Method for detecting infant atopic dermatitis

Abstract

To provide a marker for detecting infant atopic dermatitis, and a method for detecting infant atopic dermatitis using the marker.SOLUTION: A method for detecting infant atopic dermatitis in a subject includes the step of measuring the expression level of at least one gene selected from seven gene groups consisting of IMPDH2, ERI1, FBXW2, STK17B, TAGLN2, AMICA1 and HNRNPA1 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 infantile atopic dermatitis using an infantile atopic dermatitis marker.

Background Art

[0002] Atopic dermatitis (hereinafter also referred to as "AD") is an eczematous skin disease that mainly develops in individuals with an atopic predisposition. Typical symptoms of atopic dermatitis include chronic and recurrent itching, rashes, erythema, etc. that occur bilaterally, as well as parakeratosis, reduced barrier function, and dry skin. Most cases of atopic dermatitis develop in infants and tend to improve with growth, but in recent years, adult-type and refractory atopic dermatitis have also been increasing. Newborns / infants with genetic factors that make them prone to allergies and atopy are known to develop various allergic diseases such as infantile eczema, atopic dermatitis, food allergies, as well as bronchial asthma and allergic rhinitis as they grow (allergy march). As such, allergic diseases are likely to lead to other allergic diseases once one disease develops, and their treatment often takes a long time. Therefore, it is necessary to suppress the onset of allergic diseases in infants.

[0003]

[0004] As a method for detecting atopic dermatitis using biomarkers, the number of peripheral blood eosinophils in blood, serum total IgE value, LDH (lactate dehydrogenase) value, Thymus and Activation-Regulated Chemokine (TARC) and Squamous cell carcinoma antigen 2 (SCCA2) in serum are detected. ​​​​​​​​​​​​​To detect (Non-patent documents 1, 2, 3), and Staphylococcus aureus in the skin microbiota. Detecting the agrC mutation-dependent RNAIII gene expression level in bacteria (Patent Document 1), etc. It has been proposed. However, it remains to be seen whether these markers are applicable to atopic dermatitis in infants and young children. It is not clear.

[0005] On the other hand, analysis of nucleic acids such as DNA and RNA in biological samples reveals the current state of the human body, and furthermore... Technologies are being developed to predict future physiological states. Nucleic acid-based analysis is a comprehensive analytical method. A method has been established that allows for obtaining a wealth of information in a single analysis, and also for single nucleotide polymorphisms and RNA function, etc. Based on numerous research reports on the subject, one advantage is that it is easy to functionally link the analysis results. Nucleic acids of biological origin can be extracted from bodily fluids such as blood, secretions, tissues, etc. Recently, it has been found that skin surface lipids (SSL) are included in RNA is used as a sample for the analysis of living organisms, including the use of SSL to analyze the epidermis, sweat glands, hair follicles, and sebaceous glands. It has been reported that the marker gene can be detected (Patent Document 2). [Prior art documents] [Patent Documents]

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

[0007] [Non-Patent Document 1] Sugawara et al., Allergy (2002) 57:180-181

Non-Patent Document 2

Non-Patent Document 3

Summary of the Invention

Problems to be Solved by the Invention

[0008] The present invention relates to a marker for detecting infantile atopic dermatitis and a method for detecting infantile atopic dermatitis using the marker.

Means for Solving the Problems

[0009] The present inventors collected SSL from infants with atopic dermatitis and healthy infants without allergic predisposition, and comprehensively analyzed the expression state of RNA contained in SSL as sequence information. As a result, it was found that the expression levels of specific genes were significantly different between the two, and that infantile atopic dermatitis could be detected using this as a target. That is, the present invention relates to the following 1) to 3).

[0010] 1) A method for detecting infantile atopic dermatitis in a subject, comprising a step of measuring the expression level of at least one gene selected from a group of 7 genes consisting of IMPDH2, ERI1, FBXW2, STK17B, TAGLN2, AMICA1 and HNRNPA1 or its expression product in a biological sample collected from the subject. 1) 2) A test kit for detecting infantile atopic dermatitis used in the method of 1), containing an oligonucleotide that specifically hybridizes with the gene or an antibody that recognizes the expression product of the gene. 3) At least one gene selected from the gene groups shown in Tables B-1 to B-2 below or a detection marker for infantile atopic dermatitis comprising the expression product thereof.

Advantages of the Invention

[0011] According to the present invention, infantile atopic dermatitis can be detected simply and non-invasively at an early stage with high accuracy, sensitivity, and specificity.

Embodiments for Carrying Out the Invention

[0012] All patent documents, non-patent documents, and other publications cited in this specification are hereby incorporated by reference in their entirety. into this specification.

[0013] In the present invention, the terms "nucleic acid" or "polynucleotide" mean DNA or RNA. DNA includes cDNA, genomic DNA, and synthetic DNA, and "RNA" includes total RNA, mRNA, rRNA, tRNA, non-coding RNA, and synthetic RNA.

[0014] In the present invention, the term "gene" refers to double-stranded DNA including human genomic DNA, as well as single-stranded DNA (sense strand) including cDNA, single-stranded DNA (antisense strand) having a sequence complementary to the sense strand, and fragments thereof, which means that the sequence information of the bases constituting the DNA contains some biological information. Moreover, the "gene" is not limited to the "gene" represented by a specific base sequence, but also includes homologs (i.e., homologs or orthologs), mutants such as gene polymorphisms, and derivatives encoding nucleic acids. The gene names disclosed herein are as recorded in the NCBI ([www.ncbi.nlm.nih.gov / ]) It follows the Official Symbol listed. On the other hand, the Gene Ontology (GO) Regarding this, follow the Pathway ID listed in String([string-db.org / ]). cormorant.

[0015] In this invention, the "expression product" of a gene includes the transcript and translation product of the gene. This is a concept. A "transcript" is RNA that is produced when a gene (DNA) is transcribed. A "translation product" is a gene-encoded protein that is synthesized through translation based on RNA. It means...

[0016] In this invention, "atopic dermatitis" refers to a condition characterized by repeated exacerbations and remissions, characterized by itching and eczema. This refers to a disease whose primary cause is a rash, and many of its patients are believed to have an atopic predisposition. Topi predispositions include: i) Family history / past medical history (bronchial asthma, allergic rhinitis / conjunctivitis, etc.) (i) any or more of the totopic dermatitis diseases, or (ii) prone to producing IgE antibodies One possible contributing factor is... In this invention, infant atopic dermatitis refers to the age range from birth to school age, specifically... It primarily targets infants and toddlers from 0 to 5 years old. Skin rashes begin on the head and face in infancy. It often descends to the trunk and limbs, and in infancy, the facial rash decreases, while the rash mainly appears in the neck and limb joints. It is characterized by appearing in the mind. Differences between infantile atopic dermatitis and adult atopic dermatitis. Regarding this, in recent years, adult atopic dermatitis has been chronic compared to infantile atopic dermatitis. Abnormal epidermal keratinization associated with inflammatory abnormalities has been reported (Journal of Allergy) and clinical immunology, volume 141, issue 6, June 2018, pageg 2094-2106), report The number of reports is small, and the situation is not clear.

[0017] In this invention, "detection" of infant atopic dermatitis means the detection of infant atopic dermatitis It means clarifying existence or non-existence, and includes inspection, measurement, judgment, evaluation, or evaluation support. It can also be rephrased using different terms. Furthermore, in this specification, the terms "judgment" or "evaluation" are used in a different way. The term does not include any judgment or evaluation by a physician.

[0018] In this specification, "features" are synonymous with "explanatory variables" in machine learning, and A gene or its expression product to be used in machine learning selected from atopic dermatitis detection markers These are collectively referred to as "feature genes."

[0019] As shown in the examples described later, SSL was administered to 28 healthy children and 25 children with atopic dermatitis. Regarding the RNA expression level data (read count value) extracted from DESeq2( Count values ​​corrected using Love MI et al. Genome Biol. 2014 (Normaliz Using the ED count value, the likelihood in children with atopic dermatitis compared to healthy children was calculated. By extracting RNA whose p-value correction value (FDR) from the ratio test is less than 0.25, 61 genes with increased expression, 310 genes with decreased expression (total of 371 genes (Table 1-1~1-9) ) was identified. In the table, the gene indicated by "UP" is expressed in children with atopic dermatitis. The gene that causes the bell to rise, and the gene indicated as "DOWN," is associated with children with atopic dermatitis. This is a gene whose expression level decreases. Therefore, the gene selected from this group of 371 genes or its expression product is milk This could serve as an infant atopic dermatitis marker for detecting atopic dermatitis in young children. Of these genes, 318 genes (indicated with an asterisk and in bold in Tables 1-1 to 1-9) are... This gene has not been previously reported to be associated with atopic dermatitis.

[0020] In addition, data on the expression levels of all SSL-derived RNAs detected from the subjects (3486 genes) The child's Log2(RPM+1) value was used as an explanatory variable, comparing healthy individuals with infants and children with atopic dermatitis. The target variable is Random Forest (Breiman L. Machine) as the machine learning algorithm. We attempted to extract feature genes and construct a predictive model using Learning (2001) 45;5-32). As shown in the examples described later, the top 100 genes in terms of variable importance based on the Gini coefficient (Table 3) -1~3-3) were selected as feature genes, and a model using these was developed for atopic dermatitis in infants. It was shown that dermatitis can be predicted. Therefore, the gene selected from this group of 100 genes or its expression product is milk This could be a suitable infant atopic dermatitis marker for detecting atopic dermatitis in infants. Of the genes in question, 92 genes (marked with * and shown in bold in Tables 3-1 to 3-3) are This gene has not been previously reported to be associated with atopic dermatitis, and is a novel gene for infants. These are atopic dermatitis markers. As shown in the examples described later, these novel infant a Predictive models using atopic dermatitis markers can also predict atopic dermatitis in infants and young children. be.

[0021] Furthermore, the above 371 genes or Using the expression level data (Log2(RPM+1) values) of 318 of those genes, When we attempted to build a predictive model using random forests, in all cases, infants It has been shown that childhood atopic dermatitis can be predicted.

[0022] Furthermore, as a machine learning algorithm, the Boruta method (Kursa et al. Fundamental Infor Feature gene extraction using maticae (2010) 101;271-286) (maximum number of trials: 1000, When the p-value (less than 0.01) was analyzed, nine genes (Table 4) were extracted as characteristic genes. As shown in the examples described later, a prediction model using random forests with this method can predict milk It has been shown that it is possible to predict atopic dermatitis in infants. Therefore, the gene selected from these nine gene groups or its expression product is used in infants and young children. This could serve as an infant atopic dermatitis marker for detecting atopic dermatitis. Seven of the genes in the gene group (marked with * and shown in bold in Table 4) have previously been associated with atopic dermatitis. This gene has not been reported to be related to dermatitis at all, and is a novel gene for infantile atopic dermatitis. This is a car. As shown in the examples described later, these novel infant atopic dermatitis markers Predictive models using this method can also predict atopic dermatitis in infants and young children.

[0023] The 371 gene groups shown in Tables 1-1 to 1-9, extracted from the differential expression analysis described above, A) and Tables 3-1 to 3-3 show the feature genes selected by random forest analysis. A group of 100 genes (B) and a set of feature genes selected by the Boruta method. The sum of the nine gene groups (C) shown in Table 4 (A∪B∪C) is 441 genes (Table 4 A-1 to A-2) can all be markers for atopic dermatitis in infants and young children, of which 383 types exist. Tables B-1 to B-2 show novel markers for atopic dermatitis in infants and young children.

[0024] [Table A-1]

[0025] [Table A-2]

[0026] [Table B-1]

[0027] [Table B-2]

[0028] The present invention includes IMPDH2, ERI1, FBXW2, STK17B, TAGLN2, and AMI. The seven genes consisting of CA1 and HNRNPA1 were analyzed using the random forest method described above. The 100 gene groups (B) listed in Tables 3-1 to 3-3, which were selected as characteristic genes, The nine gene groups listed in Table 4 (C) were selected as characteristic genes using the Boruta method. The common gene (B∩C) has not been previously reported to have any relationship with atopic dermatitis. These are genes that are not present (indicated by an asterisk and in bold in each table). Therefore, these gene groups At least one selected gene or its expression product is used to test for atopic dermatitis in infants. It is particularly useful as a novel marker for atopic dermatitis in infants and young children. Furthermore, among these, IMPDH2, ERI1, and FBXW2 were analyzed using the differential expression analysis described above. The genes included in the 371 gene group (A) listed in Tables 1-1 to 1-9 extracted by ( Since A∩B∩C), it is considered a more preferable novel marker for atopic dermatitis in infants. El. Each of these seven genes can individually serve as a marker for atopic dermatitis in infants, It is preferable to use a combination of two or more types, preferably four or more, and more preferably six or more. Ideally, it is preferable to use all seven types.

[0029] Furthermore, the present invention relates to ABHD8, GPT2, PLIN2, FAM100B, YPEL2, M AP1LC3B2, RLF, KIAA0930, UBE2R2, HK2, USF2, PD IA3P, HNRNPUL1, SEC61G, DNAJB11, SDHD, NDUFS7 ECH1, CASS4, CLEC4A, SNRPD1, SLC7A11 and SNX8 The 23 genes listed above are those extracted from the differential expression analysis and are shown in Tables 1-1 to 1-9. A set of 371 genes (A) and genes selected as feature genes using a random forest algorithm. It is included in the common part of the 100 gene groups (B) listed in Tables 3-1 to 3-3, and conventionally atopic From genes that have not been reported to be associated with sexual dermatitis, the aforementioned IMPDH2, ERI1 and These are the genes excluding FBXW2. Therefore, fewer genes are selected from this group of genes. At least one gene or its expression product is also used to detect atopic dermatitis in infants. It is useful as a novel marker for atopic dermatitis in infants and young children.

[0030] Furthermore, the genes that can serve as markers for atopic dermatitis in infants and young children (hereinafter referred to as "target genes") are also referred to as "target genes." (Also known as ) can serve as a biomarker for detecting atopic dermatitis in infants and young children. To the extent that, a gene having a base sequence substantially identical to the base sequence of the DNA constituting the gene in question. This also includes. Here, substantially identical base sequences refer to, for example, homology calculation algorithms. Using NCBI BLAST, expect value = 10; allow gaps; filtering = ON; match score When searching under the conditions A=1; mismatch score=-3, the DN constituting the gene in question is... The base sequence of A is 90% or more, preferably 95% or more, and more preferably 98% or more. It means having one nature.

[0031] The present invention relates to a method for detecting atopic dermatitis in infants and young children, and involves a biological sample taken from a subject. For example, target genes include IMPDH2, ERI1, FBXW2, STK17B, Selected from a group of seven genes consisting of TAGLN2, AMICA1, and HNRNPA1. The process includes measuring the expression level of at least one gene or its expression product.

[0032] The biological sample used in this invention is a sample that shows the onset and progression of atopic dermatitis. Any tissue or biomaterial in which the gene expression of the invention is altered is acceptable. Specifically, organs, skin, blood Prepared from bodily fluids such as blood, urine, saliva, sweat, surface lipids (SSL), tissue exudate, etc. Examples include serum, plasma, feces, hair, etc., preferably skin or surface lipids (SS). L), more preferably, surface lipids (SSL) of the skin. Furthermore, the subjects from whom biological samples are collected are not particularly limited in terms of gender, race, etc., as long as they are infants or young children. However, for infants and young children who require detection of atopic dermatitis or who are suspected of having atopic dermatitis Infants and toddlers are preferred.

[0033] Here, "superficial lipids (SSL)" refers to the lipid-soluble fraction present on the surface of the skin. It is sometimes called sebum. Generally, sebum is secreted from exocrine glands such as sebaceous glands in the skin. It mainly contains secretions and exists on the skin surface in the form of a thin layer covering the skin surface. This includes RNA expressed in skin cells. (See Patent Document 3 above). Furthermore, in this specification Unless otherwise specified, "skin" refers to the epidermis, dermis, hair follicles, sweat glands, and sebaceous glands of the body surface. It is a general term for the region including other tissues such as glands.

[0034] For collecting SSL from the subject's skin, a method used for recovering or removing SSL from the skin is used. Any means available can be employed. Preferably, the SSL absorbent material described later, S SL adhesive material or an instrument for scraping SSL off the skin can be used. As for L absorbent materials or SSL adhesive materials, any material that has affinity for SSL is particularly suitable. The following are examples of materials, but are not limited to polypropylene, pulp, etc. The method of collecting SSL from the skin. A more detailed example of the order is SS applied to sheet-like materials such as oil-blotting paper and oil-blotting film. Methods for absorbing L, methods for adhering SSL to glass plates, tape, etc., spatula, screw Methods include scraping off and recovering SSL using a rapist, etc. SSL adsorption To improve performance, SSL absorbent materials that have been pre-soaked with a highly lipid-soluble solvent may be used. On the other hand, if SSL absorbent material contains highly water-soluble solvents or moisture, the adsorption of SSL is inhibited. Because it is harmful, a highly water-soluble solvent or one with a low water content is preferable. SSL absorbency The material is preferably used in a dry state. The skin site from which SSL is collected is as follows: The skin is not particularly limited and includes the skin of any part of the body such as the head, face, neck, trunk, limbs, etc. Areas with high oil secretion, such as the skin of the face, are preferable.

[0035] RNA-containing SSLs collected from subjects may be stored for a certain period of time. To minimize the degradation of the RNA contained in L, it is necessary to process it under low-temperature conditions as quickly as possible after collection. Storage is preferable. The storage temperature conditions for the RNA-containing SSL in this invention are 0°C. The following is acceptable, preferably -20±20℃ to -80±20℃, more preferably -20 ±10℃ to -80±10℃, more preferably -20±20℃ to -40±20℃, and further Preferably -20±10℃ to -40±10℃, more preferably -20±10℃, further Preferably, the temperature is -20±5℃. The storage period for the RNA-containing SSL under these low-temperature conditions is particularly While not limited to, preferably 12 months or less, for example, 6 hours or more and 12 months or less, more preferably Or less than 6 months, for example, more than 1 day and less than 6 months, more preferably less than 3 months, for example It is between 3 days and 3 months.

[0036] In the present invention, the target for measuring the expression level of the target gene or its expression product is RN cDNA artificially synthesized from A, DNA that encodes its RNA, and the RNA The encoded protein, the molecule that interacts with the protein, and the RNA that interacts with it. Examples include molecules that interact with DNA, or molecules that interact with DNA. Here, RNA, DNA or Molecules that interact with proteins include DNA, RNA, proteins, polysaccharides, and oligosaccharides. Sugars, monosaccharides, lipids, fatty acids, and their phosphorylated, alkylated, and glycoadduct compounds, etc. The following complexes are examples. Furthermore, the expression level refers to the expression of the gene or expression product in question. It comprehensively refers to the actual amount and activity.

[0037] In the method of the present invention, in a preferred embodiment, SSL is used as the biological sample. In this case, the expression level of RNA contained in SSL is analyzed, specifically the RNA expression level is reversed. After being converted to cDNA by transcription, the cDNA or its amplified product is measured. RNA extraction from SSL is typically 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 Methods using columns such as l(registered trademark), special magnetic particles coated with silica Method of use, Solid Phase Reversible Immobilizat Methods using ion magnetic particles, extraction using commercially available RNA extraction reagents such as ISOGEN, etc. It can be used.

[0038] For this reverse transcription, primers targeting the specific RNA to be analyzed may be used, For comprehensive nucleic acid storage and analysis, it is preferable to use random primers. For the reverse transcription, a general reverse transcriptase or reverse transcription reagent kit can be used. For this purpose, highly accurate and efficient reverse transcriptase or reverse transcriptase reagent kits are used, for example, For example, M-MLV Reverse Transcriptase and its variants, or commercially available reverse transcriptase or reverse transcriptase reagent kit, for example, PrimeScript (registered trademark) ) Reverse Transcriptase series (Takara Bio Inc.), Supe rScript (registered trademark) Reverse Transcriptase series (T Examples include Hermo Scientific, etc. (Registered) Trademark)III Reverse Transcriptase, SuperScript( (Registered Trademark) VILO cDNA Synthesis kit (all Thermo Scientific (and others) are preferably used. The extension reaction in the reverse transcription is carried out at a temperature preferably 42°C ± 1°C, more preferably 42°C. The temperature is adjusted to ±0.5℃, more preferably to 42℃±0.25℃, while the reaction time is preferably It is preferable to adjust the time to 60 minutes or more, more preferably to 80 to 120 minutes.

[0039] Methods for measuring expression levels include, when targeting RNA, cDNA, or DNA, these PCR method using DNA that hybridizes to the target DNA as a primer, real-time RT-PCR method Nucleic acid amplification methods such as multiplex PCR, SmartAmp, and LAMP. , hybridization methods that use nucleic acids that hybridize to these as probes ( DNA chips, DNA microarrays, dot blot hybridization, slot Toblot hybridization, Northern blot hybridization, etc., salt Choose from methods for determining the base sequence (sequencing), or a combination of both. It is possible.

[0040] PCR uses a primer pair that targets the specific DNA to be analyzed. While it is possible to amplify only the DNA of a species, multiple primer pairs can be used to amplify multiple specific DNAs simultaneously. DNA may be amplified. Preferably, the PCR is multiplex PCR. Chiplex PCR uses multiple primer pairs simultaneously in the PCR reaction system. Multiplex PCR is a method for simultaneously amplifying multiple gene regions. (For example, Ion AmpliSeqTranscriptome Human Gen Using the e Expression Kit (Life Technologies Japan Co., Ltd., etc.) It can be implemented. The temperatures for the annealing and extension reactions in this PCR depend on the primers used. Therefore, it cannot be said in general, but when using the above multiplex PCR kit, it is preferable The temperature is 62℃±1℃, more preferably 62℃±0.5℃, and even more preferably 62℃±0.25℃. The temperature is °C. Therefore, in this PCR, preferably the annealing and extension reactions are performed in one step. The process is carried out by the annealing and extension reaction. The time of the steps in the annealing and extension reaction is determined by the amount of DNA to be amplified. The duration can be adjusted depending on factors such as Iz, but is preferably 14 to 18 minutes. The conditions for the denaturation reaction can be adjusted depending on the DNA to be amplified, but preferably 95-99 The time is 10-60 seconds at °C. Reverse transcription and PCR at the above temperatures and times are common. This can be performed using a thermal cycler used for PCR.

[0041] The purification of the reaction product obtained by the PCR is performed by size separation of the reaction product. This is preferable. By size separation, the target PCR reaction product is separated from the PCR reaction solution. It can be separated from lymers and other impurities. DNA size separation is, for example, Size separation columns, size separation chips, magnetic beads that can be used for size separation, etc. This can be done. A preferred example of magnetic beads that can be used for size separation is Ampu Solid Phase Reversible Immobilizers such as re XP Examples include tion(SPRI) magnetic beads.

[0042] Further processing of the purified PCR reaction product is required for subsequent quantitative analysis. For example, the purified PCR reaction product may be used for DNA sequencing. ...to prepare appropriate buffer solutions, or to prepare PCR-amplified DNA containing PCR plastic Even if you cut the imer region or add an adapter sequence to the amplified DNA Good. For example, prepare the purified PCR reaction product in a buffer solution and apply it to the amplified DNA. Then, the PCR primer sequence was removed and adapter ligation was performed, and the reaction product obtained This can be amplified as needed to prepare a library for quantitative analysis. These operations are, for example, SuperScript® VILO cDNA Sy The 5x included with the inthesis kit (Life Technologies Japan Co., Ltd.) VILO RT Reaction Mix and Ion AmpliSeq Tran scriptome Human Gene Expression Kit (LifeTe The 5x Ion AmpliSeq HiF included with the Knology Japan Co., Ltd. i Mix and Ion AmpliSeq Transcriptome Human Using the Gene Expression Core Panel, the parts included with each kit This can be done according to Rotcor.

[0043] Northern blot hybridization is used to obtain a target gene or a gene derived from it. When measuring nucleic acid expression levels, for example, first, the probe DNA is treated with a radioisotope, a fluorescent substance, etc. The DNA is labeled with a substance, and then the resulting labeled DNA is transferred to a nylon membrane or the like according to a standard method. The RNA derived from the transferred biological sample is hybridized with the formed label. The double helix of DNA and RNA is measured by detecting a signal derived from a label. One possible method is to...

[0044] When measuring the expression level of a target gene or nucleic acid derived therefrom using the RT-PCR method: For example, first prepare cDNA from RNA derived from a biological sample according to a standard method, and use this as a template. A pair of primers (the above cDNA()) are prepared to amplify the target gene of the present invention. The positive chain (bonded to the negative chain) and the reverse chain (bonded to the positive chain) are hybridized with this. Then, PCR is performed according to standard procedures, and the resulting amplified double-stranded DNA is detected. For DNA detection, primers that have been pre-labeled with RI, fluorescent substances, etc., are used in the above PC Methods for detecting labeled double-stranded DNA produced by performing R can be used. ru.

[0045] DNA microarrays are used to measure the expression levels of target genes or nucleic acids derived therefrom. For example, if the support contains a small amount of nucleic acid (cDNA or DNA) derived from the target gene of the present invention Using an array immobilized with at least one species, labeled cDNA or cRN was prepared from mRNA. By binding A to a microarray and detecting the label on the microarray, m RNA expression levels can be measured. The nucleic acids immobilized on the array are specific under stringent conditions (that is, Any nucleic acid that hybridizes (effectively only to the target nucleic acid) is sufficient, for example, this The nucleic acid may contain the complete sequence of the target gene, or it may consist of a partial sequence. Good. Here, "partial sequence" refers to nucleic acids consisting of at least 15 to 25 base pairs. The stringent conditions here are typically around "1×SSC, 0.1%SDS, 37°C". The washing conditions can be listed as follows, and a more stringent hybridization condition is "0.5×SS C, 0.1% SDS, 42°C is a reasonable estimate, and even stricter hybridization conditions include "0.1 Conditions such as "×SSC, 0.1% SDS, 65°C" can be cited. Hybridize Conditions are as per J. Sambrook et al., Molecular Cloning: A Laboratory Manual, Thrd Edition. This is described in publications such as Cold Spring Harbor Laboratory Press (2001).

[0046] When measuring the expression level of a target gene or nucleic acid derived therefrom by sequencing: For example, next-generation sequencers (e.g., Ion S5 / XL system, Life Technology) One example is that it is analyzed using (Series Japan Co., Ltd.). RNA expression can be quantified based on the number of reads (read count).

[0047] The probe or primer used in the above measurement, i.e., the target gene of the present invention or A primer for specifically recognizing and amplifying nucleic acids derived therefrom, or the RNA or it This includes probes for specifically detecting nucleic acids derived from the said, It can be designed based on the base sequence that makes up the target gene. Here, "specifically recognized" "To do so" means, for example, in the Northern blot method, substantially the target gene of the present invention or its It is possible to detect only nucleic acids derived from [the specified source], and for example, in the RT-PCR method, it is practically [the specified source]. The detection or product is such that only the nucleic acid is amplified, and the nucleus derived from the gene is amplified. This means that it can be determined to be an acid. Specifically, the DNA or its complementary strand consisting of the base sequence constituting the target gene of the present invention Oligonucleotides containing a certain number of complementary nucleotides can be used. The "complementary strand" refers to a double-stranded DNA consisting of A:T (U in the case of RNA) and G:C base pairs. This refers to one chain in relation to the other. Furthermore, "complementary" means that a certain number of consecutive nuclei This is not limited to cases where the rheotide region is a completely complementary sequence, but is preferably 80% or more, more preferably Alternatively, it is sufficient if the base sequences have 90% or more identity, and more preferably 95% or more. The identity of the base sequence can be determined by algorithms such as BLAST. When such oligonucleotides are used as primers, specific annealing is required. It is sufficient that the chain can be extended, and usually, for example, 10 bases or more, preferably 15 bases or more. Preferably 20 bases or more, and for example 100 bases or less, preferably 50 bases or less. Preferably, it has a chain length of 35 bases or less. Also, it can be used as a probe. In such cases, it is sufficient if specific hybridization can be achieved, and the target gene of the present invention is constructed. The sequence of at least part or all of the DNA (or its complementary strand) consisting of the following base sequence It has, for example, 10 bases or more, preferably 15 bases or more, and for example 100 bases or less, preferably A chain length of 50 bases or less, more preferably 25 bases or less, is used. Note that "oligonucleotide" here can be DNA or RNA. It may be synthetic or natural. Or the probe used for hybridization is Usually, labeled versions are used.

[0048] Furthermore, the translation product (protein) of the target gene of the present invention, and the interaction with said protein When measuring molecules, molecules that interact with RNA, or molecules that interact with DNA, Rotain chip analysis, immunoassay (e.g., ELISA), mass spectrometry (e.g., LC- MS / MS, MALDI-TOF / MS), 1-hybrid method (PNAS 100, 12271-1227) Methods such as 6(2003)) and the 2-hybrid method (Biol. Reprod. 58, 302-311 (1998)) It can be used and selected appropriately depending on the subject. For example, when a protein is used as the target of measurement, the antibody against the expression product of the present invention The body is brought into contact with a biological sample, and the protein in the sample bound to the antibody is detected, and the level This is performed by measuring the primary antibody. For example, according to the Western blot method, the primary antibody After using the above antibody as the primary antibody, a radioactive isotope, fluorescent substance, or enzyme is used as the secondary antibody. Using antibodies that bind to identified primary antibodies, the primary antibodies are labeled, and these labeled substances are derived from The signal is measured using radiation detectors, fluorescence detectors, etc. Furthermore, antibodies against the above translation products may be polyclonal antibodies or monoclonal antibodies. It may be the body. These antibodies can be manufactured according to known methods. For polyclonal antibodies, proteins expressed and purified in E. coli and other bacteria according to standard methods are used. Alternatively, by synthesizing the partial polypeptide of the protein in accordance with conventional methods, non-hybrid rabbits and other animals It is possible to immunize animals and obtain it from the serum of the immunized animals according to a standard method. On the other hand, monoclonal antibodies are proteins expressed and purified in E. coli, etc., according to conventional methods. The partial polypeptide of the protein was immunized to non-human animals such as mice, and the resulting spleen cells were... It can be obtained from hybridoma cells prepared by cell fusion with myeloma cells. Monoclonal antibodies may be prepared using phage display (Griffiths, A .D.; Duncan, AR, Current Opinion in Biotechnology, Volume 9, Number 1, February y 1998, pp. 102-108(7).

[0049] Thus, the target gene of the present invention or its expression product in the biological sample collected from the subject The expression level is measured, and infantile atopic dermatitis is detected based on that expression level. The detection specifically involves determining the expression level of the target gene or its expression product of the present invention. This is done by comparing it to a control level. When analyzing the expression levels of multiple target genes using sequencing, as described above... The data includes the read count value, which is the expression level data, and the total number of samples between the read count values. The RPM value corrected for differences in the number of codes, and the value obtained by converting the said RPM value to a base-2 logarithm (Log2). The RPM value) or the base-2 logarithm with an integer 1 added (Log2(RPM+1) value), or The count value corrected using DESeq2 (Normalized count value) This uses the base-2 logarithm (Log2(count+1) value) obtained by adding the integer 1 as an index. It is 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 The value may also be calculated using methods such as ts per million (TPM). Alternatively, the signal value obtained by the microarray method, and its corrected value, may also be used. Furthermore, when analyzing the expression level of only specific target genes using methods such as RT-PCR, The expression level of the target gene is changed to a relative expression level based on the expression level of housekeeping genes. Methods for analysis using relative quantification, or absolute quantification using a plasmid containing the region of the target gene. A method that quantifies (absolutely quantifies) the copy number and analyzes it is preferred. The number of copies obtained may also be acceptable. Here, "control level" refers, for example, to the target gene or its expression level in a healthy person. The expression level of a substance can be cited. The expression level in healthy individuals is measured from a healthy population of genes. This may also be a statistical value (e.g., mean value) of the expression level of the offspring or its expression product. Target gene If there are multiple offspring, determine the baseline expression level for each gene or its expression product. It is preferable.

[0050] Furthermore, the detection of atopic dermatitis in infants and young children in this invention is performed by identifying the target gene or its This can also be done by increasing / decreasing the expression level of the expression product. In this case, the expression level is derived from the subject. The expression level of the target gene or its expression product in a biological sample is determined by the expression level of each gene or its expression product. It is compared to the cutoff value (reference value) of the object. The cutoff value is determined in advance by the standard value for a healthy person. The expression level of the target gene or its expression product is obtained as reference data, and based on that... The appropriate determination should be made based on statistical values ​​such as the mean and standard deviation of the expression level.

[0051] Furthermore, the expression levels of target genes or their expression products derived from children with atopic dermatitis, Using measurements of the expression levels of target genes or their expression products derived from healthy children, atopic dermatitis We constructed a discriminant formula (predictive model) to distinguish between children with dermatitis and healthy children, and used this discriminant formula. It can detect atopic dermatitis in infants and young children. In other words, it can detect children suffering from atopic dermatitis. The expression level of the target gene or its expression product derived from the source, and the expression level of the target gene or its expression product derived from a healthy child. Using the measured expression levels of the actual product as a training sample, children with atopic dermatitis and healthy children were compared. A discriminant formula (predictive model) is constructed to differentiate between children with atopic dermatitis and children with atopic dermatitis. We will determine the cutoff value (reference value) for distinguishing between healthy and non-healthy children. In creating the discriminant formula, Dimensionality reduction can be performed using component analysis (PCA), allowing the major components to be used as explanatory variables. Furthermore, the level of the target gene or its expression product from biological samples taken from the subjects is determined to be the same. The measurement is performed in the specified manner, the obtained measurement value is substituted into the discriminant formula, and the result obtained from the discriminant formula is referred to. By comparing the values, the presence or absence of atopic dermatitis in infants and young children in the subjects can be determined. It's commendable.

[0052] The variables used to construct a discriminant include explanatory variables and dependent variables. Examples of explanatory variables include... The expression level of the target gene or its expression product selected by the following method can be used. The dependent variable could be, for example, whether the sample is derived from healthy children or from children with atopic dermatitis. You can use ka.

[0053] Feature selection should include statistically significant differences between the two groups being classified, for example, a difference in expression level between two groups. Using the expression levels of genes (differential expression genes) or their expression products that differ significantly between groups It is possible to utilize publicly available algorithms and other tools used in machine learning to identify features. It is possible to extract genes and use their expression levels. For example, the following R Using the expression levels of genes or their expression products with high variable importance in the random forest Alternatively, feature genes can be extracted using the "Boruta" package in the R language, and Expression levels can be used, for example.

[0054] The algorithm used in constructing the discriminant is a publicly known algorithm, such as those used in machine learning. This can be used. An example of a machine learning algorithm is Random Forest. Random forest, Linear kernel support vector machine (SVM li near), rbf kernel support vector machine (SVM rbf) neural network Network (Natural net), Generalized Linear Model (Generalized Linear Model) inear model), regularized linear discriminant analysis (Regularized linea discriminant analysis) (discriminant analysis), regularized logistic regression (Re Examples include gularized logistic regression. The built prediction model is then populated with validation data to calculate the predicted value, and the predicted value is found to be the closest to the actual value. The most suitable model, for example, the model with the highest accuracy, is selected as the optimal prediction model. It can be selected as Dell. Also, the detection rate (Recall) can be calculated from the predicted value and the actual value. The precision and its harmonic mean, the F-score, are calculated, and the F-score is the most accurate. A larger model can be selected as the optimal predictive model.

[0055] When using the Random Forest algorithm in constructing a discriminant, the predictive model As an indicator of accuracy, the out-of-bounds (OOB) error rate is used for estimations of unknown data. This can be calculated (Breiman L. Machine Learning (2001) 45;5-32).

[0056] In random forests, the bootstrap method is used to select all samples. By allowing duplicates, approximately two-thirds of the sample size are randomly selected and called a decision tree. Create a classifier that detects the problem. Samples that are not extracted at this time will be considered Out of Bug (OO This method, called B), uses a single decision tree to predict the objective variable in an out-of-bounds (OOB) scenario, and compares the correct label with the corresponding value. By comparing them, the error rate can be calculated (OOB error rate in decision trees). (ate). Repeat the same process 500 times and perform OOB er in 500 decision trees. The average value of the ror rate is used as the OOB err of the random forest model. It can be expressed as "or rate".

[0057] Note that the number of decision trees (ntree value) used to construct a random forest model is the default value. The standard configuration has 500 strings, but this can be changed to any number from one to any number as needed. Furthermore, the number of variables (mtry value) used to create the discriminant expression for a sample in a single decision tree. By default, this value is the square root of the number of explanatory variables, but it can be changed from one to all as needed. The value can be changed to any of the values ​​up to the number of explanatory variables.

[0058] The “caret” package in the R language can be used to determine the mtry value. Specify a random forest as the method of the "aret" package and select 8 different mtry values. Try different values ​​and select the mtry value that maximizes accuracy as the optimal mtry value. This is possible. Note that the number of trials in the mtry value can be changed to any number of trials as needed. It is possible.

[0059] When using the Random Forest algorithm in constructing a discriminant, the model structure The importance of the explanatory variables used in the construction can be quantified (variable importance). For example, the mean decrease in the Gini coefficient can be used. It is possible.

[0060] The method for determining the cutoff value (reference value) is not particularly restricted and can be determined according to a known method. This can be done. For example, an ROC (Receiver Operator) created using a discriminant can be created. It can be obtained from the characteristic curve. In the ROC curve, the vertical axis represents the probability (sensitivity) of a positive result in a positive patient, and the horizontal axis represents the probability of a negative result. The probability of getting a negative result in a sexually transmitted infection (specificity) is subtracted from 1 (false positive rate), which is the probable result. The ROC curve shows "true positive (sensitivity)" and "false positive (1-specificity)". Regarding this, the value that maximizes "true positive (sensitivity)" - "false positive (1 - specificity)" (Youden The index can be used as the cutoff value (reference value).

[0061] As already mentioned, IMPDH2, ERI1, FBXW2, STK17B, TAGLN 2. Seven genes consisting of AMICA1 and HNRNPA1, and containing the said seven genes The 100 genes shown in Tables 3-1 to 3-3, or the 9 genes shown in Table 4, or Table 1 The 371 genes indicated by -1 to 1-9 were used as characteristic genes for atopic skin in infants. It is possible to build a predictive model that can predict the intensity of flames. Therefore, we will create a discriminant formula to separate the group of infants with atopic dermatitis from the group of healthy infants. In some cases, the target genes are IMPDH2, ERI1, FBXW2, STK17B, TAG One gene selected from seven genes consisting of LN2, AMICA1, and HNRNPA1. Select two or more genes, preferably five or more, more preferably all seven, and their genes or This uses the expression data of the expression product. Furthermore, when selecting multiple genes, these genes In Tables 3-1 to 3-3 of the genes, the feature genes are listed in order of their variable importance, from most important to least important. It is preferable to select and create a discriminant formula for the seven genes shown in Table A above. 441 genes are included, 100 genes are shown in Tables 3-1 to 3-3, and 9 genes are shown in Table 4. In the children, or among the 371 genes shown in Tables 1-1 to 1-9, genes other than the seven genes in question At least one, five or more, ten or more, twenty or more, fifty or more genes selected from or By appropriately adding expression data of the expression product, a discriminant formula can be created to identify atopic skin in infants and young children. It is also possible to detect flames. Genes other than the seven specified genes are shown in Tables 3-1 to 3-3. When selecting from the 100 genes shown, choose from the genes with the highest variable importance, in descending order. Feature inheritance is derived from genes with a variable importance ranking within the top 50, preferably within the top 30. You may select offspring. Furthermore, select genes other than the seven genes mentioned above as feature genes. If you do so, mark the table with an asterisk (*) and use bold text in Tables 1-1 to 1-9, 3-1 to 3-3, and 4. It is preferable to select characteristic genes from the novel atopic dermatitis markers presented. Furthermore, when adding the 371 genes shown in Tables 1-1 to 1-9, the target genes are IMPDH2, ERI1, FBXW2, STK17B, TAGLN2, AMICA1 and In addition to the seven genes consisting of HNRNPA1, ABHD8 and GPT2 are among the 371 genes. , PLIN2, FAM100B, YPEL2, MAP1LC3B2, RLF, KIAA0 930, UBE2R2, HK2, USF2, PDIA3P, HNRNPUL1, SEC6 1G, DNAJB11, SDHD, NDUFS7, ECH1, CASS4, IL7R, C A group of 25 genes: LEC4A, AREG, SNRPD1, SLC7A11, and SNX8. More selected at least one, preferably one of these genes in Tables 3-1 to 3-3 Choose at least one, five or more, ten or more, or twenty or more genes of high importance. A discriminant formula may be created by appropriately adding the expression data of the expression product. The 25 genes in question are listed in Tables 1-1 to 1-9, which were extracted from the differential expression analysis described above. A set of 371 genes (A) and genes selected as feature genes using a random forest algorithm. These are genes included in the common region of the 100 gene groups (B) listed in Tables 3-1 to 3-3. . Preferably, the above 7 genes, the 371 genes shown in Tables 1-1 to 1-9, or 318 genes. Children (shown in bold with an asterisk in Tables 1-1 to 19), 100 shown in Tables 3-1 to 3-3 The genes or 92 genes (shown in bold with an asterisk in Tables 3-1 to 3-3), or shown in Table 4. One possible discriminant formula is one in which the nine genes being considered are feature genes. More preferably, the above 7 genes, 100 genes shown in Tables 3-1 to 3-3, or 92 genes. The genes (marked with * and shown in bold in Tables 3-1 to 3-3), or the nine genes shown in Table 4 One example is a discriminant formula used as a feature gene.

[0062] The present invention provides a test kit for detecting atopic dermatitis in infants and young children, which is isolated from patients. A test for measuring the expression level of the target gene or its expression product in a biological sample. It contains a reagent for testing. Specifically, it contains the target gene of the present invention or nucleic acids derived therefrom. Oligonucleotides that specifically bind (hybridize) to (for example, pry for PCR) Reagents for nucleic acid amplification, hybridization, or the target of the present invention, including (M) Examples include reagents for immunological measurements, including antibodies that recognize gene expression products (proteins). The oligonucleotides, antibodies, etc. included in the kit are publicly known as described above. It can be obtained by this method. In addition to the antibodies and nucleic acids mentioned above, the test kit also contains labeling reagents, buffer solutions, chromogenic substrates, and two Subsequent antibodies, blocking agents, and necessary equipment for the test, as well as positive and negative controls. Control reagents to be used as controls, tools for collecting biological samples (for example, This may include things like oil-absorbing film for collecting SSL (Slipid Saturation).

[0063] Aspects and preferred embodiments of the present invention are shown below. <1> Regarding biological samples taken from subjects, IMPDH2, ERI1, FBXW2 Seven genes consisting of STK17B, TAGLN2, AMICA1, and HNRNPA1. A step of measuring the expression level of at least one gene or its expression product selected from the group. A method for detecting atopic dermatitis in infants and young children, including the subject in question. <2> The gene is selected from a group of three genes consisting of IMPDH2, ERI1, and FBXW2. This includes at least measuring the expression level of a gene or its expression product, <1> Atopic dermatitis in infants and toddlers Methods for detecting sexual dermatitis. <3> The expression level of a gene or its expression product is measured by the amount of mRNA expression. <1> or <2> The method. <4> The gene or its expression product is RNA contained in the lipids on the skin surface of the subject. <1> ~ <3> One of the following methods. <5> The measured expression levels are compared with the reference values ​​of each gene or its expression product, and infants To assess the presence or absence of atopic dermatitis. <1> ~ <4> One of the following methods. <6> The expression level of the aforementioned gene or its expression product derived from a child suffering from atopic dermatitis, and the expression level of a healthy child. Using the measured expression levels of the same gene or its expression product from a normal child as a training sample, A discriminant formula was created to distinguish between children with atopic dermatitis and healthy children, and biological samples were taken from the subjects. The measured expression levels of the gene or its expression product obtained are substituted into the discriminant formula, and By comparing the results with reference values, the presence of atopic dermatitis in infants and young children in the subjects can be determined. To evaluate whether something is present or absent. <1> ~ <4> One of the following methods. <7> The algorithms used in constructing the discriminant include random forests and linear kernels. Support vector machines, RFB kernel support vector machines, neural networks The methods include a general linear model, regularized linear discriminant analysis, or regularized logistic regression.<6> <8> The expression levels of all genes or their expression products from the aforementioned seven gene groups are measured. , <6> or <7> The method. <9> In addition to at least one gene selected from the seven gene groups mentioned above, see Table 3 below. Of the 100 species shown in 1-3-3 and the 9 species shown in Table 4, the genes of the aforementioned 7 species have been excluded. The expression level of at least one gene or its expression product selected from the gene group is measured. ru, <6> ~ <8> One of the following methods. <10> The 100 species shown in Tables 3-1 to 3-3 below were extracted using random forest. These are the characteristic genes that were identified. <9> The method. <11> The nine genes shown in Table 4 below are feature genes extracted using the Boruta method. be, <9> The method. <12> In addition to at least one gene selected from the seven gene groups mentioned above, see Table 1 below. From the 371 types represented by -1 to 1-9, excluding the genes included in the aforementioned 7 types, The expression level of at least one selected gene or its expression product is measured. <6> ~ <8> One of the following methods. <13> In addition to at least one gene selected from the seven gene groups mentioned above, the following two The expression level of at least one gene selected from the five gene groups, or its expression product, is Measured, <11> or <12> The method, ABHD8, GPT2, PLIN2, FAM100B, YPEL2, MAP1LC3B 2, RLF, KIAA0930, UBE2R2, HK2, USF2, PDIA3P, HN RNPUL1, SEC61G, DNAJB11, SDHD, NDUFS7, ECH1, C ASS4, IL7R, CLEC4A, AREG, SNRPD1, SLC7A11 and SN X8. <14> Oligonucleotides that specifically hybridize with the aforementioned gene or nucleic acid derived therefrom A product containing a rheotide or an antibody that recognizes the expression product of the gene, <1> ~ <13> Noi A test kit for detecting atopic dermatitis in infants and young children, used in either of the following methods. <15> Selected from the 383 gene groups shown in B-1 to B-2 above, at least A detection marker for atopic dermatitis in infants and young children, consisting of a single gene or its expression product. <16> IMPDH2, ERI1, FBXW2, STK17B, TAGLN2, AMI At least one gene selected from a group of seven genes consisting of CA1 and HNRNPA1 The child or its expression product, <15> A detection marker for atopic dermatitis in infants and young children. <17> ABHD8, GPT2, PLIN2, FAM100B, YPEL2, MAP1 LC3B2, RLF, KIAA0930, UBE2R2, HK2, USF2, PDIA3 P, HNRNPUL1, SEC61G, DNAJB11, SDHD, NDUFS7, EC It consists of H1, CASS4, CLEC4A, SNRPD1, SLC7A11, and SNX8. <1 5> A detection marker for atopic dermatitis in infants and young children. [Examples]

[0064] The present invention will be described in more detail below based on examples, but the present invention is not limited thereto. No.

[0065] Example 1: Expression changes related to atopic dermatitis in RNA extracted from SSL in infants and young children. Gene detection 1) SSL collection 28 healthy-skinned infants and toddlers (HL) (6 months to 5 years old, male and female), and infants and toddlers with atopic dermatitis. The study included 25 children (boys and girls) aged 6 months to 5 years with atopic dermatitis. This is a diagnosis of mild to moderate atopic dermatitis, characterized by skin lesions covering the entire face, as determined by a dermatologist. They have been diagnosed with eczema. Oil-absorbing solution was applied to the entire face of each subject (including the affected area in the case of AD). After collecting sebum using a film (5 x 8 cm, made of polypropylene, 3M), the oil remover The film was transferred to a vial and stored at -80°C for approximately one month until it was ready for RNA extraction. .

[0066] 2) RNA preparation and sequencing Cut the oil-absorbing film from 1) above to an appropriate size, and apply QIAzol Lysis RNA was extracted using Reagent (Qiagen) according to the provided protocol. Based on the extracted RNA, SuperScript VILO cDNA Synth Using the esis kit (Life Technologies Japan Co., Ltd.), at 42°C for 90 minutes. Reverse transcription was performed to synthesize cDNA. The primers for the reverse transcription reaction were those included in the kit. Random primers were used. From the obtained cDNA, multiplex PCR was performed. A library containing DNA derived from 20802 genes was prepared. SPCR is a type of Ion AmpliSeqTranscriptome Human Geometry. Using the ne Expression Kit (Life Technologies Japan Co., Ltd.) Then, [99℃, 2 minutes → (99℃, 15 seconds → 62℃, 16 minutes) × 20 cycles → 4℃, Ho The procedure was performed under the conditions of [ld]. The obtained PCR product was Ampure XP (Beckman Co.). After purification by Luther Co., Ltd., buffer reconstitution, primer sequence digestion, and adaptation are performed. Ligation, purification, and amplification were performed to prepare the library. - Load into Ion 540 Chip and Ion S5 / XL system (life Sequencing was performed using Technologies Japan Co., Ltd.

[0067] 3) Data Analysis i) Data used The RNA expression data (read count value) from the subject measured in 2) above The data was then corrected using the DESeq2 method. However, the expression data for all sample subjects was not used. Of these, 348 sample subjects had non-missing expression level data available for more than 90%. Only 6 genes were used for the following analysis. The analysis was corrected using the DESeq2 method. The normalized count value was used. ii) RNA expression analysis The SSL-derived RNA expression levels (normalized) of healthy individuals and AD as measured in i) above. Based on the count value, the corrected value of the p-value obtained by the likelihood ratio test in AD compared to healthy individuals. RNAs (differentially expressed genes) with an FDR of less than 0.25 were identified. As a result, 31 0 types of RNA decreased (DOWN), and 61 types of RNA increased (UP) (Table 1-1~1- 9).

[0068] [Table 1-1]

[0069] [Table 1-2]

[0070] [Table 1-3]

[0071] [Table 1-4]

[0072] [Table 1-5]

[0073] [Table 1-6]

[0074] [Table 1-7]

[0075] [Table 1-8]

[0076] [Table 1-9]

[0077] For the 371 genes shown in Tables 1-1 to 1-9, the public database STRI Biologic analysis using gene ontology (GO) enrichment with NG We investigated the al-process (BP). As a result, we found that its expression was reduced in AD patients. 144 BPs associated with the gene group were obtained, and they were involved in cell death, keratinization, and immune response (neutrophils, leukocytes). It has been shown that the term includes degranulation of myeloid cells, activation of myeloid cells, and lipid metabolism. (Table 2). In addition, 44 BPs associated with the gene group with increased expression were obtained, and exogenous antigens This included terms related to the immune response to (Tables 2-1 to 2-4). On the other hand, the above table Of the 371 genes shown in 1-1 to 1-9, 318 genes (each table is marked with an asterisk in bold) Regarding (displayed in text), there have been no reports to date suggesting a link to atopic dermatitis. Therefore, it was determined that it could serve as a novel marker for atopic dermatitis.

[0078]

Table 2-1

[0079]

Table 2-2

[0080]

Table 2-3

[0081]

Table 2-4

[0082] Example 2: Construction of a discrimination model using genes with high variable importance in random forest 1) Data used In the same way as in Example 1, data on the expression levels of SSL-derived RNA from subjects (read count values) were obtained and converted to RPM values corrected for differences in the total number of reads between samples. However, only the expression level data of 3486 genes that were not missing values in 90% or more of all samples were used for the following analysis. For constructing the machine learning model, to approximate the RPM values following the negative binomial distribution to the normal distribution, the base-2 logarithm value (Log2(RPM + 1) value) added with integer 1 was used. However, only the expression level data of 3486 genes that were not missing values in 90% or more of all samples were used for the following analysis. For constructing the machine learning model, to approximate the RPM values following the negative binomial distribution to the normal distribution, the base-2 logarithm value (Log2(RPM + 1) value) added with integer 1 was used. However, only the expression level data of 3486 genes that were not missing values in 90% or more of all samples were used for the following analysis. For constructing the machine learning model, to approximate the RPM values following the negative binomial distribution to the normal distribution, the base-2 logarithm value (Log2(RPM + 1) value) added with integer 1 was used. However, only the expression level data of 3486 genes that were not missing values in 90% or more of all samples were used for the following analysis. For constructing the machine learning model, to approximate the RPM values following the negative binomial distribution to the normal distribution, the base-2 logarithm value (Log2(RPM + 1) value) added with integer 1 was used. However, only the expression level data of 3486 genes that were not missing values in 90% or more of all samples were used for the following analysis. For constructing the machine learning model, to approximate the RPM values following the negative binomial distribution to the normal distribution, the base-2 logarithm value (Log2(RPM + 1) value) added with integer 1 was used. However, only the expression level data of 3486 genes that were not missing values in 90% or more of all samples were used for the following analysis. For constructing the machine learning model, to approximate the RPM values following the negative binomial distribution to the normal distribution, the base-2 logarithm value (Log2(RPM + 1) value) added with integer 1 was used.

[0083] 2) Selection of feature genes To select feature genes using the random forest algorithm, the Log2(RPM + 1) values of 3486 genes for which expression level data were obtained without missing values in 90% or more of all samples were used as explanatory variables, and healthy subjects (HL) and AD were used as target variables. The random forest algorithm in the "caret" package of the R language To select feature genes using the random forest algorithm, the Log2(RPM + 1) values of 3486 genes for which expression level data were obtained without missing values in 90% or more of all samples were used as explanatory variables, and healthy subjects (HL) and AD were used as target variables. The random forest algorithm in the "caret" package of the R language To select feature genes using the random forest algorithm, the Log2(RPM + 1) values of 3486 genes for which expression level data were obtained without missing values in 90% or more of all samples were used as explanatory variables, and healthy subjects (HL) and AD were used as target variables. The random forest algorithm in the "caret" package of the R language To select feature genes using the random forest algorithm, the Log2(RPM + 1) values of 3486 genes for which expression level data were obtained without missing values in 90% or more of all samples were used as explanatory variables, and healthy subjects (HL) and AD were used as target variables. The random forest algorithm in the "caret" package of the R language was specified as a method, and the optimal value of the number of variables (mtry value) used for constructing a single decision tree was tuned. Using the mtry value determined by tuning, the random forest algorithm was executed, and the top 100 genes of variable importance based on the Gini coefficient were calculated (Tables 3-1 to 3-3). These 100 genes or 92 genes (marked with * and shown in bold in each table) that have not been reported to be related to atopic dermatitis so far were selected as feature genes.

[0084]

Table 3-1

[0085]

Table 3-2

[0086]

Table 3-3

[0087] 3) Model construction The Log2(RPM+1) values of the above 100 genes or 92 genes were used as explanatory variables, and HL and AD were used as target variables. In the "caret" package of the R language, the random forest algorithm was specified as a method, and the optimal value of the number of variables (mtry value) used for constructing a single decision tree was tuned. Using the mtry value determined by tuning, the random forest algorithm was executed, and the out-of-bag error rate (OOB error r rate) was calculated. As a result, the OOB error rate of the model using 100 genes was 9.43%, and that of the model using 92 genes was 13.21%. there were.

[0088] Example 3: Construction of a discrimination model using differentially expressed genes 1) Data used Using the same method as in Example 1, data on the expression level of SSL-derived RNA from subjects (readcards) The Count value was obtained and converted to an RPM value corrected for differences in the total number of reads between samples. To build a machine learning model, we approximate the RPM values, which follow a negative binomial distribution, to a normal distribution using integers. The base-2 logarithm (log2(RPM+1) value) with 1 added was used.

[0089] 2) Selection of feature genes In Example 2, 371 individuals showed significantly altered expression in AD compared to healthy individuals (HL). The genes (Tables 1-1 to 1-9), or those of which have been reported to date to be related to atopic dermatitis. The 318 genes that have not been reported (marked with an asterisk and shown in bold in each table) are used as characteristic genes. I selected it.

[0090] 3) Model Construction The Log2(RPM+1) values ​​of the above 371 or 318 genes are used as explanatory variables. L and AD were used as the target variables. Randomization was performed using the "caret" package in the R language. Specify the forest algorithm as a method, and the variables used to construct a decision tree in one go. The optimal value of the number (mtry value) was tuned. Using the values, run the Random Forest algorithm and calculate the OOB error rate. The following was calculated. As a result, the model using 371 genes showed an OOB error ra The percentage of te was 26.42%, while in the model using 318 genes it was 30.19%.

[0091] Example 4: Construction of a discriminant model using feature genes extracted by the Boruta method 1) Usage data Using the same method as in Example 1, data on the expression level of SSL-derived RNA from subjects (readcards) The read count was obtained and converted to an RPM value corrected for the difference in the total number of reads between samples. Furthermore, non-missing expression level data was obtained for over 90% of all samples. Only 3486 genes were used for the following analysis. A negative binomial distribution was used to construct the machine learning model. To approximate the RPM values ​​to a normal distribution, we add an integer 1 to the base 2 logarithm (log2(R The PM+1 value was used.

[0092] 2) Selection of feature genes Expression level data without missing values ​​was obtained for more than 90% of all samples.34 The Log2(RPM+1) values ​​of 86 genes were used as explanatory variables, and healthy individuals (HL) and individuals with AD were used as objective variables. The number was used, and the algorithm from the "Boruta" package in the R language was executed. The number of rows was set to 1000, and nine genes with a p-value of less than 0.01 were identified (Table 4). Table 4 The nine genes shown, or those of which have been reported to be related to atopic dermatitis. Seven genes that were not present (marked with an asterisk and shown in bold in Table 4) were selected as characteristic genes.

[0093] [Table 4]

[0094] 3) Model Construction The Log2(RPM+1) values ​​of the above 9 genes or 7 genes are used as explanatory variables, and HL and AD The target variable was used. Random forests were used in the "caret" package of the R language. Specify the algorithm as the method, and the number of variables used to construct one decision tree (mt The optimal value of the ry value was tuned. The mtry value determined by tuning was used. Then, the Random Forest algorithm is executed and the OOB error rate is calculated. As a result, the OOB error rate in the model using 9 genes was 9.4 In a model using 3% and 7 genes, the figure was 15.09%.

Claims

1. A method for detecting infantile atopic dermatitis in a subject, comprising measuring the expression level of mRNA of at least one gene selected from a group of three genes consisting of IMPDH2, ERI1, and FBXW2 in surface lipids on the skin collected from the entire face of the subject.

2. The method described in claim 1, wherein the detection comprises comparing the measured expression levels with reference values ​​of the mRNA expression levels of each of the genes to assess the presence or absence of infantile atopic dermatitis.