Method for detecting atopic dermatitis

Abstract

To provide a marker for detecting atopic dermatitis, and a method for detecting atopic dermatitis using the marker.SOLUTION: A method for detecting atopic dermatitis in a subject includes the step of measuring the expression level of at least one gene selected from seventeen gene groups consisting of MECR, RASA4CP, ARRDC4, EIF1AD, FDFT1, ZNF706, TEX2, TMPRSS11E, RPS6KB2, CTBP1, ZNF335, DGKA, PPP1R9B, SPDYE7P, DNASE1L1, GNB2 and CSNK1G2 or an expression product thereof in a biological sample taken from the subject.SELECTED DRAWING: None

Description

[Technical Field]

[0001] The present invention relates to a method for detecting atopic dermatitis using an atopic dermatitis marker. [Background technology]

[0002] Atopic dermatitis (hereinafter also referred to as "AD") occurs mainly in people with a predisposition to atopy. Atopic dermatitis is a skin disease characterized by eczema. The typical symptoms of atopic dermatitis are bilateral and contralateral. It can cause chronic and recurring itching, rash, erythema, etc., as well as keratinization failure, impaired barrier function, dry skin, etc. Atopic dermatitis occurs mostly in infants and tends to improve as the child grows. In recent years, adult-type and intractable atopic dermatitis has also been increasing.

[0003] Currently, the severity of atopic dermatitis is judged based on visual findings. Findings include dryness, erythema, scales, papules, excoriation, edema, crusting, and urination. There are various types, such as blisters, erosions, and pruritic nodules, but since each of these involves subjective factors, doctors There is a possibility that the judgment may differ depending on the skill of the doctor. The problem is that the judgment lacks objectivity.

[0004] Recently, peripheral blood tests have been used as a method for detecting atopic dermatitis using biomarkers. Eosinophil count, serum total IgE level, LDH (lactate dehydrogenase) level, serum thymus and Activation-Regulated Chemokine(TARC ) and Squamous cell carcinoma antigen 2 (SCCA2) (Non-patent documents 1, 2, 3) and the aggregative activity of Staphylococcus aureus in the skin flora. It has been proposed to detect the expression level of RNAIII gene dependent on rC mutation (Patent Document 1). However, it cannot be said that diagnosis is always possible with sufficient accuracy.

[0005] On the other hand, analysis of nucleic acids such as DNA and RNA in biological samples has revealed the current state of the human body and even Technologies for investigating future physiological states are being developed. Analysis using nucleic acids is a comprehensive analysis method. It has been established that a wealth of information can be obtained with a single analysis, and it is possible to obtain information on single nucleotide polymorphisms, RNA functions, etc. Based on many research reports on the subject, it has the advantage of being easy to link the analysis results functionally. Nucleic acids derived from living organisms can be extracted from body fluids such as blood, secretions, tissues, etc. Recently, it has been reported that α-glucan is contained in skin surface lipids (SSL). Use of RNA as a sample for biological analysis, and analysis of the epidermis, sweat glands, hair follicles, and sebaceous glands from SSL. It has been reported that the above marker genes can be detected (Patent Document 2). [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Japanese Patent Application Publication No. 2019-30272 [Patent Document 2] International Publication No. 2018 / 008319 [Non-patent literature]

[0007] [Non-Patent Document 1] Sugawara et al., Allergy (2002) 57:180-181 [Non-patent document 2] Ohta et al., Ann Clin Biochem.(2012) 49:277-84 [Non-patent document 3] Kato et al., Journal of the Japanese Dermatological Association (2018) 128: 2431-2502 Summary of the Invention [Problem to be solved by the invention]

[0008] The present invention relates to a marker for detecting atopic dermatitis and a method for detecting atopic dermatitis using the marker. The present invention relates to a method for detecting eczema. [Means for solving the problem]

[0009] The present inventors collected SSL from adult patients with atopic dermatitis and healthy individuals, and The expression state of the RNA contained in the gene was comprehensively analyzed as sequence information, and specific genes were identified. The expression level of α-glucan is significantly different between the two, and this can be used as an indicator to detect atopic dermatitis. We found that this is the case.

[0010] That is, the present invention relates to the following 1) to 3). 1) Regarding biological samples collected from subjects, MECR, RASA4CP, ARRDC 4, EIF1AD, FDFT1, ZNF706, TEX2, TMPRSS11E, RPS 6KB2, CTBP1, ZNF335, DGKA, PPP1R9B, SPDYE7P, D Selected from a group of 17 genes consisting of NASE1L1, GNB2, and CSNK1G2 The test method includes measuring the expression level of at least one gene or its expression product. Method for detecting atopic dermatitis in individuals. 2) an oligonucleotide that specifically hybridizes with the gene, or A method for detecting atopic dermatitis used in the method 1) containing an antibody that recognizes the expression product. A test kit for this purpose. 3) At least one gene selected from the 210 gene groups shown in Table B below is a marker for detecting atopic dermatitis, consisting of its expression product. [Effects of the Invention]

[0011] According to the present invention, atopic dermatitis can be diagnosed early and with high accuracy, sensitivity, and convenience in a simple and non-invasive manner. It is possible to detect with high specificity. DETAILED DESCRIPTION OF THE INVENTION

[0012] All patents, non-patent documents, and other publications cited herein are hereby incorporated by reference in their entirety. The bodies are incorporated herein by reference.

[0013] In the present invention, the term "nucleic acid" or "polynucleotide" refers to DNA or RNA. DNA includes cDNA, genomic DNA, and synthetic DNA. , "RNA" includes total RNA, mRNA, rRNA, tRNA, non-co This includes both ding RNA and synthetic RNA.

[0014] In the present invention, the term "gene" refers to double-stranded DNA including human genomic DNA, as well as cDNA. Single-stranded DNA (positive strand) containing the above, and single-stranded DNA (complementary strand) having a sequence complementary to the positive strand. and fragments thereof, and the sequence information of the bases constituting the DNA includes It means something that contains some biological information. In addition, the "gene" in question is not just a "gene" represented by a specific base sequence, but also Homologues (i.e., homologs or orthologs), variants such as genetic polymorphisms, and derivatives Nucleic acids encoding the The names of the genes disclosed herein are listed in NCBI ([www.ncbi.nlm.nih.gov / ]). On the other hand, the official symbols listed in the Gene Ontology (GO) For details, please refer to the Pathway ID listed in String ([string-db.org / ]). cormorant.

[0015] In the present invention, the "expression product" of a gene includes a transcription product and a translation product of the gene. A "transcript" is an RNA that is produced by transcription from a gene (DNA). "Translation product" refers to a protein encoded by a gene that is translated and synthesized based on RNA. means.

[0016] In the present invention, "atopic dermatitis" refers to itchy, moist skin that repeatedly worsens and improves. It refers to a disease whose main pathogenic factor is a rash, and many of its patients are said to have a predisposition to atopy. The predisposing factors for asthma are: i) family history and medical history (bronchial asthma, allergic rhinitis, conjunctivitis, allergic ii) one or more of the following diseases: atopic dermatitis; or ii) IgE antibody production Factors that make it easier to do so include:

[0017] In the present invention, "detection" of atopic dermatitis means detecting the presence or absence of atopic dermatitis. This means clarifying the existence of a problem, and can be expressed in terms of inspection, measurement, judgment, evaluation, or evaluation support. In this specification, the terms "assessment" and "evaluation" are used to refer to It does not include judgments or evaluations based on the above.

[0018] In this specification, "feature" is synonymous with "explanatory variable" in machine learning, and A gene or its expression product to be used for machine learning selected from the markers for detecting eczema Together they are called "feature genes."

[0019] As shown in the Examples below, 14 healthy adults and 29 patients with atopic dermatitis The expression level data (read count value) of RNA extracted from SSL was analyzed using DESe Count values ​​(Norma) were corrected using q2 (Love MI et al. Genome Biol. 2014). Using the quantified count value, the difference in the number of patients with atopic dermatitis compared with healthy controls was RNA with a p-value corrected by likelihood ratio test (FDR) of less than 0.05 was extracted. As a result, 48 genes with increased expression and 75 genes with decreased expression (total of 123 genes (Table 1-1 to 1- 3) were identified. In the table, genes indicated with "UP" are those whose expression levels are higher in patients with atopic dermatitis. The genes indicated by "DOWN" are those with atopic dermatitis. These are genes whose expression levels decrease. Therefore, the genes or their expression products selected from the group of 123 genes are These genes can be used as markers for detecting atopic dermatitis. Of these, 107 genes (marked with * in bold in Tables 1-1 to 1-3) have been This is a gene that has not been reported to be associated with dermatitis eczema.

[0020] In addition, the expression data of all SSL-derived RNA from the subjects (Lo g2(RPM+1) value) was used as an explanatory variable, and healthy subjects and patients with atopic dermatitis were used as the objective variables. Random forests (Breiman L. Machine Learning (200 1) 45;5-32) was used to extract feature genes and construct a prediction model. As shown in the example, the top 150 genes in terms of variable importance based on the Gini coefficient (Tables 3-1 to 3-4) We selected these as feature genes and used them to build a prediction model, which revealed that atopic dermatitis It was shown that it is possible to predict dermatitis. Therefore, the genes or their expression products selected from the group of 150 genes are It can be a suitable marker for detecting atopic dermatitis. Of these, 127 genes (shown in bold with an * in Tables 3-1 to 3-4) have been This is a novel atopic dermatitis marker that has not been reported to be associated with atopic dermatitis. As shown in the examples below, a prediction model using these new atopic dermatitis markers was developed. It is possible to predict atopic dermatitis even with this method.

[0021] In addition, the above 123 genes or their equivalents whose expression levels were altered between healthy individuals and patients with atopic dermatitis were Similarly, using the expression data (Log2(RPM+1) value) of 107 genes, When we attempted to build a prediction model using random forest, we found that atopic dermatitis was not associated with any of the It has been shown that it is possible to predict allergic dermatitis.

[0022] In addition, the Boruta method (Kursa et al. Fundamental Information maticae (2010) 101;271-286) to extract feature genes (maximum number of trials: 1000, p-value less than 0.01), 45 genes (Table 4) were extracted as feature genes. As will be shown in the examples below, a random forest prediction model using this method can be used to It has been shown that it is possible to predict dermatitis. Therefore, the genes or their expression products selected from the group of 45 genes are This can be a suitable marker for detecting atopic dermatitis. Of these, 39 genes (marked with an * in bold in Table 4) have been reported to date to be associated with atopic dermatitis. It is a novel atopic dermatitis marker that has not been reported to be related to As shown, the prediction model using these new atopic dermatitis markers also showed atopic dermatitis Dermatitis can be predicted.

[0023] The 123 gene groups (shown in Tables 1-1 to 1-3) extracted by the above-mentioned expression analysis A) and the genes selected as feature genes by random forest are shown in Tables 3-1 to 3-4. A group of 150 genes (B) selected as feature genes by the Boruta method. The sum (A∪B∪C) of the 45 gene groups (C) shown in Table 4 was 245 genes (Table A ) are atopic dermatitis markers, of which 210 genes (Table B) are novel atopic dermatitis markers. It is a marker for skin dermatitis.

[0024] [Table A]

[0025] [Table B]

[0026] MECR, RASA4CP, ARRDC4, EIF1AD, FDFT1, ZN of the present invention F706, TEX2, TMPRSS11E, RPS6KB2, CTBP1, ZNF335 , DGKA, PPP1R9B, SPDYE7P, DNASE1L1, GNB2 and CSN The 17 genes comprising K1G2 are listed in Tables 1-1 to 1- 3, the 123 gene group (A) shown in Fig. 3 and the random forest were used to select the feature genes. The 150 gene groups (B) shown in Tables 3-1 to 3-4 selected by the Boruta method were The 45 genes in Table 4 (C) were selected as feature genes by Genes (A∩B∩C) that have not previously been associated with atopic dermatitis (In each table, the genes are marked with an * and written in bold.) At least one gene or its expression product is a novel method for detecting atopic dermatitis. It is particularly useful as a novel marker for atopic dermatitis. Each of these 17 genes can be a marker for atopic dermatitis. It is preferable to use a combination of at least 5 kinds, more preferably at least 10 kinds. It is even more preferable to use all 17 species.

[0027] The above genes that can be markers for atopic dermatitis (hereinafter referred to as "target genes") The gene is included in the list of biomarkers for detecting atopic dermatitis, as long as it can be used as a biomarker for detecting atopic dermatitis. It also includes genes that have a base sequence that is substantially identical to the base sequence of the DNA that constitutes the gene. Here, the term "substantially identical base sequences" refers to sequences that are substantially identical to each other, for example, sequences that are identical to each other using the homology calculation algorithm NCBI BLAST. Use, expectation=10; gaps allowed; filtering=ON; match score=1; misses When searching with a match score of -3, the base sequence of the DNA that constitutes the gene and the sequence has at least 90%, preferably at least 95%, and even more preferably at least 98% identity with the sequence. This means:

[0028] The method for detecting atopic dermatitis of the present invention comprises: In one embodiment, the target gene is MECR, RASA4CP, ARRDC4, EIF1AD, F DFT1, ZNF706, TEX2, TMPRSS11E, RPS6KB2, CTBP1 , ZNF335, DGKA, PPP1R9B, SPDYE7P, DNASE1L1, GN At least one gene selected from a group of 17 genes consisting of B2 and CSNK1G2 The method includes measuring the expression level of the gene or its expression product.

[0029] The biological sample used in the present invention is a sample containing the above-mentioned components that are involved in the onset and progression of atopic dermatitis. Any tissue or biomaterial in which the gene of the invention changes its expression may be used. Specifically, organs, skin, blood prepared from body fluids such as urine, saliva, sweat, skin surface lipids (SSL), tissue exudates, and blood Examples include serum, plasma, feces, hair, etc., and preferably skin or skin surface lipids (SS L), and more preferably skin surface lipids (SSL). In addition, the subjects from whom biological samples are collected are those who need to be detected for atopic dermatitis or those who have atopic dermatitis. The target population is preferably people suspected of developing eczema, and there are no limitations on gender or age, but the target population is preferably people suspected of developing eczema. is an adult, specifically a person aged 16 or over, preferably 20 or over.

[0030] Here, "skin surface lipids (SSL)" refers to the fat-soluble fraction present on the surface of the skin, It is sometimes called sebum. Generally, SSL is secreted from exocrine glands such as sebaceous glands in the skin. It is mainly composed of secretions released from the skin and exists on the surface of the skin as a thin layer covering the surface of the skin. contains RNA expressed in skin cells (see Patent Document 3). Unless otherwise specified, "skin" refers to the epidermis, dermis, hair follicles, sweat glands, and sebaceous glands on the surface of the body. It is a general term for the area that includes tissues such as the lining of the stomach and other glands.

[0031] The SSLs were collected from the subject's skin using a device used to retrieve or remove the SSLs from the skin. Any means that can be used for the absorption of the material may be employed. Preferably, the material is an SSL absorbent material, S SL adhesive material or a tool to scrape SSL from the skin can be used. As for absorbent materials or SSL adhesive materials, materials that have affinity for SSL are particularly suitable. Examples of the material include, but are not limited to, polypropylene and pulp. A more detailed example of the order is the SS on sheet-like materials such as oil blotting paper and oil blotting film. How to absorb L, how to attach SSL to glass plates, tape, etc., how to use a spatula, a scraper, etc. SSL can be scraped off and collected using a rapper or similar method. To improve the properties of the material, SSL absorbents pre-soaked with a highly lipid-soluble solvent may be used. On the other hand, SSL absorbent materials are hindered from adsorption when they contain highly water-soluble solvents or moisture. Therefore, it is preferable to use solvents with high water solubility and low water content. It is preferable to use the material in a dry state. The area of ​​the skin from which SSL is collected is The skin may be any part of the body, such as the head, face, neck, trunk, limbs, etc., but is not particularly limited thereto. Areas where oil is secreted in large amounts, such as facial skin, are preferred.

[0032] The RNA-containing SSL collected from the subject may be stored for a period of time. L should be stored under low temperature conditions as soon as possible after collection to minimize degradation of the RNA contained in it. In the present invention, the RNA-containing SSL is preferably stored at a temperature of 0°C. The temperature is preferably -20±20°C to -80±20°C, more preferably -20 ±10°C to -80±10°C, more preferably -20±20°C to -40±20°C, Preferably, the temperature is -20±10°C to -40±10°C, more preferably -20±10°C, and even more preferably The temperature is preferably −20±5° C. The period of storage of the RNA-containing SSL under the low temperature condition is particularly Although not limited to, it is preferably 12 months or less, for example, 6 hours or more and 12 months or less, more preferably or 6 months or less, for example, 1 day or more and 6 months or less, more preferably 3 months or less, for example More than 3 days and less than 3 months.

[0033] In the present invention, the expression level of a target gene or its expression product is measured using RNase I. cDNA artificially synthesized from A, DNA encoding that RNA, and that RNA The encoded protein, the molecules that interact with the protein, and the molecules that interact with the RNA Here, RNA, DNA or molecules that interact with the DNA are included. Molecules that interact with proteins include DNA, RNA, proteins, polysaccharides, and oligosaccharides. Sugars, monosaccharides, lipids, fatty acids, and their phosphorylation products, alkylation products, sugar adducts, etc., and the above The expression level refers to the expression level of the gene or expression product. It comprehensively refers to abundance and activity.

[0034] In a preferred embodiment of the method of the present invention, SSL is used as the biological sample. In this case, the expression level of the RNA contained in the SSL is analyzed, specifically, the RNA is reverse-translated. After conversion to cDNA by transcription, the cDNA or its amplification product is measured. RNA extraction from SSLs involves the use of methods commonly used for the extraction or purification of RNA from biological samples. methods such as the phenol / chloroform method, AGPC (acid guanidin ium thiocyanate-phenol-chloroform extrac tion) method, or TRIzol®, RNeasy®, QIAzo Column-based methods such as 1 (registered trademark) and special magnetic particles coated with silica The method used is Solid Phase Reversible Immobilization Methods using ion magnetic particles, extraction using commercially available RNA extraction reagents such as ISOGEN, etc. It can be used.

[0035] For the reverse transcription, a primer targeting a specific RNA to be analyzed may be used. For more comprehensive nucleic acid storage and analysis, it is preferable to use random primers. For the reverse transcription, a general reverse transcriptase or a reverse transcription reagent kit can be used. For this purpose, highly accurate and efficient reverse transcriptase or reverse transcription reagent kits are used. M-MLV reverse transcriptase and its variants, Alternatively, commercially available reverse transcriptase or reverse transcription reagent kits, such as PrimeScript (registered trademark), may be used. ) Reverse Transcriptase series (Takara Bio Inc.), Super rScript® Reverse Transcriptase Series (T Hermo Scientific) and others. Trademark)III Reverse Transcriptase, SuperScript( VILO cDNA Synthesis kit (registered trademark) Scientific) is preferably used. The temperature for the extension reaction in the reverse transcription is preferably 42°C ± 1°C, more preferably 42°C ± 1°C. ±0.5°C, more preferably 42°C ±0.25°C, while the reaction time is preferably The time is preferably adjusted to 60 minutes or more, more preferably 80 to 120 minutes.

[0036] When RNA, cDNA, or DNA is used as the target, the expression level can be measured using these methods. PCR method using DNA that hybridizes to the primer, real-time RT-PCR method Nucleic acid amplification methods such as multiplex PCR, SmartAmp, and LAMP , a hybridization method using a nucleic acid that hybridizes to these as a probe ( DNA chip, DNA microarray, dot blot hybridization, slot blot hybridization, Northern blot hybridization, etc.), salt Select from methods for determining base sequences (sequencing), or a combination of these methods. can be done.

[0037] In PCR, a primer pair targeting a specific DNA to be analyzed is used to amplify that specific DNA. Although it is possible to amplify only the DNA of one species, multiple primer pairs can be used to amplify multiple specific DNA simultaneously. The DNA may be amplified. Preferably, the PCR is a multiplex PCR. Multiplex PCR is a method for generating multiple PCR products by using multiple primer pairs simultaneously in a PCR reaction system. Multiplex PCR is a method for amplifying multiple gene regions simultaneously. (e.g., Ion AmpliSeqTranscriptome Human Gen e Expression Kit; Life Technologies Japan, Inc.) This can be implemented. The annealing and extension reaction temperatures in the PCR depend on the primers used. Therefore, it is difficult to generalize, but when using the above multiplex PCR kit, it is preferable to is 62°C ± 1°C, more preferably 62°C ± 0.5°C, and even more preferably 62°C ± 0.25°C Therefore, in the PCR, the annealing and extension reactions are preferably carried out in one step. The time for the annealing and extension steps depends on the size of the DNA to be amplified. The time can be adjusted depending on the size, etc., but is preferably 14 to 18 minutes. The conditions for the denaturation reaction can be adjusted depending on the DNA to be amplified, but are preferably 95 to 99 ℃ for 10 to 60 seconds. Reverse transcription and PCR at the above temperatures and times are generally This can be performed using a thermal cycler commonly used for PCR.

[0038] The PCR product is purified by size separation. It is preferable to separate the target PCR reaction product into the target PCR reaction product by size separation. DNA can be separated from primers and other impurities by size separation, e.g. By using size separation columns, size separation chips, magnetic beads that can be used for size separation, etc. A preferred example of magnetic beads that can be used for size separation is Ampu re XP etc. Solid Phase Reversible Immobiliza Examples include SPRI magnetic beads.

[0039] Further processing of purified PCR products required for subsequent quantitative analysis For example, purified PCR reaction products may be subjected to Prepare the PCR product in an appropriate buffer solution or use it as a PCR primer in PCR-amplified DNA. Even if the amplicon region is cut or an adapter sequence is added to the amplified DNA, For example, purified PCR reaction products can be prepared in a buffer solution and then used to amplify the amplified DNA. The PCR primer sequence was removed and adapter ligation was performed. The resulting DNA can be amplified as needed to prepare a library for quantitative analysis. These procedures can be carried out, for example, using the SuperScript® VILO cDNA System. 5x IVF supplied with the IVF Enthesis Kit (Life Technologies Japan, Inc.) VILO RT Reaction Mix and Ion AmpliSeq Tran Scriptome Human Gene Expression Kit (Lifetech) 5x Ion AmpliSeq HiF kit included with the Ion Technologies Japan Co., Ltd. i Mix and Ion AmpliSeq Transcriptome Human Using the Gene Expression Core Panel, the primers included with each kit were This can be done according to a protocol.

[0040] Northern blot hybridization was used to identify target genes or their derivatives. When measuring the expression level of a nucleic acid, for example, first, probe DNA is fused to a radioisotope or a fluorescent substance. The labeled DNA is then transferred to a nylon membrane or the like in a conventional manner. The resulting target is then hybridized with RNA from a transfected biological sample. The double strand of DNA and RNA is measured by detecting the signal derived from the label. One method is to

[0041] When measuring the expression level of a target gene or a nucleic acid derived therefrom using RT-PCR, For example, first, cDNA is prepared from RNA derived from a biological sample according to a conventional method, and then the cDNA is used as a template. A pair of primers (the above-mentioned cDNA ( The positive strand that binds to the negative strand and the reverse strand that binds to the positive strand are then hybridized with this. PCR is performed according to standard methods, and the resulting amplified double-stranded DNA is detected. For DNA detection, primers labeled with RI, fluorescent substances, etc. are used to detect the above PC A method for detecting labeled double-stranded DNA produced by performing R can be used. do.

[0042] Measuring the expression level of a target gene or a nucleic acid derived therefrom using a DNA microarray In this case, for example, a small amount of nucleic acid (cDNA or DNA) derived from the target gene of the present invention is attached to a support. Using an array immobilizing at least one species, labeled cDNA or cRNA prepared from mRNA By binding A to the microarray and detecting the label on the microarray, m The amount of RNA expression can be measured. The nucleic acids immobilized on the array are specific (i.e., That is, any nucleic acid that hybridizes to the target nucleic acid (substantially only to the target nucleic acid) is acceptable. The nucleic acid may be a nucleic acid having the entire sequence of a target gene, or a nucleic acid consisting of a partial sequence. Here, the "partial sequence" includes a nucleic acid consisting of at least 15 to 25 bases. Here, stringent conditions are usually around "1x SSC, 0.1% SDS, 37℃". The washing conditions are as follows: 0.5x SS, and the more stringent hybridization conditions are as follows: For hybridization conditions, 0.1% SDS, 42°C is recommended. Hybridization conditions include approximately 100x SSC, 0.1% SDS, and 65°C. Conditions are as per J. Sambrook et al., Molecular Cloning: A Laboratory Manual, Thrd Edition. , Cold Spring Harbor Laboratory Press (2001), etc.

[0043] When measuring the expression level of a target gene or a nucleic acid derived therefrom by sequencing, For example, next-generation sequencers (e.g., Ion S5 / XL System, Life Technologies) The analysis can be performed using a system developed by the Japanese Ministry of Health, Labour and Welfare (MHLW). Based on the number of reads obtained (read count), RNA expression can be quantified.

[0044] The probes or primers used in the above measurement, i.e., the target gene or A primer for specifically recognizing and amplifying a nucleic acid derived therefrom, or the RNA or This includes probes for specifically detecting nucleic acids derived from the It can be designed based on the base sequence that constitutes the target gene. The term "substantially detects the target gene of the present invention or a gene thereof in, for example, Northern blotting" means that the target gene of the present invention or a gene thereof is substantially detected. and in the RT-PCR method, for example, the nucleic acid derived from the The detection or product is the gene or nucleic acid derived therefrom so that only the nucleic acid is amplified. This means that it can be determined to be an acid. Specifically, the target gene of the present invention is a DNA having a base sequence constituting the target gene or a complementary strand thereof. Oligonucleotides containing a certain number of complementary nucleotides can be used. The "complementary strand" is a double-stranded DNA consisting of A:T (U in the case of RNA) and G:C base pairs. The term "complementary" refers to a certain number of consecutive nucleotides in a sequence. It is not limited to a completely complementary sequence in the nucleic acid region, but is preferably 80% or more, more preferably It is sufficient if the base sequence has at least 90% identity, and more preferably at least 95% identity. The identity of the nucleotide sequences can be determined using an algorithm such as the BLAST algorithm. Such oligonucleotides, when used as primers, exhibit specific annealing properties. It is sufficient if the fragment can be fused and elongated, and usually, for example, 10 bases or more, preferably 15 bases or more, More preferably, it is 20 bases or more, and for example, 100 bases or less, preferably 50 bases or less, Preferably, the chain length is 35 bases or less. In this case, specific hybridization is sufficient, and the target gene of the present invention is constituted. At least a part or all of the sequence of DNA (or its complementary strand) consisting of the base sequence For example, it has 10 bases or more, preferably 15 bases or more, and for example, 100 bases or less, preferably Preferably, the chain length is 50 bases or less, more preferably 25 bases or less. Here, the "oligonucleotide" can be DNA or RNA. The probes used for hybridization may be synthetic or natural. , usually labeled ones are used.

[0045] In addition, the translation product (protein) of the target gene of the present invention, and a gene that interacts with the protein If you are measuring molecules, molecules that interact with RNA, or molecules that interact with DNA, Protein chip analysis, immunoassays (e.g., ELISA), mass spectrometry (e.g., LC- MS / MS, MALDI-TOF / MS), one-hybrid method (PNAS 100, 12271-1227 6 (2003)) and the two-hybrid method (Biol. Reprod. 58, 302-311 (1998)). It can be used and can be selected appropriately depending on the subject. For example, when a protein is used as the measurement target, an antibody against the expression product of the present invention is used. The body is contacted with a biological sample, and proteins in the sample that bind to the antibody are detected and their levels are determined. For example, Western blotting is performed by measuring the primary antibody After using the above antibody as a primary antibody, a secondary antibody labeled with a radioisotope, fluorescent substance, enzyme, etc. The primary antibody is labeled with an antibody that binds to the primary antibody, and the resulting labeled substance is used to The signal is measured using a radiation detector, a fluorescence detector, or the like. The antibody against the translation product may be a polyclonal antibody or a monoclonal antibody. These antibodies can be produced according to known methods. Polyclonal antibodies are produced using proteins that have been expressed in E. coli or other bacteria and purified according to standard methods. Alternatively, a partial polypeptide of the protein may be synthesized by a conventional method and then cultured in a non-human animal such as a rabbit. The antibody can be obtained by immunizing a human animal and obtaining the antibody from the serum of the immunized animal in a conventional manner. On the other hand, monoclonal antibodies are proteins that are expressed in E. coli or the like and purified according to standard methods. A partial polypeptide of the protein is immunized into a non-human animal such as a mouse, and the resulting spleen cells are then used to It can be obtained from hybridoma cells prepared by cell fusion with myeloma cells. Alternatively, monoclonal antibodies may be generated using phage display (Griffiths, A .D.; Duncan, AR, Current Opinion in Biotechnology, Volume 9, Number 1, February y 1998 , pp. 102-108(7)).

[0046] Thus, the target gene of the present invention or its expression product in a biological sample collected from a subject can be detected. The expression level is measured, and atopic dermatitis is detected based on the expression level. Specifically, the measured expression level of the target gene of the present invention or its expression product is compared with a control level. This is done by comparing it to a bell. When analyzing the expression levels of multiple target genes by sequencing, As shown in the figure, the read count value, which is the data on the expression level, and the read count value are used to calculate the total read count between samples. RPM value corrected for differences in the number of columns, and the value converted to a logarithmic value with base 2 (Log2 RPM value) or the logarithm value to the base 2 with an integer 1 added (Log2(RPM+1) value), or Normalized count values ​​using DESeq2 or uses the logarithm value of base 2 (Log2(count+1) value) added with integer 1 as an index. It is also preferable to use fragments p as a quantitative value for RNA-seq. er kilobase of exon per million reads ma pped (FPKM), reads per kilobase of exon p er million reads mapped (RPKM), transcript It may also be a value calculated by ts per million (TPM) or the like. Alternatively, the signal value obtained by the microarray method and its corrected value may be used. In addition, when analyzing the expression level of only a specific target gene using RT-PCR, etc., The expression level of the target gene is converted to a relative expression level based on the expression level of the housekeeping gene. Alternatively, a plasmid containing the target gene region can be used to analyze the target gene. The preferred method is to quantify (absolutely quantify) the number of copies of a gene by digital PCR. The number of copies obtained by the method may also be the number of copies obtained by the method. Here, the "control level" refers to, for example, the level of the target gene or its expression product in a healthy individual. The expression level of a gene in a healthy individual is measured from a population of healthy individuals. It may also be a statistical value (e.g., average value) of the expression level of the target gene or its expression product. If there are multiple genes, determine the reference expression level for each gene or its expression product. is preferred.

[0047] Furthermore, the detection of atopic dermatitis in the present invention can be carried out by detecting the target gene of the present invention or its expression product. This can also be achieved by increasing / decreasing the expression level of a gene. The expression level of the target gene or its expression product in the sample is The cut-off value is determined in advance by comparing the target gene expression level in healthy individuals with the reference value. The expression level of the gene or its expression product is obtained as reference data, and the expression level based on this is calculated. It may be determined appropriately based on statistical values ​​such as the average value and standard deviation of the values.

[0048] Furthermore, the expression levels of target genes or their expression products derived from atopic dermatitis patients and healthy individuals were compared. Using the measured expression levels of target genes or their expression products derived from normal subjects, We constructed a discriminant (prediction model) to distinguish between dermatitis patients and healthy individuals, and used this discriminant to It is possible to detect atopic dermatitis. Expression level of a gene or its expression product, and expression of a target gene or its expression product derived from a healthy individual Using the measured values ​​of the level as a training sample, a discriminant ( A predictive model is constructed and atopic dermatitis patients and healthy individuals are distinguished based on the discriminant. The cutoff value (reference value) is calculated. In creating the discriminant equation, principal component analysis (PCA) ) to perform dimensionality reduction, and the principal components can be used as explanatory variables. The level of the target gene or its expression product is then measured from the biological sample collected from the subject. Measure the same way, substitute the measured value into the discriminant, and refer to the result obtained from the discriminant. By comparing the values, the presence or absence of atopic dermatitis in the subject can be evaluated. Cut.

[0049] The variables used to construct the discriminant equation are explanatory variables and target variables. The expression level of a target gene or its expression product selected by the following method can be used. The target variable is, for example, whether the sample is from a healthy individual or a patient with atopic dermatitis. , can be used.

[0050] The selection of features is based on statistically significant differences between the two groups to be discriminated, e.g., expression levels of 2 Using the expression levels of genes that vary significantly between groups (differentially expressed genes) or their expression products In addition, it is possible to use known algorithms such as those used in machine learning to extract features. For example, the following gene expression level can be used. Using the expression levels of genes or their expression products with high variable importance in random forests Or, we extract feature genes using the "Boruta" package in R language, and Expression levels can be used.

[0051] As the algorithm for constructing the discriminant, publicly available algorithms such as those used for machine learning are used. Examples of machine learning algorithms include Random Forest. Random forest, linear kernel support vector machine (SVM) linear), rbf kernel support vector machine (SVM rbf) Neural net, Generalized Linear Model d linear model), regularized linear discriminant analysis (Regularized li near discriminant analysis), regularized logistic regression (Regularized logistic regression) The validation data is input into the constructed prediction model to calculate a predicted value, and the predicted value is compared with the actual measurement value. The model that best fits the model, e.g., the model with the highest accuracy, is selected as the optimal model. It can be selected as a prediction model. In addition, the detection rate (Recalculated) can be calculated from the predicted value and the actual measured value. l), precision, and their harmonic mean F-measure, and The model with the highest value can be selected as the optimal predictive model.

[0052] When using the random forest algorithm to build the discriminant, the predictive model As an index of accuracy, the OOB error rate is used. (Breiman L. Machine Learning (2001) 45;5-32).

[0053] In random forests, a method called bootstrap is used to extract the data from all samples. A sample of about two-thirds of the total number of samples is randomly extracted from the tree, allowing overlaps, and a decision tree is created. The samples that are not extracted at this time are considered to be out of bug (OO) samples. B), which uses a single decision tree to predict the OOB target variable and compare it with the correct label. By comparing the OOB error rate in the decision tree, the error rate can be calculated. The same process was repeated 500 times to obtain the OOB er The average error rate is used as the OOB error rate for the random forest model. or rate.

[0054] The number of decision trees (ntree value) used to build the random forest model is set to the default. The default is 500, but this can be changed to any number as needed. The number of variables (metric values) used to create the discriminant equation for a sample in one decision tree is By default, this is the square root of the number of explanatory variables, but you can change it to one or all explanatory variables as needed. The value can be changed to any of the following values:

[0055] The mtry value can be determined using the R package “caret.” We specified the random forest method in the “aret” package and used eight different mtry values. Try it out and select the mtry value that maximizes accuracy as the optimal mtry value. The number of attempts for the mtry value can be changed to any number of attempts as needed. It is possible.

[0056] When using the random forest algorithm to build the discriminant, the model structure The importance of the explanatory variables used in the construction can be expressed as a numerical value (variable importance). For example, the decrease in the Gini coefficient (Mean Decrease Gini) can be used. can be done.

[0057] The method for determining the cutoff value (reference value) is not particularly limited, and can be determined according to a known method. For example, the ROC (Receiver Operator Criteria) created using the discriminant It can be obtained from the (Lattice Characteristic Curve) In an ROC curve, the vertical axis represents the probability of a positive result in a positive patient (sensitivity) and the horizontal axis represents the probability of a negative result. The false positive rate is calculated by subtracting the probability of a negative result (specificity) from 1 in a patient. The ROC curve shows "true positive (sensitivity)" and "false positive (1-specificity)". Regarding the sensitivity, the value at which "true positive (sensitivity)" - "false positive (1 - specificity)" is maximized (Youden The index) can be used as a cutoff value (reference value).

[0058] As already mentioned, MECR, RASA4CP, ARRDC4, EIF1AD, and FDF T1, ZNF706, TEX2, TMPRSS11E, RPS6KB2, CTBP1, Z NF335, DGKA, PPP1R9B, SPDYE7P, DNASE1L1, GNB2 and CSNK1G2, and Table 1-1 123 genes shown in Tables 1-3, 150 genes shown in Tables 3-1 to 3-4, or Table 4 A prediction model that can predict atopic dermatitis using the 45 genes shown as feature genes can be constructed.

[0059] Therefore, when creating a discriminant to separate the atopic dermatitis patient group from the healthy subject group, The feature genes were MECR, RASA4CP, ARRDC4, EIF1AD, and FD. FT1, ZNF706, TEX2, TMPRSS11E, RPS6KB2, CTBP1, ZNF335, DGKA, PPP1R9B, SPDYE7P, DNASE1L1, GNB One or more genes selected from 17 genes consisting of CSNK1G2 and CSNK1G2 Preferably, 5 or more types, more preferably 10 or more types, and even more preferably all 17 types are selected. Select a gene and use the expression data of that gene or its expression product. In this case, the genes with higher variable importance in Tables 3-1 to 3-4 are selected. It is preferable to select these as feature genes in order and create a discriminant. The genes include 245 genes shown in Table A, 123 genes shown in Tables 1-1 to 1-3, Among the 150 genes shown in Tables 3-1 to 3-4 or the 45 genes shown in Table 4, At least one, five or more, ten or more, or twenty or more genes selected from genes other than the 17 genes Alternatively, by appropriately adding expression data of 50 or more genes or their expression products, the discriminant formula can be It is also possible to create a gene encoding a gene other than the 17 genes and detect atopic dermatitis. When selecting genes from the 150 genes shown in Tables 3-1 to 3-4, the importance of variables is In order from the top genes, or within the top 50, preferably 30 or more, in terms of variable importance. Feature genes may be selected from the genes in the 17 types of genes. When selecting the gene as a feature gene, refer to Tables 1-1 to 1-3, Tables 3-1 to 3-4, and Table 4, the new atopic dermatitis markers are marked with * and shown in bold. It is preferable to select Preferably, the above 17 genes, the 123 genes or the 107 genes shown in Tables 1-1 to 1-3 Genes (indicated in bold with an * in Tables 1-1 to 1-3), 1 shown in Tables 3-1 to 3-4 50 genes or 127 genes (indicated in bold with an * in Tables 3-1 to 3-4), or Table The 45 genes or 39 genes shown in 4 (indicated in bold with an * in Table 4) were used as feature genes. A discriminant that uses genes is listed.

[0060] The test kit for detecting atopic dermatitis of the present invention is a test kit for detecting atopic dermatitis using a biological sample isolated from a patient. Test reagent for measuring the expression level of the target gene of the present invention or its expression product in a sample Specifically, the target gene of the present invention or a nucleic acid derived therefrom is specifically specifically binds (hybridizes) to the target oligonucleotide (e.g., a primer for PCR) A reagent for nucleic acid amplification or hybridization, or a target gene of the present invention, comprising Examples of such reagents include reagents for immunological assays containing antibodies that recognize the expression products (proteins) of The oligonucleotides, antibodies, etc. contained in the kit can be prepared by known methods as described above. can be obtained by In addition to the above antibodies and nucleic acids, the test kit also contains labeling reagents, buffer solutions, coloring substrates, and The secondary antibody, blocking agent, and the necessary equipment for the test, as well as positive and negative controls control reagents used as controls, and equipment for collecting biological samples (e.g., It may include items such as oil removal films for collecting SSL.

[0061] Aspects and preferred embodiments of the present invention are set out below. <1> Regarding biological samples collected from subjects, MECR, RASA4CP, and ARRD C4, EIF1AD, FDFT1, ZNF706, TEX2, TMPRSS11E, RP S6KB2, CTBP1, ZNF335, DGKA, PPP1R9B, SPDYE7P, Selected from a group of 17 genes consisting of DNASE1L1, GNB2, and CSNK1G2 measuring the expression level of at least one gene or its expression product in the subject A method for detecting atopic dermatitis in a subject. <2> The expression level of a gene or its expression product is measured by measuring the expression amount of mRNA. <1> How to do it. <3> The gene or its expression product is RNA contained in the lipids on the skin surface of the subject. <1> or <2> How to do it. <4> The measured expression levels are compared with reference values ​​for each of the genes or their expression products to identify atopic Evaluate the presence or absence of eczema, <1> ~ <3> Either way. <5> The expression level of the gene or its expression product derived from atopic skin patients and that of healthy individuals The measured expression level of the gene or its expression product was used as a teacher sample. A discriminant formula was created to distinguish between patients with allergic dermatitis and healthy individuals, and the results obtained from biological samples collected from the subjects were used. The measured values ​​of the expression levels of the gene or its expression products were substituted into the discriminant, and the results obtained were The results are compared with the reference values ​​to determine the presence or absence of atopic dermatitis in the subject. Evaluate, <1> ~ <3> Either way. <6> The algorithm for constructing the discriminant is random forest, linear kernel support. Support Vector Machine, Support Vector Machine with RBF kernel, Neural Network The method in <5> is a regularized logistic regression, general linear model, regularized linear discriminant analysis, or regularized logistic regression. <7> The expression levels of all genes in the 17 gene groups or their expression products are measured. Ru, <5> or <6> How to do it. <8> In addition to at least one gene selected from the seven gene groups, 1123 species shown in Tables 1 to 1-3, 150 species shown in Tables 3-1 to 3-4, and 150 species shown in Table 4. At least one gene selected from the group of 45 genes excluding the 17 genes The expression level of the gene or its expression product is measured. <5> ~ <7> Either way . <9> The 150 species shown in Tables 3-1 to 3-4 below were extracted using random forests. The selected feature genes are <8> How to do it. <10> The 45 types shown in Table 4 below are feature genes extracted using the Boruta method. That is, <8> How to do it. <11> An oligonucleotide that specifically hybridizes with the gene or a nucleic acid derived therefrom or an antibody that recognizes the expression product of the gene, <1> ~ <10> Noi A test kit for detecting atopic dermatitis to be used in any of the methods. <12> At least one gene selected from the group of 210 genes shown in B or a marker for detecting atopic dermatitis, comprising the gene encoding the gene or its expression product. <13> MECR, RASA4CP, ARRDC4, EIF1AD, FDFT1, ZN F706, TEX2, TMPRSS11E, RPS6KB2, CTBP1, ZNF335 , DGKA, PPP1R9B, SPDYE7P, DNASE1L1, GNB2 and CSN At least one gene selected from the group of 17 genes consisting of K1G2 or its expression It is a product, <12> A marker for detecting atopic dermatitis. [Example]

[0062] The present invention will be described in more detail below with reference to examples, but the present invention is not limited to these examples. It's not that.

[0063] Example 1: Atopic dermatitis-related differentially expressed genes in RNA extracted from SSL Detection of 1) SSL collection Fourteen healthy adult subjects (HL) (25-57 years old, male) and fourteen adult subjects with atopic skin The subjects were 29 men with atopic dermatitis (AD) (aged 23-56 years). A dermatologist has determined that the rash is at least on the face and is mild or moderate in severity. Each subject was diagnosed with atopic dermatitis. Remove sebum from the skin using an oil absorbing film (5 x 8 cm, polypropylene, 3M). After collection, the oil blotting film was transferred to a vial and stored at -80°C until use in RNA extraction. and stored for about a month.

[0064] 2) RNA preparation and sequencing Cut the oil absorbing film from 1) above to an appropriate size and place it in QIAzol Lysis RNA was extracted using Qiagen Reagent according to the attached protocol. The extracted RNA was used for cDNA synthesis using SuperScript VILO. SIS kit (Life Technologies Japan) at 42°C for 90 minutes Reverse transcription was performed to synthesize cDNA. The primers for the reverse transcription reaction were the same as those included in the kit. The random primers used were from the cDNA obtained, and multiplex PCR was performed. A library containing DNA derived from the 20802 gene was prepared. PCR was performed using the Ion AmpliSeq Transcriptome Human Genomics ne Expression Kit (Life Technologies Japan, Inc.) Then, [99°C, 2 min → (99°C, 15 sec → 62°C, 16 min) × 20 cycles → 4°C, Ho The PCR products were purified using Ampure XP (Beckman, Co., Ltd.). After purification by the HPLC (Luther Co., Ltd.), buffer reconstitution, digestion of primer sequences, and adaptor Ligation, purification, and amplification were performed to prepare a library. The data was loaded onto the Ion 540 Chip and then run on the Ion S5 / XL system (lifecycle Sequencing was performed using a PCR technology developed by Sigma-Aldrich Technologies Japan Co., Ltd.

[0065] 3) Data analysis i) Usage Data The data on the expression level of RNA derived from the subject measured in 2) above (read count value) The expression level data for all sample subjects was corrected using a method called DESeq2. Of these, 742 samples had non-missing expression data for over 90% of the subjects. Only 9 genes were used in the following analysis. The analysis was performed using a method called DESeq2. The normalized count value was used.

[0066] ii) RNA expression analysis The expression levels of SSL-derived RNA in healthy subjects and AD subjects measured in i above (Normalized Based on the count value, the adjusted p-value by likelihood ratio test was calculated for AD compared with healthy controls. We identified RNAs (differentially expressed genes) with a FDR of less than 0.05. 1 species of RNA are decreased (DOWN) in AD, 48 species of RNA are increased (UP) in AD ) (Tables 1-1 to 1-3).

[0067] [Table 1-1]

[0068] [Table 1-2]

[0069] [Table 1-3]

[0070] The 123 genes shown in Tables 1-1 to 1-3 were analyzed using the public database STRI. Gene Ontology (GO) enrichment analysis using NG for biologic As a result, we found that the expression of BP was decreased in AD patients. 27 BPs related to the gene cluster were obtained, including terms related to lipid metabolism and amino acid metabolism. In addition, four BPs associated with the gene clusters whose expression levels were elevated were obtained (Table 2). , and terms related to leukocyte activation (Table 2). Of the 123 genes shown in Tables 1-1 to 1-3, 107 genes (in each table, * is indicated) Regarding the substances (shown in bold), there have been reports suggesting a relationship with atopic dermatitis. Therefore, it was determined that this compound could be a novel marker for atopic dermatitis.

[0071] [Table 2]

[0072] Example 2: Construction of a discriminant model using genes with high variable importance in random forests 1) Usage Data In the same manner as in Example 1, data on the expression level of SSL-derived RNA from subjects (read column) were collected. The RPM values ​​were then converted to RPM values ​​corrected for differences in the total number of reads between samples. However, expression data that are not missing values ​​were obtained for more than 90% of all samples. Only 7429 genes were used in the following analysis. To build the machine learning model, we used a negative binomial distribution. To approximate the RPM values ​​to a normal distribution, we add an integer 1 to the logarithm of base 2 (Log2(R PM+1) values ​​were used.

[0073] 2) Selection of feature genes In order to select feature genes using the random forest algorithm, we 7429 genes for which non-missing expression data were available for more than 90% of the samples in the pool. The Log2(RPM+1) value of the child was used as the explanatory variable, and the healthy subjects (HL) and AD subjects were used as the objective variables. We used the random forest algorithm in the "caret" package of the R language. Specify as the method and find the optimal value of the number of variables (mtry value) used to build a decision tree. The mtry value determined by tuning was used to perform random forwarding. Run the Rest algorithm and calculate the top 150 genes by variable importance based on the Gini coefficient. (Tables 3-1 to 3-4). 127 genes that have not been reported to be related to atopic dermatitis (marked with * in bold in each table) (shown in red) were selected as feature genes.

[0074] [Table 3-1]

[0075] [Table 3-2]

[0076] [Table 3-3]

[0077] [Table 3-4]

[0078] 3) Model construction The Log2(RPM+1) values ​​of the above 150 genes or 127 genes were used as explanatory variables, and H L and AD were used as the objective variables. Specify the forest algorithm as the method and select the variables used to build a decision tree. The optimum value of mtry was tuned. Using the values, we ran the random forest algorithm and estimated the OOB error rate (OOB error As a result, the OOB rate was calculated using the 150 gene model. The error rate was 6.98%, and the model using 127 genes was 6.98%. It was.

[0079] Example 3: Construction of a discrimination model using differentially expressed genes 1) Usage Data In the same manner as in Example 1, data on the expression level of SSL-derived RNA from subjects (read column) were collected. The total number of reads (RPM) was calculated and converted to an RPM value corrected for differences in the total number of reads between samples. To build a machine learning model, we use integers to approximate the RPM values ​​that follow a negative binomial distribution to a normal distribution. The logarithm value of base 2 with 1 added (log2(RPM+1) value) was used.

[0080] 2) Selection of feature genes In Example 1, 123 genes whose expression was significantly altered in AD compared to healthy subjects (HL) were identified. Genes (Table 1-1 to 1-3) or genes that have been reported to be associated with atopic dermatitis 107 unreported genes (marked with * in bold in each table) were used as feature genes. Selected.

[0081] 3) Model construction The Log2(RPM+1) values ​​of the above 123 genes or 107 genes were used as explanatory variables, and H L and AD were used as the objective variables. Specify the forest algorithm as the method and select the variables used to build a decision tree. The optimum value of mtry was tuned. Using the values, we ran the random forest algorithm to calculate the OOB error rate. As a result, the OOB error ra was calculated using the model with 123 genes. te was 13.95%, and the model using 107 genes was 13.95%.

[0082] Example 4: Construction of a discriminant model using feature genes extracted by the Boruta method 1) Usage Data In the same manner as in Example 1, data on the expression level of SSL-derived RNA from subjects (read column) were collected. The RPM values ​​were then converted to RPM values ​​corrected for differences in the total number of reads between samples. However, expression data that are not missing values ​​were obtained for more than 90% of all samples. Only 7429 genes were used in the following analysis. To build the machine learning model, we used a negative binomial distribution. To approximate the RPM values ​​to a normal distribution, we add an integer 1 to the logarithm of base 2 (log2(R PM+1) values ​​were used.

[0083] 2) Selection of feature genes Non-missing expression data were obtained for more than 90% of all samples74 The Log2(RPM+1) values ​​of 29 genes were used as explanatory variables, and the objective was to differentiate between healthy subjects (HL) and AD. The algorithm in the "Boruta" package in R was implemented using the maximum The number of runs was set to 1000, and 45 genes with p-values ​​less than 0.01 were calculated (Table 4). These 45 genes or genes that have not been reported to be associated with atopic dermatitis 39 genes (marked with * in bold in Table 4) that were not present were selected as feature genes.

[0084] [Table 4]

[0085] 3) Model construction The Log2(RPM+1) values ​​of the above 45 genes or 39 genes were used as explanatory variables, and HL and AD was used as the target variable. Specify the rest algorithm as the method and the number of variables used to build a decision tree ( The optimum value of mtry was tuned. The mtry value determined by tuning was Random forest algorithm was run using the data to calculate the OOB error rate. As a result, the OOB error rate was 1.0 in the model using 45 genes. The model using 39 genes yielded 6.98% and 9.3% respectively.

[0086] Example 5 Construction of a discrimination model based on feature genes used in multiple examples 1) Usage Data In the same manner as in Example 1, data on the expression level of SSL-derived RNA from subjects (read column) were collected. The RMP value was calculated by correcting for the difference in the total number of reads between samples. To build a machine learning model, we use integers to approximate the RPM values ​​that follow a negative binomial distribution to a normal distribution. The logarithm value of base 2 with 1 added (log2(RPM+1) value) was used.

[0087] 2) Selection of feature genes Of the feature genes used in Examples 2 to 4, in all Examples 2 to 4 The genes used were MECR, RASA4CP, HMGCS1, ARRDC4, and EIF1AD. , FDFT1, ZNF706, TEX2, TMPRSS11E, RPS6KB2, CTB P1, ZNF335, CAPN1, DGKA, PPP1R9B, SPDYE7P, DNA These 19 genes were SE1L1, GNB2, and CSNK1G2 (Table 5). Among the 17 genes in the study, no relationship to atopic dermatitis has been reported (see Table 5). The genes marked with * and shown in bold were selected as feature genes.

[0088] 3) Model construction The Log2(RPM+1) values ​​of the above 17 genes were used as explanatory variables, and HL and AD were used as objective variables. The random forest algorithm was used in the “caret” package of the R language. Specify the algorithm as the method and set the maximum number of variables (mtry value) used to build a decision tree. The optimum value was tuned. The mtry value determined by tuning was used to randomly The forest algorithm was run and the OOB error rate was calculated. The OOB error rate was 6.98%.

[0089] [Table 5]

Claims

1. A method for detecting atopic dermatitis in a subject, comprising measuring the expression level of mRNA of at least one gene selected from a group of 17 genes consisting of MECR, RASA4CP, ARRDC4, EIF1AD, FDFT1, ZNF706, TEX2, TMPRSS11E, RPS6KB2, CTBP1, ZNF335, DGKA, PPP1R9B, SPDYE7P, DNASE1L1, GNB2 and CSNK1G2 in surface skin lipids collected from the entire face of the subject.

2. The method of claim 1, wherein the detection comprises comparing a measured value of the expression level with a reference value of the expression level of mRNA of each of the genes to evaluate the presence or absence of atopic dermatitis.

3. A method for evaluating the presence or absence of atopic dermatitis in a subject, wherein the detection comprises creating a discriminant for distinguishing between atopic dermatitis patients and healthy subjects using the mRNA expression level of the gene derived from atopic dermatitis patients and the measured value of the mRNA expression level of the same gene derived from healthy subjects as teacher samples, substituting the measured value of the mRNA expression level of the same gene obtained from surface lipids of the skin collected from the entire face of the subject into the discriminant, and comparing the obtained result with a reference value, wherein the mRNA expression levels of all genes in the 17 gene group are measured, the method described in claim 1 or 2.