in the trained model based on the input data. It is a process of predicting the risk that gingitis of a patient with periodontitis progresses to periodontitis (more specifically, predicting whether or not it progresses).

It should be noted that each step is performed on a computer, and the present disclosure also includes a computer on which such a trained model is implemented. The present disclosure also includes, for example, a trained model manufacturing method including step A.

Further, as described above, the data (preferably the abundance ratio data) reflecting the abundance of the bacteria (1) to (5) or the bacteria (1) to (6) in the oral cavity of the gingival inflammation patient is. It is useful for predicting the risk of gingival inflammation in the patient with periodontitis progressing to periodontitis. Therefore, the present disclosure describes data (preferably abundance ratio data) that reflect the abundance of bacteria (1) to (5) or bacteria (1) to (6) in the oral cavity of a patient with periodontitis. It also preferably includes its use for predicting the risk of gingival inflammation progressing to periodontitis. In addition, when the data is a set of data obtained from a plurality of gingival inflammation patients, it may be called a data group.

Furthermore, as described above, the data reflecting the abundance of the bacteria (1) to (5) or the bacteria (1) to (6) in the oral cavity of the periodontitis patient (preferably abundance ratio data) The data indicating whether or not the gingival inflammation of the gingival inflammation patient has progressed to periodontitis is calculated as to whether or not the gingival inflammation of a gingival inflammation patient other than the gingival inflammation patient has progressed to periodontitis. It is useful for manufacturing trained models to make the computer work. Therefore, the present disclosure includes data (preferably abundance ratio data) reflecting the abundance of bacteria (1) to (5) or bacteria (1) to (6) in the oral cavity of a patient with periodontitis. To calculate whether the gingival inflammation of a gingival inflammation patient other than the gingival inflammation patient has progressed to periodontitis with the data indicating whether or not the gingival inflammation of the gingival inflammation patient has progressed to periodontitis. It also preferably includes its use for the manufacture of trained models for the functioning of computers. When the data is a set of data obtained from a plurality of gingival inflammation patients, it may be referred to as a data group.

In addition, in this specification, "including" also includes "consisting essentially" and "consisting of" (The term "comprising" includes "consisting essentially of" and "consisting of."). The present disclosure also includes all combinations of the constituent requirements described herein.

In addition, the various properties (property, structure, function, etc.) described for each embodiment of the present disclosure described above may be combined in any way in specifying the subject matter included in the present disclosure. That is, the present disclosure includes all subjects consisting of any combination of each combinable property described herein.

Hereinafter, embodiments of the present disclosure will be described in more detail with reference to examples, but the embodiments of the present disclosure are not limited to the following examples.

DNA was extracted from saliva samples collected from subjects (patients with periodontitis) using the QIAamp DNA Mini Kit (Qiagen). For the sequence library, a DNA solution amplified by a two-step PCR method targeting the V3-V4 region of 16S rDNA was used. The first step was to amplify the V3 to V4 variable regions of the bacterial 16S rRNA gene contained in the DNA solution. As a specific method, a nucleic acid amplification reaction was carried out by the PCR method using the following primer set and 2 × KAPA HiFi HotStart Ready Mix (Kapa Biosystems). The composition of the reaction solution is shown below. The amplification reaction is carried out by heating at 95 ° C for 3 minutes, 95 ° C-30 seconds, 62.3 ° C-30 seconds, 72 ° C-30 seconds as one cycle, repeating this 25 times, and then heating at 72 ° C-5 minutes. It was.
[Primer set]
Forward primer: TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTACGGGNGGCWGCAG
Reverse primer: GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGACTACHVGGGTATCTAATCC

[Composition of reaction solution]
・ DNA-extracted sample 10.5 μL
・ 16S V3 Forward primer 5 μM 1.0 μL
・ 16S V4 Reverse primer 5 μM 1.0 μL
・ 2 × KAPA HiFi HotStart Ready Mix 12.5 μL
25.0 μL

The amplified sample was purified using magnet beads AM PureXP (Beckman Coulter) to remove free primers and primer dimers.

As a second step, dual index and illuminator sequence adapters were added to each sample to identify each sample.

Using 25 μL of the purified PCR product as a template, a dual index and an illumination sequence adapter were added using a second primer set (Nextera XT Index Kit). As a specific method, a nucleic acid amplification reaction was carried out by the PCR method using Nextera XT index primer (Illumina) and 2 × KAPA HiFi HotStart Ready Mix (Kapa Biosystems). The composition of the reaction solution is shown below. The amplification reaction is carried out by heating at 95 ° C for 3 minutes, 95 ° C-30 seconds, 55 ° C-30 seconds, 72 ° C-30 seconds as one cycle, repeating this 8 times, and then heating at 72 ° C-5 minutes. It was.

[Composition of reaction solution]
・ Sample 5.0 μL obtained in the first step
・ Nextera XT Index Primer 1 5.0 μL
・ Nextera XT Index Primer 2 5.0 μL
・ 2 × KAPA HiFi HotStart Ready Mix 25.0 μL
50.0 μL

The amplified sample was purified using magnet beads AM PureXP (Beckman Coulter) to remove free primers and primer dimers. For the purified PCR product, the DNA concentration was measured with the Quant-iT PicoGreen dsDNA Assay kit (manufactured by Life Technologies). Each sample was diluted to the same concentration, mixed, and sequence analysis was performed with MiSeq using the MiSeq v3 Reagent kit (Illumina).

For the DNA sequence obtained by sequence analysis, a sequence having 97% or more homology with the HOMD (Human Oral Microbial Database) database was picked up using QIIME (analysis software), and the sample was used as the relative bacterial count. Calculated for each.

Then, by multiplying the data obtained by the analysis and the subject information, 28 subjects whose gingival inflammation progressed to periodontitis two years after the analysis were "progressed" (Progress), and the gingival inflammation was maintained. Alternatively, 85 improved subjects were set as “no progression risk” (Nonprogress), and a group was set (Table 1: “ID” is an identification number that distinguishes each subject).

Then, using the relative number of bacteria (abundance ratio) for each subject obtained by sequence analysis, the bacteria that had a significant difference between "at risk of progression" and "without risk of progression" by the Kruskal-Wallis test were found. Selected. As a result, 13 kinds of bacteria were selected. Table 2 shows the relative number (abundance ratio) of these 13 types of bacteria in each subject.

Next, the data on the presence or absence of progression risk (Table 1) for 113 subjects and the data on the relative ratio (abundance ratio) of these bacteria (Table 2) were randomly used as training data and test data. Divide it in half and let a random forest model (specifically, RandomForestClassifier in the scikit-learn module of Anaconda) feed the training data (Tables 1 and 2) to train and test the obtained trained model. We entered the data (Table 2) and examined whether the presence or absence of progression risk (Table 1) could be answered correctly. Table 3 (Examination Example 1) shows the importance of each bacterial species data (feature amount) and the calculated correct answer rate. The importance of the features of the data analyzed in the random forest is 1 when all are added together.

Furthermore, among the 13 bacterial species, the less important ones were deleted from the data in order, and the examination was repeated using the random forest model in the same manner (Examination Examples 2 to 12). Table 3 also shows the importance of each bacterial species data (feature amount) in these studies and the calculated accuracy rate. In Table 3, the blanks indicate that the data of the bacterial species was not used for the examination (learning and examination of the correct answer).

From the results, when using the data of (1) Megaspherea bacterium, (2) Bergeyella bacterium, (3) Atopobium bacterium, (4) Actinomyces bacterium, and (5) Haemophilus bacterium, the correct answer rate However, it was found that if even one of these five types of bacteria lacks data, the accuracy rate drops significantly. Furthermore, it was also suggested that the accuracy rate could be further increased when the data of (6) Bacteria belonging to the genus Lachnospira were also used. Therefore, it was confirmed that the risk of gingival inflammation progressing to periodontitis can be predicted with high accuracy by constructing a trained model using the abundance ratio data of these 5 or 6 types of bacteria in the oral cavity. It was.

Claims (18)

A kit for detecting bacteria in the oral cavity
(1) Probe for detecting bacteria belonging to the genus Megasphaera,
(2) Probe for detecting bacteria of the genus Bergeyella,
(3) Probe for detecting Atopobium bacteria,
A kit comprising a probe for detecting bacteria 1, 2, 3, 4, or 5 selected from the group consisting of (4) a probe for detecting Actinomyces bacteria and (5) a probe for detecting Haemophilus bacteria. The probes for detecting bacteria in (1) to (5) are the same or different,
Includes an antibody or fragment thereof that recognizes the bacterium, a nucleic acid that recognizes the nucleic acid possessed by the bacterium, a substrate catalyzed by the bacterium, or a combination thereof.
The kit according to claim 1. The kit according to claim 2, wherein the bacterium detection probe of (1) to (5) contains a substrate to catalyze and an indicator for recognizing the catalytic reaction. The kit according to claim 3, wherein the bacterium detection probe of (1) to (5) contains a primer that specifically amplifies the nucleic acid possessed by the bacterium. The kit according to any one of claims 1 to 4, further comprising a probe for detecting all the bacteria (1) to (5). (1) Bacteria of the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
For sequencing 16S ribosomal DNA and / or 16S ribosomal RNA of 1, 2, 3, 4, or 5 oral bacteria selected from the group consisting of (4) Actinomyces bacteria and (5) Haemophilus bacteria. is there,
A kit comprising a primer set for 16S ribosomal DNA and / or 16S ribosomal RNA amplification, and a reagent for sequencing the amplified nucleic acid by the primer set. The kit according to any one of claims 1 to 6, for measuring the abundance ratio of the bacteria of (1) to (5) in the oral cavity. A kit for measuring the abundance ratio of bacteria in the oral cavity.
A kit comprising a primer set for 16S ribosomal DNA and / or 16S ribosomal RNA amplification, and a reagent for sequencing the amplified nucleic acid by the primer set. The kit according to any one of claims 1 to 8, wherein the oral bacterium is an oral bacterium of a patient with gingival inflammation. The kit according to claim 9, which is used for data acquisition for measuring the risk of progression from gingival inflammation to periodontitis. In the oral cavity of patients with periodontitis (1) Bacteria belonging to the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
A group of data reflecting the abundance of (4) Actinomyces bacteria and (5) Haemophilus bacteria, and
A step A in which a data group indicating whether or not the patient's gingitis has progressed to periodontitis is input to the artificial intelligence model as learning data and the artificial intelligence model is trained.
In the oral cavity of a periodontitis patient different from the periodontitis patient (1) Bacteria of the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
Step B to acquire data reflecting the abundance of (4) Actinomyces bacteria and (5) Haemophilus bacteria, and
In step C, in which the data acquired in step B is input to the trained artificial intelligence model,
The trained artificial intelligence model includes step D for calculating whether the patient's gingitis that has acquired data in step B progresses to periodontitis.
A method of predicting the risk of gingival inflammation progressing to periodontitis. In the oral cavity of patients with periodontitis (1) Bacteria belonging to the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
Based on data that reflect the abundance of (4) Actinomyces and (5) Haemophilus.
A trained model for operating a computer to calculate whether the patient's gingival inflammation progresses to periodontitis.
In the oral cavity of patients with periodontitis (1) Bacteria belonging to the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
A group of data reflecting the abundance of (4) Actinomyces bacteria and (5) Haemophilus bacteria, and
A trained model in which a data group indicating whether or not the patient's gingitis has progressed to periodontitis is input to an artificial intelligence model as training data and trained. In the oral cavity of patients with periodontitis (1) Bacteria belonging to the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
Step B to acquire data reflecting the abundance of (4) Actinomyces bacteria and (5) Haemophilus bacteria, and
In step C, in which the data acquired in step B is input to the trained model according to claim 12,
The trained model includes step D for calculating whether the patient's gingitis, for which data was acquired in step B, progresses to periodontitis.
A method of predicting the risk of gingival inflammation progressing to periodontitis. In the oral cavity of patients with periodontitis (1) Bacteria belonging to the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
Data reflecting the abundance of (4) Actinomyces bacteria and (5) Haemophilus bacteria
Step C to input to the trained model according to claim 12,
The trained model includes a step D of calculating whether the gingival inflammation of the patient who acquired the data group progresses to periodontitis.
A method of predicting the risk of gingival inflammation progressing to periodontitis. In the oral cavity of patients with periodontitis (1) Bacteria belonging to the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
Based on a group of data reflecting the abundance of (4) Actinomyces bacteria and (5) Haemophilus bacteria.
A method of manufacturing a trained model for operating a computer to calculate whether the patient's gingival inflammation has progressed to periodontitis.
In the oral cavity of patients with periodontitis (1) Bacteria belonging to the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
A group of data reflecting the abundance of (4) Actinomyces bacteria and (5) Haemophilus bacteria, and
A method for manufacturing a trained model, which comprises inputting a data group indicating whether or not the patient's gingitis has progressed to periodontitis into an artificial intelligence model as training data for training. In the oral cavity of patients with periodontitis (1) Bacteria belonging to the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
Of the data reflecting the abundance of (4) Actinomyces bacteria and (5) Haemophilus bacteria.
Use to predict the risk of gingival inflammation progressing to periodontitis. In the oral cavity of patients with periodontitis (1) Bacteria belonging to the genus Megasphaera,
(2) Bacteria of the genus Bergeyella,
(3) Atopobium bacteria,
Data reflecting the abundance of (4) Actinomyces bacteria and (5) Haemophilus bacteria,
With data showing whether or not the patient's gingival inflammation progressed to periodontitis,
Use for the production of a trained model for operating a computer to calculate whether gingival inflammation in a gingival inflammation patient other than the gingival inflammation patient has progressed to periodontitis. A computer that implements the trained model according to claim 12.