Method for predicting change in body fat

Abstract

To provide a technique for predicting a change in body fat for each subject due to the intake of chlorogenic acids.SOLUTION: A method for predicting a change in body fat according to the present invention is a computer-executed method for predicting a change in body fat in a subject due to the intake of chlorogenic acids, and includes: obtaining gene expression information, including information on the expression level of at least one associated gene or its expression product associated with the change in body fat due to the intake of chlorogenic acids, in association with subject identification information; and predicting the change in body fat in the subject due to the intake of chlorogenic acids from the gene expression information, based on correlation information indicating the correlation between the change in body fat in a plurality of sample subjects who have taken chlorogenic acids and the expression level of the associated gene or its expression product associated with the change in body fat.SELECTED DRAWING: Figure 4

Description

Technical Field

[0001] The present invention relates to a technique for predicting changes in body fat of a subject.

Background Art

[0002] From the viewpoint of preventing lifestyle-related diseases and maintaining health, methods for efficiently reducing body fat such as visceral fat have been explored. As one of them, a body fat-reducing effect by ingestion of a composition containing coffee chlorogenic acids or chlorogenic acids which are its main components is known. For example, Patent Document 1 discloses an action of suppressing weight gain and reducing visceral fat mass by ingestion of chlorogenic acids. For example, Non-Patent Documents 1, 2, and 3 report actions such as a decrease in visceral fat area, a decrease in subcutaneous fat area, a decrease in body weight, a decrease in abdominal circumference, an increase in energy consumption, and an increase in lipid combustion amount by ingestion of chlorogenic acids.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Non-Patent Documents

[0004]

Non-Patent Document 1

Non-Patent Document 2

Non-Patent Document 3

Summary of the Invention

Problems to be Solved by the Invention

[0005] On the other hand, these documents do not describe a technique for predicting changes in body fat for each subject in consideration of individual differences in changes in body fat due to the intake of chlorogenic acids.

[0006] The present invention relates to a technique for predicting changes in body fat for each subject due to the intake of chlorogenic acids.

Means for Solving the Problems

[0007] A method for predicting changes in body fat according to one embodiment of the present invention is a method for predicting changes in body fat of a subject due to the intake of chlorogenic acids, which is executed by a computer, and includes: obtaining gene expression information including information on the expression level of at least one related gene or its expression product related to changes in body fat due to the intake of chlorogenic acids, in association with subject information including identification information for identifying the subject; predicting changes in body fat of the subject due to the intake of chlorogenic acids from the gene expression information based on correlation information indicating the correlation between changes in body fat in a plurality of sample subjects who have ingested chlorogenic acids and the expression level of the related gene or its expression product related to the changes in body fat.

[0008] A body fat change prediction system according to another embodiment of the present invention is a body fat change prediction system for predicting changes in body fat of a subject due to the intake of chlorogenic acids, and includes a storage unit, an acquisition unit, and a prediction unit. The storage unit stores correlation information indicating the correlation between changes in body fat in a plurality of sample subjects who have ingested chlorogenic acids and the expression level of a related gene or its expression product related to the changes in body fat. The acquisition unit acquires gene expression information including information on the expression level of the related gene or its expression product, in association with subject information including identification information for identifying the subject. The prediction unit predicts changes in body fat of the subject due to the intake of chlorogenic acids from the gene expression information based on the correlation information.

[0009] A program according to another aspect of the present invention is a program for predicting changes in body fat of a subject due to the intake of chlorogenic acids, which causes a computer to acquire gene expression information including information on the expression levels of at least one related gene or its expression product related to changes in body fat in a plurality of sample subjects who have ingested the chlorogenic acids, in association with subject information including identification information for identifying the subjects; predict changes in body fat of the subject due to the intake of the chlorogenic acids for each piece of the identification information of the subject from the gene expression information based on correlation information indicating the correlation between the changes in body fat and the expression levels of the related gene or its expression product in the plurality of sample subjects; and execute the steps.

Advantages of the Invention

[0010] According to the present invention, it becomes possible to predict changes in body fat due to the intake of chlorogenic acids for each subject.

Brief Description of the Drawings

[0011]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Figure 17

Figure 18

Figure 19

[0012] [Summary of the Present Invention] The present invention relates to a technique for predicting changes in body fat due to the intake of chlorogenic acids in a subject by a computer (information processing device). Conventionally, although the effect of reducing body fat by the intake of chlorogenic acids has been known (see Patent Document 1 and Non-Patent Documents 1 to 3), genes related to changes in body fat due to the intake of chlorogenic acids have not been known. Therefore, in the present invention, a correlation is found between the expression level of a related gene or its expression product extracted as being related to changes in body fat after the intake of chlorogenic acids and changes in body fat in a plurality of sample subjects, and based on the correlation information indicating this correlation, from the gene expression information of the subject, changes in body fat due to the intake of chlorogenic acids in the subject are predicted.

[0013] In one embodiment of the present invention, specifically, the chlorogenic acids include monocatechoylquinic acids containing 3-caffeoylquinic acid, 4-caffeoylquinic acid, and 5-caffeoylquinic acid; monophenoylquinic acids containing 3-feruloylquinic acid, 4-feruloylquinic acid, and 5-feruloylquinic acid; and dicaffeoylquinic acids containing 3,4-dicaffeoylquinic acid, 3,5-dicaffeoylquinic acid, and 4,5-dicaffeoylquinic acid. The chlorogenic acids can be any one or a combination of two or more of the compounds listed above. It is preferable that the chlorogenic acids include the above nine types. In the present invention, the content of the chlorogenic acids is defined based on the total amount of the above nine types.

[0014] In one embodiment of the present invention, the source of intake of the chlorogenic acids may be one type or a plurality of types. In the present specification, a composition containing chlorogenic acids and ingestible by humans is also referred to as a "chlorogenic acid-containing composition". The form of the chlorogenic acid-containing composition is not particularly limited, and is appropriately selected from, for example, foods including beverages, preparations including capsules and tablets, and the like.

[0015] In one embodiment of the present invention, the intake of chlorogenic acids for body fat change is preferably continuous intake over a predetermined period, and the change in body fat is preferably measured before and after the predetermined period. The predetermined period is preferably 1 week or longer, more preferably 10 days or longer. Further, when the subject is an adult, the lower limit of the daily intake of chlorogenic acids during the intake period is preferably 10 mg or more, more preferably 50 mg or more, and still more preferably 100 mg or more from the viewpoint of obtaining the body fat reducing effect. Further, although the upper limit of the intake amount is not particularly limited, it is preferably 30,000 mg or less, more preferably 10,000 mg or less, and still more preferably 5,000 mg or less in consideration of the balance with other nutrients etc.

[0016] In one embodiment of the present invention, the gene includes double-stranded DNA containing human genomic DNA, single-stranded DNA (sense strand) containing cDNA, single-stranded DNA (complementary strand) having a sequence complementary to the sense strand, and fragments thereof, and means those in which some biological information is contained in the sequence information of the bases constituting the DNA. Further, the "gene" includes not only the "gene" represented by a specific base sequence, but also nucleic acids encoding homologs (i.e., homologs or orthologs) thereof, mutants such as gene polymorphisms, and derivatives. In the present invention, the gene name follows the Official Symbol described in NCBI ([www.ncbi.nlm.nih.gov / ]).

[0017] In the present invention, the "expression product" of a gene is a concept encompassing the transcriptional product and the translational product of the gene. The "transcriptional product" is RNA generated by transcription from a gene (DNA), and the "translational product" means a protein encoded by the gene that is translationally synthesized based on the RNA. Note that "RNA" includes all of total RNA, mRNA, rRNA, tRNA, non-coding RNA, and synthetic RNA. Further, as the measurement target of the expression level in the present invention, RNA, DNA encoding the RNA, protein encoded by the RNA, molecule that interacts with the protein, molecule that interacts with the RNA, or molecule that interacts with the DNA, etc. may be mentioned, RNA is preferred, and mRNA is more preferred. Here, as the molecule that interacts with RNA, DNA, or protein, DNA, RNA, protein, polysaccharide, oligosaccharide, monosaccharide, lipid, fatty acid, and their phosphorylated products, alkylated products, sugar adducts, etc., and complexes of any of the above may be mentioned. Further, the expression level comprehensively means the expression amount and activity of the gene or the expression product.

[0018] In one embodiment of the present invention, body fat means fat stored in the body of mammals including humans. Body fat includes subcutaneous fat located under the skin and visceral fat located around the viscera. In one embodiment of the present invention, as the body fat change, it is preferable to predict the change in visceral fat which is highly related to lifestyle-related diseases. Further, the "change in body fat" is represented by the amount of change in body fat, the change level of body fat represented stepwise, etc. Examples of the amount of change in body fat include the volume of body fat, the area of body fat at a predetermined cross-sectional site, the mass of body fat, the body fat percentage, etc. In the case of the amount of change in visceral fat, the cross-sectional area of visceral fat in the abdomen is preferable. The change level of body fat may be represented by qualitative expressions such as "body fat is likely to decrease", "body fat is difficult to decrease", "body fat is likely to increase", as well as numbers, alphabets, symbols, etc. indicating classes.

[0019] Note that the method for predicting the change in body fat of the present invention is not carried out for medical purposes for humans, and each step does not include any therapeutic or diagnostic act for medical purposes.

[0020] In the present invention, the "system" shall include one or more computers. For example, the system may be composed of a single computer or multiple computers. In the following embodiments, an example of the latter will be shown. Hereinafter, embodiments of the present invention will be described with reference to the drawings.

[0021] [Configuration example of the system] The system according to an embodiment of the present invention includes a server 100 on the Internet 50, a plurality of subject terminals 200, and a measurement terminal 300.

[0022] The server 100 can be, for example, a web server operated by an operator of a website (body fat prediction site) capable of providing a service for predicting body fat change due to the intake of chlorogenic acids. Note that the server 100 may be composed of one computer or multiple computers. The server 100 is connected to, for example, a plurality of subject terminals 200 and the measurement terminal 300 via the Internet 50.

[0023] The subject terminals 200 (200A, 200B, 200C...) can be terminals used by subjects (subject A, subject B, subject C) who are users of the body fat change prediction service, such as smartphones, mobile phones, tablet PCs (Personal Computers), notebook PCs, desktop PCs, etc. The subject terminal 200 accesses, for example, the server 100, receives a web page or the like generated by the server 100, and displays it on the screen by a browser or the like.

[0024] The measurement terminal 300 analyzes the expression level of a gene or its expression product extracted from a biological sample of a subject. The measurement terminal 300 may be connected to a measurement device (not shown) that measures the expression level of the gene or its expression product. The measurement terminal 300 may be used by a person who analyzes the biological sample of the subject. In the example of FIG. 1, it is used by the service provider, but it may also be a person commissioned by the service provider or a person in partnership with the service provider. In the following description, for convenience, the user of the measurement terminal 300 is referred to as the "service provider". Similar to the subject terminal 200, the measurement terminal 300 can be, for example, a smartphone, a mobile phone, a tablet PC, a notebook PC, or a desktop PC.

[0025] In the example of FIG. 1, in order to obtain a biological sample from the subject, the service provider sends a biological sampling kit to each subject. Then, the subject returns the biological sample collected by the kit to the service provider.

[0026] In the present embodiment, the biological sample can be cells, body fluids (such as blood), urine, secretions (such as saliva, skin surface lipids, etc.). However, from the viewpoint of convenience in collecting gene expression information and the viewpoint of being able to identify the skin surface site, it is preferable to use skin surface lipids. Here, "skin surface lipids (SSL)" refers to the fat-soluble fraction present on the surface of the skin and is sometimes called sebum. SSL contains RNA expressed in skin cells. Examples of the skin site from which SSL is collected include the skin of any part of the body such as the head, face, neck, trunk, hands, and feet, and a site with a large secretion of skin surface lipids, such as the skin of the face, is preferable.

[0027] The biological sampling kit includes tools for collecting the above biological sample. When the biological sample is SSL, the sampling tools included in the biological sampling kit include, for example, lipid-absorbing sheet materials such as blotting paper and blotting film, glass plates, tapes, spatulas, scrapers, etc. Among these, the sampling tool is preferably a lipid-absorbing sheet material because of its high handling convenience.

[0028] In this embodiment, the measurement terminal 300 analyzes the expression level of a gene or its expression product extracted from a biological sample to generate gene expression information, and transmits the generated gene expression information to the server 100. The server 100 predicts the change in body fat (for example, visceral fat) by a predetermined process from the gene expression information and the subject information, and transmits the prediction result to the subject terminal 200. A specific operation example of the system will be described later.

[0029] [Hardware Configuration of Information Processing Device] As shown in FIG. 2, the server 100 includes, for example, a CPU (Central Processing Unit) 11, a ROM (Read Only Memory) 12, a RAM (Random Access Memory) 13, an input / output interface 15, and a bus 14 that connects these to each other.

[0030] The CPU 11 appropriately accesses the RAM 13 or the like as necessary, and comprehensively controls the entire blocks of the server 100 while performing various arithmetic processes. The ROM 12 is a non-volatile memory in which firmware such as an OS, programs, and various parameters to be executed by the CPU 11 are fixedly stored. The RAM 13 is used as a working area of the CPU 11 or the like, and temporarily holds the OS, various applications being executed, and various data being processed.

[0031] Connected to the input / output interface 15 are a display unit 16, an operation reception unit 17, a storage unit 18, a communication unit 19, and the like.

[0032] The display unit 16 is a display device using, for example, an LCD (Liquid Crystal Display), an OELD (Organic ElectroLuminescence Display), a CRT (Cathode Ray Tube), or the like.

[0033] The operation reception unit 17 is, for example, a pointing device such as a mouse, a keyboard, a touch panel, or other input devices. When the operation reception unit 17 is a touch panel, the touch panel can be integrated with the display unit 16.

[0034] The storage unit 18 is, for example, a non-volatile memory such as an HDD (Hard Disk Drive), a flash memory (SSD; Solid State Drive), or other solid-state memories. The OS, various applications, and various data are stored in the storage unit 18.

[0035] In the present embodiment, the storage unit 18 may have databases such as a subject information database and a correlation information database, in addition to programs necessary for the prediction process of body fat change described later. These databases are referred to as needed in the prediction process of body fat change.

[0036] The communication unit 19 is, for example, various modules for wireless communication such as an NIC (Network Interface Card) for Ethernet or a wireless LAN, and is responsible for communication processing with the subject terminal 200 and the like.

[0037] Although not shown, the basic hardware configurations of the subject terminal 200 and the measurement terminal 300 can be substantially the same as the hardware configuration of the server 100.

[0038] [Database Configuration of Server] As shown in FIG. 3, the server 100 has a subject information database 31 and a correlation information database 32. Note that these databases may be stored in a storage device externally connected to the server 100 or in the server instead of the storage unit 18.

[0039] The subject information database 31 stores subject identification information and attribute information for each subject who wishes to receive a body fat change prediction service. The subject identification information may be any information that can identify the subject, such as a subject ID, name (nickname), email address, etc. The attribute information is information indicating the attributes of the subject other than the identification information, and may include, for example, general information such as gender, age (age group), occupation, address (residential area), as well as information related to the subject's physical information and lifestyle habits. Examples of the subject's physical information include height, weight, visceral fat area before chlorogenic acid intake, BMI, body fat percentage, blood test items such as blood glucose level and triglyceride level, etc. Examples of information related to lifestyle habits include alcohol consumption, smoking habits, exercise habits, eating habits, etc.

[0040] In the present embodiment, the attribute information preferably includes gender information. Further, the attribute information preferably includes information related to menopause, i.e., information that can be used to infer the presence or absence of menopause, when the gender is female. Examples of information that can be used to infer the presence or absence of menopause include, for example, age (age group), basal body temperature data, data on female hormone values such as estrogen and follicle-stimulating hormone, etc. By including such information in the attribute information, as will be described later, the prediction accuracy of body fat change can be improved.

[0041] Furthermore, the subject information database 31 stores gene expression information for each subject. The gene expression information includes information on the expression levels of related genes or their expression products measured from biological samples collected from each subject. Hereinafter, "the expression levels of related genes or their expression products" will also be referred to as "the expression levels of related genes, etc."

[0042] In this embodiment, the related gene is a gene related to the change in body fat due to the intake of chlorogenic acids, specifically, a gene related to the change in body fat in a plurality of sample subjects who have ingested chlorogenic acids. The related gene can be extracted by a statistical method from data in a plurality of sample subject populations that have ingested chlorogenic acids, for example. The related gene may be a single type of gene or a gene group consisting of a plurality of types of genes. Specific examples of the related gene will be described later.

[0043] In this embodiment, the gene expression information includes, for example, information about the expression level of mRNA, which is a transcription product of the related gene (hereinafter referred to as "RNA information"). The RNA information includes, for example, information on at least one expression level selected from the expression level of mRNA extracted from a biological sample, the expression level of cDNA reverse-transcribed from the mRNA, or the expression level of an amplification product of the cDNA. For the extraction of RNA, reverse transcription, and measurement / analysis of the expression level from the biological sample in this case, known apparatuses, means, and algorithms can be used.

[0044] The correlation information database 32 stores correlation information indicating the correlation between the change in body fat in a plurality of sample subjects who have ingested chlorogenic acids and the expression level of the related gene or its expression product. The correlation information can be generated by analyzing data indicating the change in body fat due to the intake of chlorogenic acids and data indicating the expression level of the related gene or its expression product, obtained from the sample subject population. These data can be obtained, for example, by acquiring data on the change in body fat before and after the intake period and data indicating the expression level of the related gene or the like from a subject population that has ingested chlorogenic acids for a predetermined period.

[0045] From the perspective of performing highly accurate prediction processing, the correlation information preferably includes a machine learning model in which the expression levels of related genes or their expression products in a plurality of sample subjects are explanatory variables, and the changes in body fat due to the intake of chlorogenic acids in the plurality of sample subjects are objective variables. The algorithm used in the machine learning model is not particularly limited. For example, algorithms such as linear regression model (Linear model), Lasso regression (Lasso), Random Forest (Random Forest), Neural network (Neural net), Support Vector Machine with linear kernel (SVM(linear)), Support Vector Machine with rbf kernel (SVM(rbf)), decision tree, and k-nearest neighbor method can be mentioned.

[0046] In this embodiment, the expression level of the related gene or its expression product has a correlation with the change amount of body fat in a plurality of sample subjects, and the correlation information may be information indicating the correlation between the change amount of body fat and the expression level of the related gene or its expression product in a plurality of sample subjects. Thereby, the change amount of body fat can be predicted from the gene expression information. A prediction model for realizing such correlation information is, for example, a regression model. Specific examples include Random Forest and Support Vector Machine with linear kernel.

[0047] In this embodiment, the correlation information includes a plurality of correlation information for each of a plurality of attributes that a subject can have. Each correlation information for each attribute shows the relationship between the change in body fat in a plurality of sample subjects having the corresponding attribute among the plurality of sample subjects and the expression level of the related gene or its expression product related to the change in body fat. Note that the related genes of each correlation information for each attribute may be extracted from the data of the sample subject population having each attribute.

[0048] In this embodiment, the correlation information by attribute includes, for example, male correlation information and female correlation information in relation to gender. The male correlation information indicates the correlation between the change in body fat in a plurality of male sample subjects among a plurality of sample subjects and the expression level of a male-related gene or its expression product as a related gene related to the change in body fat. The female correlation information indicates the correlation between the change in body fat in a plurality of female sample subjects among a plurality of sample subjects and the expression level of a female-related gene or its expression product as a related gene related to the change in body fat. The male-related gene and the female-related gene are composed of, for example, different gene groups.

[0049] Furthermore, in this embodiment, the female correlation information includes pre-menopausal female correlation information and post-menopausal female correlation information in relation to the presence or absence of menopause. The pre-menopausal female correlation information indicates the correlation between the change in body fat in a plurality of sample subjects who are pre-menopausal women among a plurality of sample subjects and the expression level of a pre-menopausal female-related gene or its expression product as a related gene related to the change in body fat. The post-menopausal female correlation information indicates the correlation between the change in body fat in a plurality of sample subjects who are post-menopausal women among a plurality of sample subjects and the expression level of a post-menopausal female-related gene or its expression product as a related gene related to the change in body fat. The pre-menopausal female-related gene and the post-menopausal female-related gene are composed of, for example, different gene groups.

[0050] [Operation Example of the System] Next, an operation example of the body fat change prediction system configured as described above will be described using the sequence diagram of FIG. 4. The operations of the server 100 described below are executed by the cooperation of the hardware such as the CPU 11 and the communication unit 19 of the server 100 and the software stored in the storage unit 18. Similarly, the operations of the subject terminal 200 and the measurement terminal 300 are also executed by the cooperation of the hardware such as the CPU and the communication unit and the software stored in the storage unit. Hereinafter, the main body of the operation of each device is the CPU. In this operation example, an example of predicting the change in body fat (visceral fat here) in male or female subjects due to continuous intake of a chlorogenic acid-containing beverage will be described.

[0051] In this example, first, a subject who wishes to use the body fat prediction service browses the body fat prediction site using the subject terminal 200. The CPU of the subject terminal 200 receives an input of subject information including subject identification information such as an ID, password, name, and address, and attribute information such as gender and age, through an input operation of the subject on the body fat prediction site, for example (S201).

[0052] The CPU of the subject terminal 200 transmits the acquired subject information (S202), and the CPU 11 of the server 100 receives (acquires) the subject information including the subject identification information and the attribute information (S101). In this step, the CPU 11 functions as an "acquisition unit that acquires the subject's attribute information in association with the subject identification information". Note that in this specification, "acquisition" includes any of receiving information via the Internet 50, receiving information from another electronic device using short-range wireless communication or wired communication, acquiring information received through a user's input operation, etc., and acquiring information stored in the storage unit 18 or an external storage device.

[0053] The CPU 11 stores the received subject information in the subject information database 31 (S102). Thereafter, although not shown, the CPU 11 can perform processing for sending a biological sample collection kit to the subject. Examples of such processing include notifying the terminal used by the person in charge of sending the biological sample collection kit. As a result, the biological sample collection kit is sent to the subject.

[0054] The subject who has received the biological sample collection kit collects a biological sample such as an SSL using the kit, for example. The subject sends the collected biological sample to the service operator together with the subject identification information or information that can access it. The subject identification information corresponding to the biological sample is stored in the storage unit of the measurement terminal 300 (not shown), for example.

[0055] The service provider extracts, for example, mRNA from a biological sample and measures the expression level of the transcription product of the mRNA of the relevant gene using a measuring device and a measurement terminal 300 connected thereto. Note that the relevant gene to be measured may be only the relevant gene corresponding to the subject's attributes, or may be the relevant genes corresponding to all the attributes to be processed.

[0056] The CPU of the measurement terminal 300 analyzes the measurement data by the measuring device, calculates the expression level of the transcription product of the mRNA of the relevant gene, and generates gene expression information which is RNA information (S301). The CPU of the measurement terminal 300 associates the subject identification information corresponding to the biological sample with the generated gene expression information of the relevant gene and transmits it to the server 100 (S302).

[0057] The CPU 11 of the server 100 receives (acquires) the gene expression information in association with the subject identification information (S103). In this step, for example, the CPU 11 functions as an "acquisition unit that acquires the gene expression information in association with the subject identification information". The CPU 11 stores the gene expression information in the subject information database 31 in association with the subject identification information (S104). Thereby, the subject information such as attribute information and the gene expression information of the subject are associated with each other.

[0058] Subsequently, the CPU 11 determines the attributes of the subject based on the attribute information of the subject stored in the subject information database 31. In this step, the CPU 11 functions as a "determination unit". In this operation example, the CPU 11 determines the gender of the subject based on the acquired attribute information of the subject, and for example, determines whether the subject is male or not (S105). The CPU 11 can determine the gender of the subject based on, for example, the gender of the subject stored in the subject information database 31.

[0059] When it is determined that the subject is male (Yes in S105), the CPU 11 predicts the change in body fat due to the intake of chlorogenic acids in the subject's gene expression information based on male correlation information (S106). In this step, the CPU 11 functions as a "prediction unit". In this operation example, the male correlation information uses the expression levels of mRNAs, which are transcription products of male-related genes in a plurality of male sample subjects, as explanatory variables, and the change in body fat due to the intake of chlorogenic acids in a plurality of male sample subjects as the target variable, and consists of a machine learning model. For example, the CPU 11 can select a machine learning model related to male correlation information from the correlation information database 32 and apply the subject's RNA information to the model to predict the change in body fat of the subject.

[0060] In this embodiment, the male-related genes are at least one of 26 genes selected from FOXD4L6, KCTD18, TCHP, THOC3, TMEM57, C19orf53, CDC40, CMC2, CRCP, DDX19B, DHX57, DRG2, EIF2A, FABP3, MCC, NFE2L3, NOD2, PIGX, PSMC6, R3HCC1L, RNF7, RNPEPL1, SLC11A1, TRIM23, UBE2L6, and UNC13D. Preferably, the male-related genes are at least one of 5 genes selected from FOXD4L6, KCTD18, TCHP, THOC3, and TMEM57, and more preferably, they are FOXD4L6, KCTD18, TCHP, THOC3, and TMEM57.

[0061] The above 26 male-related genes were extracted from the data of multiple male sample subjects who ingested chlorogenic acids as genes that have a significant correlation between the expression level of their mRNA and the change in body fat. Since each of these 26 genes has a significant correlation between the expression level of its mRNA and the change in body fat, it is possible to construct a prediction model for body fat change by using at least one gene as a feature quantity. However, by combining multiple genes among the 26 genes as feature quantities, a more accurate prediction model can be constructed. Furthermore, when the population of male sample subjects is divided into multiple groups, the above 5 male-related genes were extracted as genes that commonly have a significant correlation between the expression level of their mRNA and the change in body fat in any group. Thus, by using one or more, preferably all, of the above 5 genes as feature quantities, it becomes possible to construct a more accurate prediction model.

[0062] On the other hand, when it is determined that the subject is female (No in S105), the CPU 11 determines the presence or absence of menopause in the subject based on menopause-related information. In this operation example, the CPU 11 determines whether the subject is a menopausal woman (S107). In this step, the CPU 11 functions as a "determination unit". The CPU 11 can perform the determination process of the presence or absence of menopause according to the type of menopause-related information. For example, when the menopause-related information includes age, if the age (decade) of the subject is equal to or greater than the average age of menopause, the CPU 11 determines that the subject is a menopausal woman, and if it is less than the average age, the CPU 11 can determine that the subject is a premenopausal woman. Also, when the menopause-related information includes multiple pieces of information, etc., the CPU 11 may determine the presence or absence of menopause by using a machine learning model for predicting the presence or absence of menopause from the menopause-related information of the subject.

[0063] When it is determined that the subject is a premenopausal woman (No in S107), the CPU 11 predicts the change in body fat due to the intake of chlorogenic acids in the subject from the gene expression information based on the premenopausal woman correlation information (S108). In this step, the CPU 11 functions as a "prediction unit". In this operation example, the premenopausal woman correlation information consists of a machine learning model that uses the expression level of mRNA, which is the transcription product of premenopausal woman-related genes in a plurality of premenopausal woman sample subjects, as an explanatory variable, and the change in body fat due to the intake of chlorogenic acids in the plurality of premenopausal woman sample subjects as an objective variable. For example, the CPU 11 can select a machine learning model related to the premenopausal woman correlation information from the correlation information database 32 and apply the RNA information of the subject to the model to predict the change in body fat of the subject.

[0064] In this operation example, the genes related to menstruating women are at least one of 96 genes selected from C19orf71, DDX39A, PTPN1, ALDH9A1, ANKRD33B, APBA3, ARHGEF2, ARID5B, ARRDC2, ATAD3B, B4GALT1, BIRC3, BRMS10, C10orf128, GINM1, CC2D1A, CCDC101, CCDC130, CCDC71, CDK5, CHST7, COG3, COX6A1, CSNK2B, DDX17, ELF4, ENC1, ENGASE, ERAP2, EXOSC8, FAM110A, FAM210A, FCGBP, FCGR2C, FOXJ3, FZD5, FZD7, GABPB1, GGA3, GRN, H2AFY2, HSPA7, HSPA9, ITGA5, KCNAB2, KIAA0020, KIF1B, LIMA1, LOC100132247, LOC100289019, LOC595101, MARCKSL1, MFSD10, MGST3, MIB2, MLLT6, MYO1G, NGLY1, P2RX4, PIK3C3, PMS2P3, POLG, PPCS, PRICKLE3, PRKAA1, PTPN2, PVR, RAB31, RAB7L1, RCE1, RNF113A, RRAGC, SLC5A6, TACC1, TBC1D9B, TCIRG1, TMEM176B, TNFRSF14, TRAF3, VEGFA, XPC, YEATS20, ZFAND2A, ZFP36L2, ZFYVE26, ZNF101, ZNF33A, ZNF350, ZNF783, ABCA11P, FAM223B, LOC100271836, LOC100288778, LOC728875, RPS106AP10, and SH3GL1P2. Among these genes related to menstruating women, it is preferable to select from 3 genes of C19orf71, DDX39A, and PTPN1, and it is more preferable to include all of C19orf71, DDX39A, and PTPN1.

[0065] The above 96 genes related to premenopausal women were extracted from the data of multiple premenopausal women sample subjects who ingested chlorogenic acids as genes that have a significant correlation between the expression level of their mRNA and the change in body fat. Since these 96 genes each have a significant correlation between the expression level of their respective mRNA and the change in body fat, it is possible to construct a prediction model for body fat change by using at least one gene as a feature quantity. However, by combining a plurality of genes among the 96 genes as feature quantities, a more accurate prediction model can be constructed. Furthermore, when the above 3 genes related to premenopausal women are extracted as genes that commonly have a significant correlation between the expression level of their mRNA and the change in body fat in any group when multiple premenopausal women sample subjects are divided into multiple groups. Thus, by using one or more, preferably all, of the above 3 genes as feature quantities, it becomes possible to construct an even more accurate prediction model.

[0066] When the subject is determined to be a postmenopausal woman (Yes in S107), the CPU 11 predicts the change in body fat due to the subject's ingestion of chlorogenic acids from the gene expression information based on the postmenopausal woman correlation information (S109). In this step, the CPU 11 functions as a "prediction unit". In this operation example, the postmenopausal woman correlation information consists of a machine learning model that uses the expression level of mRNA, which is the transcription product of genes related to postmenopausal women in multiple postmenopausal women sample subjects, as an explanatory variable, and the change in body fat due to the ingestion of chlorogenic acids in multiple postmenopausal women sample subjects as an objective variable. For example, the CPU 11 can select a machine learning model related to the postmenopausal woman correlation information from the correlation information database 32 and apply the subject's RNA information to the model to predict the change in body fat of the subject.

[0067] In this operation example, the genes related to menopausal women are at least one of 54 genes selected from CRAMP1L, DHRS7, NOP14, SGK3, UQCRC1, ABCC3, ALDH2, ARHGAP27, BTN3A2, C17orf79, C6orf89, C7orf13, CARD11, CCDC23, DNPEP, DSG2, DSN1, ERAL1, GOLPH3, HLA-DPA1, IL17RA, KAT7, KCTD10, KCTD20, KDM3B, LAMTOR1, MAD1L1, MAP3K2, MCCC2, MPV17, MRPS109A, MSC, NAP1L4, PARG, PFDN6, PHF13, PKN1, PLEKHG6, PRDX5, PSMA7, RBMS10, ROCK1, S1000A13, SFSWAP, SIDT2, SNORA30, TIMM10, TMEM2, TMOD2, TMUB2, TRAPPC6A, TSFM, TUBGCP2, UQCRFS10. Among these genes related to menopausal women, it is preferable to be selected from 5 genes of CRAMP1L, DHRS7, NOP14, SGK3 and UQCRC1, and more preferably to include all of CRAMP1L, DHRS7, NOP14, SGK3 and UQCRC1.

[0068] The above 54 genes related to menopausal women are genes extracted as genes having a significant correlation between the expression level of their mRNA and the change in body fat in the data of a plurality of menopausal women sample subjects who ingested chlorogenic acids. Since these 54 genes have a significant correlation between the expression level of each mRNA and the change in body fat, it is possible to construct a prediction model for body fat change by using at least one gene as a feature amount. However, a more accurate prediction model can be constructed by combining a plurality of genes among the 54 genes as a feature amount. Furthermore, when the above 5 genes related to menopausal women are divided into a plurality of groups of a plurality of menopausal women sample subjects, they are genes extracted as genes having a significant correlation between the expression level of their mRNA and the change in body fat in any group. Thereby, it becomes possible to construct a more accurate prediction model by using one or a plurality, preferably all, of the above 5 genes as a feature amount.

[0069] Subsequently, the CPU 11 transmits (outputs) prediction result information including information regarding the predicted change in body fat and subject identification information (S110). In this step, the CPU 11 functions as an "output unit". In this specification, "output" is a concept that includes not only transmission of information via the Internet 50 and transmission of information to other electronic devices using short-range wireless communication or wired communication, but also any output of information by the display unit 16 or the speaker.

[0070] For example, the CPU 11 can transmit prediction result information including the predicted change in body fat to the subject terminal 200 based on the subject identification information. Examples of the information transmission method include, for example, e-mail, various messenger applications, SNS (Social Networking Service), and the notification function on the website that provides the above body fat change prediction service.

[0071] In the prediction result information, the information regarding the predicted change in body fat includes, for example, at least one piece of information selected from information indicating the prediction result or information related to the prediction result, and preferably includes information indicating the prediction result. Further, the prediction result information may include access information to a web page for displaying this information. Note that the prediction result information does not include information related to the treatment or diagnosis of diseases.

[0072] Examples of the information indicating the prediction result include information that directly or indirectly conveys the change in body fat due to the intake of chlorogenic acids in each subject. For example, the information indicating the prediction result may include the predicted amount of change in body fat itself, or may include the level of change in body fat that represents the predicted amount of change in body fat step by step. Examples of the level of change in body fat include, in addition to qualitative expressions such as "body fat is likely to decrease", "body fat is unlikely to decrease", and "body fat is likely to increase", numbers indicating classes, alphabets, symbols, and the like.

[0073] Examples of information related to the prediction results include information generated according to the prediction results of changes in body fat due to the intake of chlorogenic acids. For example, information recommending products containing chlorogenic acids or body fat reduction assistance services using products containing chlorogenic acids can be mentioned. The CPU 11 can generate information such that, for example, the higher the predicted amount of body fat reduction (or change level) due to the intake of chlorogenic acids, the higher the recommendation degree of the chlorogenic acid-containing product and / or the service using the same.

[0074] In the present embodiment, the CPU of the subject terminal 200 receives the above prediction result information (S203) and causes the display unit to display the prediction result information (S204). Thereby, the subject can view information regarding the prediction result of the change in his / her own body fat.

[0075] As described above, according to the system of the present embodiment, by using the above correlation information, it is possible to predict the change in body fat of each subject due to the intake of chlorogenic acids. Further, in the present embodiment, by determining the gender and the presence or absence of menopause of the subject based on the attribute information, it is possible to use the correlation information for each attribute corresponding to each of men, pre-menopausal women, and post-menopausal women. Thereby, as shown also in the following test examples, the system of the present embodiment can accurately predict the change in body fat of each subject due to the intake of chlorogenic acids.

[0076] [Modification Example] As a modification of this embodiment, without determining the presence or absence of menopause, the change in body fat due to the intake of chlorogenic acids in female subjects may be predicted using female-related information. For example, in the example shown in the flowchart of FIG. 5, when it is determined that the subject is not male (No in S105), the CPU 11 predicts the change in body fat due to the intake of chlorogenic acids in the subject from the gene expression information of the subject based on the female-related information (S111). Also in this step, the CPU 11 functions as a "prediction unit". In this embodiment, the female-related information includes, for example, a machine learning model having the expression levels of female-related genes or the like in a plurality of female sample subjects as explanatory variables and the change in body fat due to the intake of chlorogenic acids in the plurality of female sample subjects as an objective variable. For example, the CPU 11 can select a machine learning model related to female-related information from the correlation information database 32 and apply the gene expression information of the subject to the model to predict the change in body fat of the subject.

[0077] In this modification, the female-related gene is preferably at least one of 20 genes selected from, for example, ARHGAP23, C19orf71, COPS20, CRLF2, DEFB4A, DLG4, EGFR, EIF2B5, ERO1LB, HDAC6, MARCKSL1, MYO1G, NFATC2IP, NOP14, NT5DC1, PLS20, PVR, RGS10, SDR16C5, ZBTB8OS. From the viewpoint of improving the prediction accuracy, the female-related genes are preferably selected from two of them, C19orf71 and NOP14, and more preferably include both C19orf71 and NOP14. The change in body fat of female subjects can also be predicted by the female-related information using these female-related genes.

[0078] [Test Example] Next, an example of the generation process of the correlation information and an example of the prediction process of the change in body fat using the same in the above embodiment will be described. Of course, the present invention is not limited to the example using the correlation information and the prediction process according to the following test example.

[0079] <Construction of a regression model for predicting visceral fat changes in male subjects by ingestion of chlorogenic acids> 1) Test participants Fifty-nine healthy men aged 20 to 50 with a BMI of 22 to 32 were selected as test participants.

[0080] 2) Collection of skin surface lipids Before the ingestion test of the chlorogenic acid-containing beverage, skin surface lipids (SSL) containing RNA were collected from the entire face of the test participants using an oil-absorbing film (5 cm × 8 cm, made of polypropylene, 3M). The oil-absorbing film was transferred to a glass vial and stored at -80 °C until used for RNA extraction.

[0081] 3) Ingestion test of chlorogenic acid-containing beverage A chlorogenic acid-containing beverage (containing 270 mg of chlorogenic acids in the beverage per day) was used as the chlorogenic acid-containing composition. The test participants were made to continuously ingest one chlorogenic acid-containing beverage once a day for 10 weeks. The ingestion time of the chlorogenic acid-containing beverage was not limited, and the entire amount was ingested within 30 minutes from the start of ingestion. The test participants measured the visceral fat area of the abdomen on an empty stomach on the day of the start of ingestion and after 10 weeks of ingestion. A visceral fat meter, Panasonic's visceral fat meter EW-FA90 (medical device approval number 22500BZX00522000), was used. As a result of confirming the change in the visceral fat area of the test participants (visceral fat area after 10 weeks of ingestion - visceral fat area on the day of the start of ingestion), it was confirmed that the amount of change in the visceral fat area varied greatly among the test participants.

[0082] 4) Selection of subjects All 59 test participants were subjected to the following analysis using the oil-absorbing film containing SSL collected in 2).

[0083] 5) RNA preparation and sequencing The defatted film was cut into an appropriate size, and RNA was extracted using QIAzol Lysis Reagent (Qiagen) according to the attached protocol. Based on the extracted RNA, reverse transcription was performed at 42 °C for 90 minutes using the SuperScript VILO cDNA Synthesis kit (Life Technologies Japan, Ltd.) to synthesize cDNA. The random primer attached to the kit was used as the primer for the reverse transcription reaction. From the obtained cDNA, a library containing DNA derived from 20,802 genes was prepared by multiplex PCR. Multiplex PCR was performed using the Ion AmpliSeq Transcriptome Human Gene Expression Kit (Life Technologies Japan, Ltd.) under the conditions of [99 °C, 2 minutes → (99 °C, 15 seconds → 62 °C, 16 minutes) × 20 cycles → 4 °C, Hold]. The obtained PCR products were purified with Ampure XP (Beckman Coulter, Inc.), followed by buffer reconstruction, digestion of the primer sequence, adapter ligation and purification, and amplification to prepare a library. The prepared library was loaded onto an Ion 540 Chip and sequenced using an Ion S5 / XL system (Life Technologies Japan, Ltd.). The gene from which each read sequence was derived was determined by gene mapping of each read sequence obtained by sequencing using hg19 AmpliSeq Transcriptome ERCC v1, which is the reference sequence of the human genome.

[0084] 6) Data analysis and selection of related genes The read count of each read obtained from the sequencing of the subject's SSL-derived RNA acquired in the above 5) was used as data on the expression level of each RNA, and the count value (Normalized count value) corrected using DESeq2 was used for the analysis. However, reads with a read count of less than 1 were treated as missing values. The 59 samples were divided into training data for constructing a prediction model in machine learning and test data for verifying the accuracy of the prediction model at 80% (47 samples) and 20% (12 samples), respectively. For genes for which expression level data that were not missing values were obtained in 90% or more of the subjects in the training data, a correlation analysis between the expression information and the change in visceral fat area (visceral fat area after 10 weeks of intake - visceral fat area on the day of the start of intake) was performed using Spearman's correlation test, and genes with p < 0.05 and a correlation coefficient < 0.4, or p < 0.05 and a correlation coefficient > 0.4 were extracted. Twenty-six genes whose expression levels were correlated with the change in visceral fat area, as shown in Table 1, were obtained. These genes were selected as male-related genes to be used in the following prediction model construction.

[0085]

Table 1

[0086] 7) Model construction The expression level data of the related genes selected in 6) were converted to the logarithm to the base 2 value (Log 2 (RPM + 1) value) obtained by adding the integer 1 to approximate the normal distribution from the RPM value following the negative binomial distribution. The Log 2A regression model for predicting the change in visceral fat area was constructed using the (RPM + 1) value as an explanatory variable and the change in visceral fat area as a target variable. In the Python language, the algorithm of the support vector machine with a linear kernel was specified as a method, and the optimal values of the hyperparameters were tuned by 10-fold cross-validation. Using the hyperparameters determined by tuning, the algorithm of the support vector machine with a linear kernel was executed, and the root mean square error (RMSE) of the difference between the predicted value and the measured value was calculated. A smaller RMSE indicates a higher accuracy of the prediction model. As a result, the RMSE of the model using the expression data of 26 genes as variables was 1.363. Thus, as also shown in Fig. 6, it was shown that the change in visceral fat due to the intake of chlorogenic acids in male subjects can be predicted using the genes shown in Table 1 as male-related genes and the above prediction model as male-correlated information.

[0087] <Test Example 2 Construction of a Regression Model for Predicting Visceral Fat Changes in Male Subjects due to Intake of Chlorogenic Acids Using Different Related Genes> 1) Data Analysis and Selection of Related Genes Five split patterns were created by changing the breakdown of the training data (48 subjects) and test data (12 subjects) in 6) of Test Example 1. For each of the five training data sets (48 subjects), a correlation analysis with the change in visceral fat area was performed based on gene expression information in the same manner as in 6) using Spearman's correlation test, and genes with p < 0.05 and a correlation coefficient < 0.4, or p < 0.05 and a correlation coefficient > 0.4 were extracted. Among the genes extracted from the five training data sets, five genes shown in Table 2 below were obtained as genes commonly extracted by four or five of them. These five genes were selected as male-related genes to be used in the following prediction model construction.

[0088]

Table 2

[0089] 2) Model Construction 1) The expression level data of the related genes selected in (1) was converted to the logarithm value to the base 2 (Log 2 (RPM + 1) value) obtained by adding 1 to the integer and approximating from the RPM value following the negative binomial distribution to the normal distribution. Using the Log 2 (RPM + 1) values of the expression level data of the obtained 26 related genes as explanatory variables and the change amount of visceral fat area as the objective variable, a regression model for predicting the change amount of visceral fat area was constructed. In the Python language, the algorithm of the support vector machine with a linear kernel was specified as a method, and the optimal values of the hyperparameters were tuned by 10-fold cross-validation. Using the hyperparameters determined by tuning, the algorithm of the support vector machine with a linear kernel was executed, and the root mean square error (RMSE) of the square of the difference between the predicted value and the measured value was calculated. As a result, in the model using the expression level data of 5 genes as variables, the RMSE was 1.090. Thus, as shown in FIG. 7, it was shown that the genes shown in Table 2 can be used as male-related genes and the above prediction model can be used as male correlation information to predict the change in visceral fat due to the intake of chlorogenic acids in male subjects. Also, it was shown that the prediction model has higher accuracy compared to the prediction model shown in Test Example 1.

[0090] <Example of constructing a regression model for predicting visceral fat change in female subjects due to intake of chlorogenic acids in Test Example 3> 1) Test participants Sixty healthy women aged 20 to 50 with a BMI of 22 to 32 were selected as test participants.

[0091] 2) Collection of sebum Similar to 2) of Test Example 1, before the intake test of the chlorogenic acid-containing beverage, sebum containing RNA was collected from the entire face of the test participants and stored.

[0092] 3) Intake test of chlorogenic acid-containing beverage Similar to 3) of Test Example 1, a test on the intake of a chlorogenic acid-containing beverage and measurement of visceral fat area were conducted on the test participants. The change in visceral fat area (visceral fat area after 10 weeks of intake - visceral fat area on the day of the start of intake) for each test participant was calculated and used for subsequent processing.

[0093] 4) Selection of Subjects Excluding one dropout, the oil-absorbing film containing sebum collected in 2) from 59 test participants who completed the test was subjected to the following analysis.

[0094] 5) RNA Preparation and Sequencing Similar to 5) of Test Example 1, RNA was extracted and prepared from the oil-absorbing film and sequenced. By mapping each read sequence obtained by sequencing to the reference sequence of the human genome, hg19 AmpliSeq Transcriptome ERCC v1, the gene from which each read sequence was derived was determined.

[0095] 6) Comparison of the Effects of the Intake Test in Terms of the Presence or Absence of Menopause (Pre-menopause, Menopause) In the intake test of the chlorogenic acid-containing beverage in 3) above, it was confirmed that the amount of change in visceral fat area varied greatly among the test participants before and after 10 weeks of intake of the chlorogenic acid-containing beverage, but a significant difference was observed between pre-menopausal pre-menopausal women and menopausal women (Table 3). The numerical values in the table are expressed as mean ± standard deviation, and the significance test for the pre-menopausal vs. menopausal groups was performed by Unpaired t-test.

[0096]

Table 3

[0097] 7) Construction of a Prediction Model Therefore, a total of 59 subjects were divided into a pre-menopausal group (37 subjects) and a post-menopausal group (22 subjects) for model construction. The read count of each read obtained from the sequencing of SSL-derived RNA of the subjects obtained in 5) above was used as data on the expression level of each RNA, and the corrected count value (Normalized count value) using DESeq2 was used for the analysis. However, reads with a read count of less than 1 were treated as missing values.

[0098] i) Data analysis in pre-menopausal women and selection of related genes The 37 samples were divided into training data for constructing a prediction model in machine learning and test data for confirming the accuracy of the prediction model at 80% (30 samples) and 20% (7 samples), respectively. For genes for which expression level data that were not missing values were obtained in 90% or more of the subjects in the training data, a correlation analysis with the change in visceral fat area (change from before ingestion of the chlorogenic acid-containing composition to after 10 weeks of ingestion) was performed using Spearman's correlation test, and genes with a correlation coefficient < 0.5 or a correlation coefficient > 0.5 were extracted. A total of 96 genes shown in Table 4, whose expression levels were correlated with the change in visceral fat area, were obtained and selected as genes related to pre-menopausal women to be used in the following prediction model construction.

[0099]

Table 4

[0100] ii) Model construction in pre-menopausal women i) The expression level data of the related genes selected in i) were converted to the logarithm to the base 2 value (Log 2 (RPM + 1) value) obtained by adding 1 to the RPM value that follows a negative binomial distribution to approximate a normal distribution. The Log 2A regression model for predicting the change in visceral fat area was constructed using the (RPM + 1) value as an explanatory variable and the change in visceral fat area as a target variable. In the Python language, the algorithm of the support vector machine with a linear kernel was specified as a method, and the optimal values of the hyperparameters were tuned by 10-fold cross-validation. Using the hyperparameters determined by tuning, the algorithm of the support vector machine with a linear kernel was executed, and the root mean square error (RMSE) of the difference between the predicted value and the measured value was calculated. As a result, in the model using the expression data of 96 genes as explanatory variables, the RMSE was 0.755. Thus, as shown in FIG. 8, it was shown that the genes shown in Table 4 can be used to predict the change in visceral fat due to the intake of chlorogenic acids in pre-menopausal female subjects, with the above prediction model as pre-menopausal female correlation information. Also, from the value of the RMSE, it was shown that the above prediction model has higher accuracy compared to the prediction model using the data of all women (59 women) shown in Test Example 6 described later.

[0101] iii) Data analysis in post-menopausal women and selection of related genes Twenty-two samples were divided into 80% (17 samples) for learning data to construct a prediction model in machine learning and 20% (5 samples) for test data to confirm the accuracy of the prediction model, respectively. In the learning data, for genes for which expression data that was not a missing value was obtained in 90% or more of the subjects, correlation analysis with the change in visceral fat area (change from before ingestion of the chlorogenic acid-containing composition to after 10 weeks of ingestion) was performed using Spearman's correlation test, and genes with a correlation coefficient <0.6 or a correlation coefficient >0.6 were extracted. A total of 54 genes shown in Table 5, whose expression levels were correlated with the change in visceral fat area, were obtained and selected as post-menopausal female-related genes to be used in the following prediction model construction.

[0102]

Table 5

[0103] iv) Model construction in post-menopausal women iii) To approximate the expression level data of the selected related genes from the RPM values following a negative binomial distribution to a normal distribution, the base-2 logarithm value (Log 2 (RPM + 1) value) obtained by adding 1 to the integer was used. Using the Log 2 (RPM + 1) values of the expression level data of the obtained 54 related genes as explanatory variables and the change amount of visceral fat area as the objective variable, a regression model for predicting the change amount of visceral fat area was constructed. In the Python language, the algorithm of the support vector machine with a linear kernel was specified as the method, and the optimal values of the hyperparameters were tuned by 10-fold cross-validation. Using the hyperparameters determined by tuning, the algorithm of the support vector machine with a linear kernel was executed, and the root mean square error (RMSE) of the square of the difference between the predicted value and the measured value was calculated. As a result, in the model using the expression level data of 54 genes as variables, the RMSE was 0.649. Thus, as shown in Fig. 9, it was shown that the genes shown in Table 5 can be used as menopause-related genes and the above prediction model can be used as menopause-related information to predict the change in visceral fat due to the intake of chlorogenic acids in menopause female subjects. Also, from the value of the RMSE, it was shown that the above prediction model has higher accuracy compared to the prediction model using the data of all women (59 women) shown in Test Example 6 described later.

[0104] <Example of constructing a regression model for predicting visceral fat changes in female subjects due to intake of chlorogenic acids using different related genes> i) Data analysis and selection of related genes in premenopausal women Thirty-seven samples were divided into training data for constructing a prediction model in machine learning and test data for verifying the accuracy of the prediction model at 80% (30 samples) and 20% (7 samples), respectively. For genes in the training data for which expression level data that were not missing values were obtained in 90% or more of the subjects, correlation analysis with the amount of change in visceral fat area (the amount of change from before ingestion of the chlorogenic acid-containing composition to after 10 weeks of ingestion) was performed using Spearman's correlation test, and genes with a correlation coefficient < 0.5 or a correlation coefficient > 0.5 were extracted. Among these, when the division pattern of the test data and the training data was changed, the three genes shown in Table 6 that were commonly extracted five times out of five division patterns were selected as genes related to premenopausal women to be used in the following prediction model construction.

[0105]

Table 6

[0106] ii) Model construction in premenopausal women i) The expression level data of the related genes selected in i) were converted to the logarithm to the base 2 value (Log 2 (RPM + 1) value) obtained by adding 1 to the integer to approximate the normal distribution from the RPM value following the negative binomial distribution. The Log 2A regression model for predicting the change in visceral fat area was constructed using the (RPM + 1) value as an explanatory variable and the change in visceral fat area as a target variable. In the Python language, the algorithm of the support vector machine with a linear kernel was specified as a method, and the optimal values of the hyperparameters were tuned by 10-fold cross-validation. Using the hyperparameters determined by tuning, the algorithm of the support vector machine with a linear kernel was executed, and the root mean square error (RMSE) of the difference between the predicted value and the measured value was calculated. As a result, the RMSE was 0.467 in the model when the expression level data of 3 genes were used as variables. Thus, as shown in Fig. 10, it was shown that the genes shown in Table 6 can be used to predict the change in visceral fat due to the intake of chlorogenic acids in pre-menopausal female subjects, with the said genes being pre-menopausal female-related genes and the said prediction model being pre-menopausal female correlation information. Also, from the value of the RMSE, it was shown that the said prediction model has higher accuracy compared to the prediction model constructed with 96 genes in Test Example 3 and the prediction model using the data of all women (59 women) shown in Test Example 6 described later.

[0107] iii) Data analysis in post-menopausal women and selection of related genes Twenty-two samples were divided into training data for constructing a prediction model in machine learning and test data for confirming the accuracy of the prediction model at 80% (17 samples) and 20% (5 samples), respectively. In the training data, for genes for which expression level data that were not missing values were obtained in 90% or more of the subjects, correlation analysis with the change in visceral fat area (change from before intake of the chlorogenic acid-containing composition to after 10 weeks of intake) was performed using Spearman's correlation test, and genes with a correlation coefficient < 0.6 or a correlation coefficient > 0.6 were extracted. Among these, the 5 genes shown in Table 7 that were commonly extracted 5 times out of 5 times when the division pattern of the test data and the training data was changed were selected as post-menopausal female-related genes to be used in the following prediction model construction.

[0108]

Table 7

[0109] iv) Model construction in postmenopausal women iii) To approximate the expression level data of the related genes selected in iii) from the RPM values following the negative binomial distribution to the normal distribution, the logarithm to the base 2 value (Log 2 (RPM + 1) value) added with the integer 1 was converted. Using the Log 2 (RPM + 1) values of the obtained expression level data of the 5 related genes as the explanatory variables and the change amount of the visceral fat area as the objective variable, a regression model for predicting the change amount of the visceral fat area was constructed. In the Python language, the algorithm of the support vector machine with a linear kernel was specified as the method, and the optimal values of the hyperparameters were tuned by 10-fold cross-validation. Using the hyperparameters determined by the tuning, the algorithm of the support vector machine with a linear kernel was executed, and the root mean square error (RMSE) of the square of the difference between the predicted value and the measured value was calculated. As a result, the RMSE in the model using the expression level data of the 5 genes as variables was 0.608. Thus, as shown in Fig. 11, it was shown that the genes shown in Table 6 could be used as postmenopausal women-related genes, and the prediction model could be used as postmenopausal women correlation information to predict the change in visceral fat due to the intake of chlorogenic acids in postmenopausal female subjects. Also, from the value of RMSE, it was found that the prediction model had higher accuracy compared to the prediction model constructed with 54 genes in Test Example 3 and the prediction model using the data of all women (59 women) shown in Test Example 6 described later.

[0110] <Example of constructing a regression model for predicting the change in visceral fat of female subjects due to the intake of chlorogenic acids using one type of related gene in Test Example 5> i) Selection of related genes and model construction in premenopausal women As the expression level data in premenopausal women, the same data as in i) of Test Example 4 was used, and 1 gene was selected from the 3 genes shown in Table 6 as the premenopausal-related gene in the following prediction model construction. To approximate the expression level data of the selected premenopausal women-related genes from the RPM values following the negative binomial distribution to the normal distribution, the logarithm to the base 2 value (Log 2(Converted to the (RPM + 1) value). Log of the expression level data of the obtained related gene of 1 2 Using the (RPM + 1) value as the explanatory variable and the change amount of visceral fat area as the objective variable, a regression model for predicting the change amount of visceral fat area was constructed. In the Python language, the algorithm of the support vector machine with a linear kernel was specified as the method, and the optimal values of the hyperparameters were tuned by 10-fold cross-validation. Using the hyperparameters determined by tuning, the algorithm of the support vector machine with a linear kernel was executed, and the root mean square error (RMSE) of the square of the difference between the predicted value and the measured value was calculated. As a result, in the model when using the expression level data of one gene, C19orf71, among the three genes in Table 6 as a variable, the RMSE was 0.612. Thereby, as also shown in FIG. 12, it was shown that the change in visceral fat due to the intake of chlorogenic acids in premenopausal female subjects can be predicted by the correlation information using one premenopausal female-related gene.

[0111] ii) Selection of related genes and model construction in postmenopausal women As the expression level data in postmenopausal women, the same data as in iii) of Test Example 4 was used, and one gene was selected from the five genes shown in Table 7 as the postmenopausal female-related gene in the following prediction model construction. The expression level data of the selected postmenopausal female-related gene was converted to the logarithm value to the base 2 (Log 2 (RPM + 1) value) for approximating the normal distribution from the RPM value following the negative binomial distribution by adding the integer 1. Log of the expression level data of the obtained related gene of 1 2A regression model for predicting the change in visceral fat area was constructed using the (RPM + 1) value as an explanatory variable and the change in visceral fat area as a target variable. In the Python language, the algorithm of the support vector machine with a linear kernel was specified as a method, and the optimal values of the hyperparameters were tuned by 10-fold cross-validation. Using the hyperparameters determined by tuning, the algorithm of the support vector machine with a linear kernel was executed, and the root mean square error (RMSE) of the difference between the predicted value and the measured value was calculated. As a result, in the model using the expression level data of one gene, CRAMP1L, among the five genes in Table 6 as a variable, the RMSE was 0.601. Thus, as also shown in Fig. 13, it was shown that the change in visceral fat due to the intake of chlorogenic acids in postmenopausal female subjects can be predicted by the correlation information using one postmenopausal female-related gene.

[0112] <Test Example 6 Construction of a Prediction Model Using All Female Data> i) Data Analysis and Selection of Related Genes A total of 59 samples were divided into training data for constructing a prediction model in machine learning and test data for confirming the accuracy of the prediction model at 80% (47 samples) and 20% (12 samples), respectively. For genes with expression level data that were not missing values in 90% or more of the subjects in the training data, a correlation analysis with the change in visceral fat area (visceral fat area after 10 weeks of intake - visceral fat area on the day of the start of intake) was performed using Spearman's correlation test based on the expression information, and genes with a correlation coefficient < 0.4 or a correlation coefficient > 0.4 were extracted. Twenty genes whose expression levels were correlated with the change in visceral fat area, as shown in Table 8, were obtained. Among these, when the division pattern of the test data and the training data was changed, two genes shown in Table 9 that were commonly extracted in all five division patterns were obtained. Each of the extracted genes was selected as a female-related gene to be used in the following prediction model construction.

[0113]

Table 8

[0114] [Table 9]

[0115] ii) Model construction In order to approximate the expression level data of the related genes selected in i) from the RPM values following the negative binomial distribution to the normal distribution, the logarithm to the base 2 value (Log 2 (RPM + 1) value) added with the integer 1 was converted. The Log 2 (RPM + 1) values of the expression level data of 20 or 2 related genes obtained were used as explanatory variables, and the change amount of visceral fat area was used as the objective variable to construct a regression model for predicting the change amount of visceral fat area. In the Python language, the algorithm of the support vector machine with a linear kernel was specified as the method, and the optimal values of the hyperparameters were tuned by 10-fold cross-validation. Using the hyperparameters determined by tuning, the algorithm of the support vector machine with a linear kernel was executed, and the root mean square error (RMSE) of the square of the difference between the predicted value and the measured value was calculated. As a result, in the model using the expression level data of the 20 genes in Table 8 as variables, RMSE was 0.998 (Figure 14). On the other hand, in the model using the expression level data of the 2 genes in Table 9 as variables, RMSE was 0.674 (Figure 15). From these results, it was shown that the above prediction model can predict the change in visceral fat due to the intake of chlorogenic acids in female subjects.

[0116] [Other Embodiments] As described above, the embodiments of the present invention have been explained. However, the present invention is not limited only to the above-described embodiments, and it goes without saying that various changes can be made without departing from the gist of the present invention.

[0117] In the above-described embodiment, an example of predicting the amount of change in body fat in a subject has been described, but the present invention is not limited thereto. For example, a related gene may be a gene (expression-variable gene) in which the expression levels are detected to be different in each of a plurality of groups when a plurality of sample subjects are classified into a plurality of groups according to the level of change in body fat. In this case, the correlation information may be information indicating the correlation between the level of change in body fat corresponding to each of the plurality of groups and the expression level of the related gene or its expression product.

[0118] In this example, the correlation information may be a machine learning model, and specifically, it may be a classification model for classifying the level of change in body fat corresponding to each of the plurality of groups using the expression level of the related gene or its expression product. The algorithm used for generating such a classification model is not particularly limited, and examples thereof include logistic regression, random forest, support vector machine, decision tree, k-nearest neighbor method, neural network, and the like.

[0119] A specific example of a method for generating correlation information will be described. First, similar to the above-described embodiment, data on the amount of change in body fat of a plurality of sample subject groups that have ingested chlorogenic acids for a predetermined period is obtained. This data is classified, for example, into a body fat reduction (effective) group and a body fat non-reduction (ineffective) group, and genes in which the expression levels of related genes and the like are significantly different between these groups are extracted, and the extracted genes are used as related genes. Then, a machine learning model is generated with the expression levels of related genes and the like in a plurality of sample subjects as explanatory variables and the level of change in body fat (body fat reduction (effective) or body fat non-reduction (ineffective)) due to the ingestion of chlorogenic acids by the plurality of sample subjects as the objective variable.

[0120] Also in this example, similar to the above-described embodiments, an acquisition step is performed, and by using the above classification model in the prediction step, it is possible to predict the change level of body fat due to the intake of chlorogenic acids by the subject. In the above specific example, the data of the sample subject population was classified into two groups, but it is not limited to this, and it may be classified into three or more groups according to the body fat change level. Further, the prediction model is not limited to a supervised machine learning model, and may be a classification model that performs cluster analysis based on the expression level of related genes or the like.

[0121] Also, the correlation information may include, as information showing the correlation with the change in body fat in the sample subjects who ingested chlorogenic acids, in addition to the expression level of related genes or the like, information having a relevance with other changes in body fat. Examples of the other relevant information include information related to the physical information and lifestyle habits of the subject. In this case, the machine learning model included in the correlation information can use, for example, the expression levels of related genes or their expression products in a plurality of sample subjects, and information related to the physical information and lifestyle habits of the subject as explanatory variables, and the change in body fat due to the intake of chlorogenic acids in a plurality of sample subjects as the target variable. By using such correlation information in the prediction process, the prediction accuracy of the change in body fat can be improved.

[0122] In addition, the attributes to be determined may be attributes other than gender and the presence or absence of menopause. Examples of such attributes include age (age group), past medical history, current medical history, and the like. Further, the attributes of the subject may be classified by combining gender and / or the presence or absence of menopause with other attributes. Even when using other attributes, similar to the above-described embodiment, the sample subject population is classified by a predetermined attribute, and from the data of the sample subject population having the predetermined attribute, genes with a high correlation between the change in body fat and the expression level (genes whose expression level correlates with the amount of change in body fat, expression fluctuation genes, etc.) are extracted, whereby relevant genes serving as feature amounts can be extracted. Furthermore, in the sample subject population having a predetermined attribute, information indicating the correlation between the change in body fat and the expression level regarding relevant genes and the like is generated by a method such as machine learning, whereby correlation information corresponding to the attribute can be generated.

[0123] In addition, the CPU 11 may predict the change in the body fat of the subject without determining the attribute information. For example, in the case of a body fat change prediction service that limits the subject of the subject to a specific gender (male or female), as shown in FIG. 16, the CPU 11 may predict the change in the body fat of the subject based on the correlation information corresponding to the specific gender without performing the determination step.

[0124] In addition, the method for acquiring subject information including the attribute information of the subject is not limited to the above example. For example, as another embodiment, the subject information is not limited to being acquired from the subject terminal 200, and may be acquired from a server or the like that provides other web services. Regarding the acquisition of the personal information of the subject, it shall be appropriately performed based on the laws of the region where the present invention is implemented.

[0125] Also, the gene expression information is not limited to the example of being acquired from the measurement terminal 300, and may be acquired from another computer that owns the gene expression information of the subject (for example, a server managed by another operator that provides a gene analysis service, etc.). In this case, as shown in FIG. 17, the system does not include the measurement terminal 300, and the server 100 may acquire the gene expression information generated from the biological sample of the user by the computer used by another operator that provides a gene analysis service or the like. Also, as another example, the subject may provide his or her own gene expression information provided by a gene analysis service or the like to the server 100 using the subject terminal 200.

[0126] In connection with these, in the above-described embodiment, the server 100 acquired the subject information (attribute information) and the gene expression information from different terminals (computers) at different timings, but they may be acquired simultaneously from the same terminal. For example, as shown in the flowchart of FIG. 18, the server 100 receives the subject information including the attribute information and the gene expression information in association with each other (S112), and stores them in the subject information database 31 (S113). The subject information including the attribute information and the gene expression information may be transmitted, for example, from a computer used by another operator or the like, or may be transmitted from the subject terminal 200 or the measurement terminal 300.

[0127] Also, the output method of the prediction result information is not limited to the above example. For example, the server 100 can also output the prediction result information to an external device such as a printer in order to provide a paper medium such as a direct mail.

[0128] In the above-described embodiment, an example in which a prediction result information transmission (output) step is performed after the prediction step has been shown, but the timing of the output step is not particularly limited. For example, the server 100 may transmit the prediction result information after receiving an inquiry about the prediction result from the subject terminal 200. As a specific example, as shown in the flowchart of FIG. 19, the CPU 11 of the server 100 stores the predicted body fat change in the subject information database 31 in association with the subject identification information after the prediction steps (S106, S108, S109) (S114). The CPU 11 determines whether it has received an inquiry about the prediction result from the subject terminal 200 (S115). If it is determined that the inquiry has been received (Yes in S115), the prediction result information is transmitted (S110). In this example, if an inquiry from the subject terminal 200 has not been received (No in S115), it is determined whether an inquiry has been received again (S115). Further, if the CPU 11 has not received an inquiry for a certain period of time, it may end the process without transmitting the prediction result information.

[0129] In the above embodiments, an example in which the correlation information consists of a machine learning model has been described, but it is not limited thereto. For example, the correlation information may be a table that stores the correspondence between the expression level of a related gene and the change in body fat. Particularly when the number of related genes is small, the change in body fat can also be predicted from the expression level of the related gene by such a table.

[0130] In the above embodiments, an example of RNA information has been given as the gene expression information, but it is not limited thereto. The gene expression information may include information on the expression levels of molecules selected from, for example, DNA, proteins encoded by RNA, molecules that interact with the protein, molecules that interact with the RNA, or molecules that interact with the DNA, as related genes or their expression products.

[0131] In the above embodiments, an example was shown in which the server 100 performs steps such as the step of acquiring subject information and gene expression information, and the step of predicting body fat change. However, each step may be performed by another computer. That is, the functional blocks such as the acquisition unit, determination unit, prediction unit, output unit, etc. of the system according to the present invention may be realized by different computers respectively.

[0132] Among the inventions described in the claims of the present application, the invention described as "method for predicting body fat change" is such that each of its steps is automatically performed by at least one device such as a computer through information processing by software, and is not performed by a human using a device such as a computer. That is, the "method for predicting body fat change" is an information processing method by computer software, and is not a method of a human operating a computing tool called a computer.

[0133] As yet another embodiment, the present invention can provide a prediction marker for predicting the change in body fat of a subject by ingestion of chlorogenic acids. The prediction marker according to one embodiment is a prediction marker for predicting the change in body fat of a male subject by ingestion of chlorogenic acids, and is composed of, for example, at least one gene selected from FOXD4L6, KCTD18, TCHP, THOC3, TMEM57, C19orf53, CDC40, CMC2, CRCP, DDX19B, DHX57, DRG2, EIF2A, FABP3, MCC, NFE2L3, NOD2, PIGX, PSMC6, R3HCC1L, RNF7, RNPEPL1, SLC11A1, TRIM23, UBE2L6, and UNC13D or its expression product. From the viewpoint of further improving the prediction accuracy, the prediction marker preferably consists of at least one gene selected from FOXD4L6, KCTD18, TCHP, THOC3, and TMEM57 or its expression product, and more preferably consists of FOXD4L6, KCTD18, TCHP, THOC3, and TMEM57 or their expression products.

[0134] A prediction marker according to still another embodiment is a prediction marker that predicts changes in body fat of premenopausal female subjects due to the intake of chlorogenic acids, and includes, for example, at least one gene selected from C19orf71, DDX39A, PTPN1, ALDH9A1, ANKRD33B, APBA3, ARHGEF2, ARID5B, ARRDC2, ATAD3B, B4GALT1, BIRC3, BRMS1, C10orf128, GINM1, CC2D1A, CCD101, CCD130, CCD71, CDK5, CHST7, COG3, COX6A1, CSNK2B, DDX17, ELF4, ENC1, ENGASE, ERAP2, EXOSC8, FAM110A, FAM210A, FCGBP, FCGR2C, FOXJ3, FZD5, FZD7, GABPB1, GGA3, GRN, H2AFY2, HSPA7, HSPA9, ITGA5, KCNAB2, KIAA0020, KIF1B, LIMA1, LOC100132247, LOC100289019, LOC595101, MARCKSL1, MFSD10, MGST3, MIB2, MLLT6, MYO1G, NGLY1, P2RX4, PIK3C3, PMS2P3, POLG, PPCS, PRICKLE3, PRKAA1, PTPN2, PVR, RAB31, RAB7L1, RCE1, RNF113A, RRAGC, SLC5A6, TACC1, TBC1D9B, TCIRG1, TMEM176B, TNFRSF14, TRAF3, VEGFA, XPC, YEATS20, ZFAND2A, ZFP36L2, ZFYVE26, ZNF101, ZNF33A, ZNF350, ZNF783, ABCA11P, FAM223B, LOC100271836, LOC100288778, LOC728875, RPS15AP10, and SH3GL1P2, or its expression product. From the viewpoint of further improving the prediction accuracy, the prediction marker preferably includes at least one gene selected from C19orf71, DDX39A, and PTPN1, or its expression product, and more preferably consists of C19orf71, DDX39A, and PTPN1, or their expression products.

[0135] A prediction marker according to still another embodiment is a prediction marker that predicts changes in body fat of postmenopausal female subjects due to the intake of chlorogenic acids, and for example, at least one gene selected from CRAMP1L, DHRS7, NOP14, SGK3, UQCRC1, ABCC3, ALDH2, ARHGAP27, BTN3A2, C17orf79, C6orf89, C7orf13, CARD11, CCDC23, DNPEP, DSG2, DSN1, ERAL1, GOLPH3, HLA-DPA1, IL17RA, KAT7, KCTD10, KCTD20, KDM3B, LAMTOR1, MAD1L1, MAP3K2, MCCC2, MPV17, MRPS18A, MSC, NAP1L4, PARG, PFDN6, PHF13, PKN1, PLEKHG6, PRDX5, PSMA7, RBMS1, ROCK1, S100A13, SFSWAP, SIDT2, SNORA30, TIMM10, TMEM2, TMOD2, TMUB2, TRAPPC6A, TSFM, TUBGCP2, UQCRFS1 or an expression product thereof. From the viewpoint of further improving the prediction accuracy, the prediction marker preferably consists of at least one gene selected from CRAMP1L, DHRS7, NOP14, SGK3 and UQCRC1 or an expression product thereof, and more preferably consists of CRAMP1L, DHRS7, NOP14, SGK3 and UQCRC1 or their expression products.

[0136] Furthermore, the present invention can provide a body fat change prediction kit used for predicting the body fat change of male subjects by the intake of chlorogenic acids. The body fat change prediction kit contains reagents for detecting the expression level of at least one target gene selected from, for example, FOXD4L6, KCTD18, TCHP, THOC3, TMEM57, C19orf53, CDC40, CMC2, CRCP, DDX19B, DHX57, DRG2, EIF2A, FABP3, MCC, NFE2L3, NOD2, PIGX, PSMC6, R3HCC1L, RNF7, RNPEPL1, SLC11A1, TRIM23, UBE2L6, and UNC13D or its expression product from a sample collected from a male subject. From the viewpoint of further improving the prediction accuracy, the target gene is preferably at least one selected from FOXD4L6, KCTD18, TCHP, THOC3, and TMEM57, and more preferably FOXD4L6, KCTD18, TCHP, THOC3, and TMEM57.

[0137] Similarly, the present invention can provide a body fat change prediction kit used for predicting the body fat change of premenopausal female subjects by the intake of chlorogenic acids. The body fat change prediction kit detects, from a sample collected from a premenopausal female subject, the expression level of at least one target gene selected from C19orf71, DDX39A, PTPN1, ALDH9A1, ANKRD33B, APBA3, ARHGEF2, ARID5B, ARRDC2, ATAD3B, B4GALT1, BIRC3, BRMS1, C10orf128, GINM1, CC2D1A, CCDC101, CCDC130, CCDC71, CDK5, CHST7, COG3, COX6A1, CSNK2B, DDX17, ELF4, ENC1, ENGASE, ERAP2, EXOSC8, FAM110A, FAM210A, FCGBP, FCGR2C, FOXJ3, FZD5, FZD7, GABPB1, GGA3, GRN, H2AFY2, HSPA7, HSPA9, ITGA5, KCNAB2, KIAA0020, KIF1B, LIMA1, LOC100132247, LOC100289019, LOC595101, MARCKSL1, MFSD10, MGST3, MIB2, MLLT6, MYO1G, NGLY1, P2RX4, PIK3C3, PMS2P3, POLG, PPCS, PRICKLE3, PRKAA1, PTPN2, PVR, RAB31, RAB7L1, RCE1, RNF113A, RRAGC, SLC5A6, TACC1, TBC1D9B, TCIRG1, TMEM176B, TNFRSF14, TRAF3, VEGFA, XPC, YEATS20, ZFAND2A, ZFP36L2, ZFYVE26, ZNF101, ZNF33A, ZNF350, ZNF783, ABCA11P, FAM223B, LOC100271836, LOC100288778, LOC728875, RPS15AP10, and SH3GL1P2 or a reagent for detecting the expression product thereof. From the viewpoint of further improving the prediction accuracy, the target gene is preferably at least one selected from C19orf71, DDX39A, and PTPN1, and more preferably C19orf71, DDX39A, and PTPN1.

[0138] Similarly, the present invention can provide a body fat change prediction kit used for predicting the body fat change of menopausal female subjects by the intake of chlorogenic acids. The body fat change prediction kit detects, from a sample collected from a menopausal female subject, for example, the expression level of at least one target gene selected from CRAMP1L, DHRS7, NOP14, SGK3, UQCRC1, ABCC3, ALDH2, ARHGAP27, BTN3A2, C17orf79, C6orf89, C7orf13, CARD11, CCDC23, DNPEP, DSG2, DSN1, ERAL1, GOLPH3, HLA-DPA1, IL17RA, KAT7, KCTD10, KCTD20, KDM3B, LAMTOR1, MAD1L1, MAP3K2, MCCC2, MPV17, MRPS18A, MSC, NAP1L4, PARG, PFDN6, PHF13, PKN1, PLEKHG6, PRDX5, PSMA7, RBMS1, ROCK1, S100A13, SFSWAP, SIDT2, SNORA30, TIMM10, TMEM2, TMOD2, TMUB2, TRAPPC6A, TSFM, TUBGCP2, UQCRFS1 or its expression product, and contains reagents therefor. From the viewpoint of further improving the prediction accuracy, the target gene is preferably at least one selected from CRAMP1L, DHRS7, NOP14, SGK3 and UQCRC1, and more preferably CRAMP1L, DHRS7, NOP14, SGK3 and UQCRC1.

[0139] Reagents for detecting the expression level of a related gene or its expression product include, for example, reagents for extracting and purifying RNA from the collected SSL, oligonucleotides (e.g., primers for PCR, adapter sequences for sequencing, etc.) that specifically bind (hybridize) to nucleic acids derived from the target gene, reagents for nucleic acid amplification or hybridization, reagents for immunological measurement including antibodies that recognize gene expression products (proteins), as well as labeling reagents, buffers, chromogenic substrates, secondary antibodies, blocking agents, control reagents used as positive controls and negative controls, instruments necessary for the test, and in addition, indicators or guidance for detecting the expression level of the target gene or its expression product.

[0140] In addition, the kit for predicting body fat change according to the present invention may contain, in addition to the above reagents, tools and reagents necessary for collecting and storing biological samples such as SSL. For example, the tools and reagents necessary for collecting and storing SSL include a fat extraction film for collecting SSL, a reagent for storing the collected SSL, a storage container, and the like.

Explanation of symbols

[0141] 11…CPU 18…Storage unit 19…Communication unit 31…Subject information database 32…Correlation information database 100…Server 200…Subject terminal 300…Measurement terminal

Claims

1. A method for predicting changes in body fat of a subject by computer, which predicts changes in body fat of the subject due to ingestion of chlorogenic acids, comprising: obtaining gene expression information including information on the expression level of at least one related gene or its expression product related to changes in body fat due to ingestion of the chlorogenic acids, in association with subject identification information for identifying the subject; predicting changes in body fat of the subject due to ingestion of the chlorogenic acids from the gene expression information based on correlation information showing the correlation between changes in body fat in a plurality of sample subjects who have ingested the chlorogenic acids and the expression level of the related gene or its expression product related to the changes in body fat. A method for predicting changes in body fat.

2. The correlation information includes: a machine learning model having, as an explanatory variable, the expression level of the related gene or its expression product in the plurality of sample subjects, and, as an objective variable, the change in body fat of the plurality of sample subjects due to ingestion of the chlorogenic acids. The method for predicting changes in body fat according to Claim 1.

3. The expression level of the related gene or its expression product has a correlation with the amount of change in body fat in the plurality of sample subjects, the correlation information is information showing the correlation between the amount of change in body fat in the plurality of sample subjects and the expression level of the related gene or its expression product, and predicting the amount of change in body fat of the subject due to ingestion of the chlorogenic acids from the gene expression information based on the correlation information. The method for predicting changes in body fat according to Claim 1 or 2.

4. The related gene is a gene in which different expression levels are detected in each of the plurality of groups when the plurality of sample subjects are classified into a plurality of groups according to the level of change in body fat, the correlation information includes: information showing the correlation between the level of change in body fat corresponding to each of the plurality of groups and the expression level of the related gene or its expression product, and predicting the level of change in body fat of the subject due to ingestion of the chlorogenic acids from the gene expression information based on the correlation information. The method for predicting changes in body fat according to Claim 1 or 2.

5. The correlation information includes a plurality of piecemeal correlation information associated with each of a plurality of attributes that the subject can have. Each of the plurality of attribute-specific correlation informations shows the relationship between the change in body fat in a plurality of sample subjects having the corresponding attribute among the plurality of sample subjects, and the expression level of the related gene or its expression product related to the change in body fat, acquire the attribute information of the subject in association with the subject identification information, judge the attribute of the subject based on the acquired attribute information of the subject, predict the change in body fat due to the intake of chlorogenic acids in the subject from the gene expression information based on the attribute-specific correlation information corresponding to the judged attribute The method for predicting a change in body fat according to claim 1 or 2.

6. The attribute information of the subject includes gender information, The plurality of attribute-specific correlation informations are male correlation information showing the correlation between the change in body fat in a plurality of male sample subjects among the plurality of sample subjects and the expression level of the male-related gene or its expression product as the related gene related to the change in body fat, female correlation information showing the correlation between the change in body fat in a plurality of female sample subjects among the plurality of sample subjects and the expression level of the female-related gene or its expression product as the related gene related to the change in body fat, judge the gender of the subject based on the acquired attribute information of the subject, when it is judged that the gender of the subject is male, predict the change in body fat due to the intake of chlorogenic acids in the subject from the gene expression information based on the male correlation information, when it is judged that the gender of the subject is female, predict the change in body fat due to the intake of chlorogenic acids in the subject from the gene expression information based on the female correlation information The method for predicting a change in body fat according to claim 5.

7. The male-related gene is selected from FOXD4L6, KCTD18, TCHP, THOC3, TMEM57, C19orf53, CDC40, CMC2, CRCP, DDX19B, DHX57, DRG2, EIF2A, FABP3, MCC, NFE2L3, NOD2, PIGX, PSMC6, R3HCC1L, RNF7, RNPEPL1, SLC11A1, TRIM23, UBE2L6 and UNC13D The method for predicting a change in body fat according to claim 6.

8. The male-related gene is selected from FOXD4L6, KCTD18, TCHP, THOC3 and TMEM57 The method for predicting body fat change according to claim 7.

9. The attribute information of the subject includes, when the gender of the subject is female, information on the presence or absence of menopause or menopause-related information that can infer the presence or absence of menopause. The plurality of correlation information by attribute includes, as the female correlation information, The correlation between the change in body fat in a plurality of sample subjects who are pre-menopausal menstruating women among the plurality of sample subjects and the expression level of the menstruating woman-related gene or its expression product as the related gene related to the change in body fat, which is menstruating woman correlation information. The correlation between the change in body fat in a plurality of sample subjects who are post-menopausal women among the plurality of sample subjects and the expression level of the post-menopausal woman-related gene or its expression product as the related gene related to the change in body fat, which is post-menopausal woman correlation information. When it is determined that the gender of the subject is female, further determine the presence or absence of menopause of the subject based on the menopause-related information. When the subject is determined to be a menstruating woman, predict the change in body fat due to the intake of chlorogenic acids by the subject from the gene expression information based on the menstruating woman correlation information. When the subject is determined to be a post-menopausal woman, predict the change in body fat due to the intake of chlorogenic acids by the subject from the gene expression information based on the post-menopausal woman correlation information. The method for predicting body fat change according to claim 6.

10. The menstruating woman-related gene is Selected from C19orf71, DDX39A, PTPN1, ALDH9A1, ANKRD33B, APBA3, ARHGEF2, ARID5B, ARRDC2, ATAAD3B, B4GALT1, BIRC3, BRMS1, C10orf128, GINM1, CC2D1A, CCD101, CCD130, CCD71, CDK5, CHST7, COG3, COX6A1, CSNK2B, DDX17, ELF4, ENC1, ENGASE, ERAP2, EXOSC8, FAM110A, FAM210A, FCGBP, FCGPR2C, FOXJ3, FZD5, FZD7, GABPB1, GGA3, GRN, H2AFY2, HSP7, HSP9, ITGA5, KCNAB2, KIAA0020, KIF1B, LIMA1, LOC100132247, LOC100289019, LOC595101, MARCKSL1, MFSD10, MGST3, MIB2, MLLT6, MYO1G, NGLY1, P2RX4, PIK3C3, PMS2P3, POLG, PPCS, PRICKLE3, PRKAA1, PTPN2, PVR, RAB31, RAB7L1, RCE1, RNF113A, RRAGC, SLC5A6, TACC1, TBC1D9B, TCIRG1, TMEM176B, TNFRSF14, TRAF3, VEGFA, XPC, YEATS2, ZFAND2A, ZFP36L2, ZFYVE26, ZNF101, ZNF33A, ZNF350, ZNF783, ABCA11P, FAM223B, LOC100271836, LOC100288778, LOC728875, RPS15AP10 and SH3GL1P2 The method for predicting body fat change according to claim 9.

11. The genes related to premenopausal women are Selected from C19orf71, DDX39A and PTPN1 The method for predicting body fat change according to claim 10.

12. The genes related to postmenopausal women are Selected from CRAMP1L, DHRS7, NOP14, SGK3, UQCRC1, ABCC3, ALDH2, ARHGAP27, BTN3A2, C17orf79, C6orf89, C7orf13, CARD11, CCDC23, DNPEP, DSG2, DSN1,ERAL1, GOLPH3, HLA-DPA1, IL17RA, KAT7, KCTD10, KCTD20, KDM3B, LAMTOR1, MAD1L1, MAP3K2, MCC2, MPV17, MRPS18A, MSC, NAP1L4, PARG, PFDAN6, PHF13, PKN1, PLEKHG6, PRDX5, PSMA7, RBMS1, ROCK1, S100A13, SFSWAP, SIDT2, SNORA30, TIMM10, TMEM2, TMOD2, TMUB2, TRAPPC6A, TSFM, TUBGCP2 and UQCRFS1 The method for predicting body fat change according to claim 9.

13. The menopause-related gene is Selected from CRAMP1L, DHRS7, NOP14, SGK3 and UQCRC1 The method for predicting body fat change according to claim 12.

14. Further output prediction result information including the predicted change in body fat and the subject identification information The method for predicting body fat change according to claim 1 or 2.

15. The gene expression information includes information on the expression level of mRNA of the related gene The method for predicting body fat change according to claim 1 or 2.

16. The change in body fat includes the change in visceral fat The method for predicting body fat change according to claim 1 or 2.

17. A body fat change prediction system for predicting the change in body fat of a subject due to the intake of chlorogenic acids, comprising A storage unit that stores correlation information indicating the correlation between the change in body fat in a plurality of sample subjects who have ingested the chlorogenic acids and the expression level of a related gene or its expression product related to the change in body fat; An acquisition unit that acquires gene expression information including information on the expression level of the related gene or its expression product in association with identification information for identifying the subject; A prediction unit that predicts the change in body fat of the subject due to the intake of the chlorogenic acids from the gene expression information based on the correlation information A body fat change prediction system comprising the above.

18. A program for predicting the change in body fat of a subject due to the intake of chlorogenic acids, which causes a computer to A step of obtaining gene expression information including information on the expression level of at least one related gene or its expression product related to the change in body fat due to the intake of the chlorogenic acids, in association with subject information including identification information for identifying the subject; A step of predicting the change in body fat due to the intake of the chlorogenic acids in the subject from the gene expression information, based on correlation information indicating the correlation between the change in body fat in the plurality of sample subjects and the expression level of the related gene or its expression product related to the change in body fat; A program for causing the above to be executed.

Images