Method for predicting risk of developing gastric cancer

WO2025058072A1PCT designated stage expired Publication Date: 2025-03-20CHIBA UNIV
View PDF 8 Cites 0 Cited by

Patent Information

Application Number
PCT/JP2024/032956
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-09-14
Filing Date
2024-09-13
Publication Date
2025-03-20

Smart Images

  • Figure JP2024032956_20032025_PF_FP_ABST
    Figure JP2024032956_20032025_PF_FP_ABST
Patent Text Reader

Abstract

The purpose of the present invention is to provide a method for predicting the risk of developing gastric cancer in a subject before the development of gastric cancer. In the present invention, the relationship between a DNA methylation level in a gastric mucosa biopsy sample collected from a patient before the development of gastric cancer and the development of gastric cancer is analyzed. As a result, a relationship between the DNA methylation in a specific gene and the development of gastric cancer is found. On the basis of this finding, a method for predicting the risk of developing gastric cancer and a system for determining the risk based on a DNA methylation level in a specific gene are provided.
Need to check novelty before this filing date? Find Prior Art

Description

Methods for predicting the risk of developing stomach cancer

[0001] A method for predicting the risk of developing gastric cancer based on the DNA methylation levels of specific genes detected in gastric biopsy samples is provided.

[0002] Gastric cancer (GC) is the fourth leading cause of cancer-related mortality worldwide. Environmental factors, such as infection with pathogens like Helicobacter pylori (H. pylori) and Epstein-Barr virus (EBV), as well as lifestyle factors like alcohol consumption and smoking, contribute to the development of gastric cancer. H. pylori, a gram-negative bacterium present in the stomachs of approximately half of the world's population, is a major risk factor for gastric cancer. Epstein-Barr virus (EBV) is a human herpesvirus acquired during childhood. It is latent in more than 95% of adults and contributes to gastric cancer in 7-16% of cases. H. pylori infection induces chronic inflammation in the gastric mucosa, which then manifests the latent EBV, leading to atrophic gastritis and intestinal metaplasia, which then progress to gastric cancer. Smoking and alcohol consumption are also factors that increase the risk of gastric cancer. Age and gender also play a role in the development of gastric cancer.

[0003] In areas with a high risk of gastric cancer, individuals infected with Helicobacter pylori often undergo regular endoscopic examinations, even if they are asymptomatic. However, the frequency of these examinations is currently not standardized, and the decision is left to the discretion of clinicians and patients. This approach results in frequent endoscopic examinations even in individuals with a low risk of developing primary gastric cancer, leading to increased patient burden and healthcare costs. Therefore, there is a need to optimize the frequency of endoscopic examinations for each patient by stratifying their risk of developing primary gastric cancer. Furthermore, implementing a personalized approach to endoscopic screening tailored to individual risk profiles may enable earlier detection of gastric cancer, reduce patient burden, and optimize the allocation of healthcare resources.

[0004] It has been reported that aberrant DNA methylation accumulates in specific genomic regions in gastric cancer biopsy samples, and that Epstein-Barr virus-positive gastric cancers exhibit particularly high levels of DNA methylation (Non-Patent Document 1: Matsusaka et al. Cancer Res. 2011;71:7187-97.). The specific genomic regions where aberrant DNA methylation accumulates are associated with genes, particularly promoter CpG islands of tumor suppressor genes, and the silencing of such genes is closely related to the development of cancer. Aberrant DNA methylation is associated with various cancer characteristics, such as altered expression of cancer driver and passenger genes, telomere shortening, and chromosomal alterations. Gastric cancer is classified into Epstein-Barr virus positive (EBV(+)), microsatellite instability-high (MSI(+)), chromosomal instability (CIN), and genomic instability (GS) subtypes. Gene methylation has been reported for each of these subtypes (Non-Patent Document 2: Cancer Genome Atlas Research Network. Nature 2014;513:202-9). Furthermore, aberrant DNA methylation in the mucosa of the remnant stomach of patients undergoing endoscopic resection for gastric cancer has been reported to predict the risk of developing secondary gastric cancer (Non-Patent Document 3: Asada K, et al. Gut. 2015;64:388-96; Non-Patent Document 4: Maeda M, et al. Gut. 2017;66:1721-3). Furthermore, it has been reported that hypermethylated genes were extracted from the mucosa surrounding gastric cancer in patients with H. pylori infection, and genes with low methylation in the mucosa of healthy individuals with H. pylori infection were selected from these to be used as new markers (Patent Document 1: JP 2018-57362 A, and Non-Patent Document 5: Nanjo S, et al., Gastric Cancer (2012) 15:382-388). However, since these reports did not involve observations of a long-term observation cohort, the term "healthy individuals" merely refers to individuals who had not developed gastric cancer at the time the mucosa was collected.To express it as a cancer risk marker, a comparison is essentially required between individuals who have actually developed gastric cancer, or more ideally, healthy individuals who will be found to develop gastric cancer in subsequent long-term follow-up, and healthy individuals who will not develop gastric cancer in subsequent long-term follow-up. In other words, it is necessary to ensure that the low-risk group of healthy individuals will not develop gastric cancer in subsequent long-term follow-up. However, such reports have only examined DNA methylation in the non-cancerous gastric mucosa of individuals who developed gastric cancer and in the gastric mucosa of healthy individuals who have not developed gastric cancer, and therefore have not yet reached the point of predicting future gastric cancer risk based on hypermethylation of selected genes before the onset of gastric cancer.

[0005] Japanese Patent Application Laid-Open No. 2018-57362

[0006] Matsusaka et al. Cancer Res. 2011;71:7187-97.Cancer Genome Atlas Research Network. Nature 2014;513:202-9.Asada K, et al. Gut.2015;64:388-96Maeda M, et al. Gut.2017;66:1721-3.Nanjo S, et al., Gastric Cancer (2012)15:382-388

[0007] The present invention aims to provide a method for predicting the risk of developing primary gastric cancer in a subject before the onset of gastric cancer.

[0008] The present inventors have focused on DNA methylation, which has been reported to accumulate in gastric cancer biopsy samples, and analyzed the relationship between DNA methylation levels in gastric mucosal biopsy samples collected from patients before the onset of gastric cancer and the onset of gastric cancer. As a result, they have clarified the association between DNA methylation in specific genes and the onset of gastric cancer, and have developed a method for predicting the risk of developing gastric cancer based on DNA methylation in specific genes after cohort analysis. Therefore, the present invention relates to the following: [1] A method for predicting the risk of developing gastric cancer in a subject, comprising: SLC6A3, GFPT2, LOC392232, CCDC37-AS1, LMX1B, ZNF790-AS1, ZSCAN23, SSTR2, FZD10, ZSCAN18, MIR3663HG, MO S, ZNF660, NAALAD2, HAGLR, CFAP100, KCNA7, ZNF549, NCAN, HS3ST2, SRD5A2, PCDH8, GHSR, WTH3DI, KCNK10, ZNF 880, CCDC67, LVRN, COMP, NEUROG1, ZNF578, LRIF1, CADM3, EID3, GRP, KCNA1, SLC2A14, ST6GAL2, DMRTA2, MYOD1 , GALR1, ISLR2, SIAH3, IRX4, INSC, DBX1, IRX6, AARD, FAM135B, HBQ1, ADCY8, LHCGR, SLITRK5, DLX5, GFRA1, MIR1 24-3, PABPC4L, TBX5-AS1, MIR9-3HG, DRD5, PCDH10, TOX2, KNSTRN, POU4F3, TMEFF2, CHRM2, ZNF582, CCDC36, CA DM1, NXPH2, MKX, GABRA5, TLX1, ZNF418, NKX2-2, COL4A2, OPRM1, TRPA1, FAR2P2, SLC27A6, CIDEA, CXCL12, PTPR T, PXDN, HIST1H1A, ZFHX4, GSX1, ADCYAP1, SYT6, ZFPM2, DMRT3, KHDRBS2, VSTM2B, LOC100287846, ELMO1, NPBWR 1, DKK2, ERVMER34-1, NKX1-1, C11orf85, CABP7, HELT, LIN7A, ALDH1L1-AS2, NKX2-1, VAX1, SYT9, SUSD5, CCDC8,CRHR2、C8orf34-AS1、CTNND2、LINC01535、FOXI2、PCSK2、LOC101928134、NT MT1、BOLL、RXFP3、HAPLN4、PNPLA5、ZNF492、WT1-AS、ZNF529、SERTM1、ERBB4、 MDGA2、DPY19L2、IGFBP7-AS1、LOC101928796、USP44、SLCO4C1、BHLHE23、PY GO1、PLXNA4、SLC35F4、LBX2、LINC01248、CACNA1B、RNPS1、CPLX2、DCLK1、NKX 2-1-AS1、RSPO3、HOXD1、PLPPR4、C1QL2、CYTL1、OPRK1、ADAMTS18、RAX、PCDH GB7、TMEM108、KAZN、TDRP、NRXN3、FENDRR、ZIC5、EDNRB、RIMS2、LINC00485、E MX2、EVC2、PEAR1、T、ZNF702P、HOXD9、CNTNAP5、CFAP161、ZNF334、ACTA1、CY YR1、COL1A2、DYDC2、FBXL7、GDNF、ELAVL2、HS3ST3A1、LOC283731、ERICH3、PC DHGA11、EFS、RALYL、EVX1、PCDHGA12、DDX25、DNAJC19、GPR139、SST、NOTE、S YN3、PCDHAC2、DOC2A、PREX2、TSPYL5、ZNF677、CBLN4、GRM2、NEUROD1、MCHR2、 C11orf87, LAYN, UNCX, HECW1, PTH2R, DYDC1, SPAG6, KCNK9, HMCN2, SNAP25 KDR、KIAA1024、ARHGEF10、HMGCLL1、SYT1、C3orf14、BARHL2、GLRA3、NWD2、SN AP91、KCNH7、ZNF582-AS1、ZNF300P1、SLC32A1、PDE10A、RGS17、CCNA1、PCSK 1、B3GAT1、HOXA9、HOXD10、DNAH8、FAM184B、HMX2、B3GAT2、NXPH1、TRAM1L1、B VES-AS1、DSCAM、GBX2、NTSR2、GRIN2A、EFTUD1P1、COL25A1、TUSC3、ACSS3、DS C3、NPR3、NEU1、ZIC4、SEMA6D、LINC01197、POU4F2、GUCY1B3、OLIG2、GALNT9、LOC101928174, NTM, FAM20A, GRIA4, CPXM2, HOXA10, LEPR, PTPN5, PRMT8, CDO1, LHFPL4, SHISA9, ALDH1L1, CPNE8, FAM19A4, C8orf88, CALN1, PDPN, SORCS1, CHL1, FSTL1, LYPD5, ADRA1D, C2orf40, SP9, ALK, PDZRN4, ZSCAN1, PRLHR, NRSN1, FBN2, TLX3, ZSWIM2, GPR146, ZIK1, TRHDE-AS1, HOXD11, GRM7, ZNF36 5、NEFL、C10orf53、ALDH1A2、FGF10-AS1、AJAP1、RELN、KCNIP1、MIR2682、TMEM132E、NECAB1、ADCY5、GJD2、EFCC1、GIPC3、LINC01231、PLPPR3、RYR3、CTD -2297D10.2、LINC00461、TBX18、CWH43、SELV、CCDC169、NID1、KCND2、ZNF835、DPP10、LOC284395、SSTR4、TRPC4、RNF212、GRIA2、EPDR1、CSMD1、RAPGEF4、 REC8、HS6ST1、AGTR1、WBSCR17、C14orf39、LRRTM1、THSD7B、LOC100289333、MIR4281、TMEM196、ROBO3、TCF15、GALNT13、GUCY1A3、ADAM33、PTH2、CDH7、E PHA7、WNT16、NKAIN4、NGF、ZNF132、PHACTR3、MEIOB、CDH18、SLC18A3、ZNF471、ZFP42、RASGRF1、GALR2、FGF3、DOCK10、TRHDE、SYT3、VWC2、SOX21、DBX2、FO XG1、ZNF667-AS1、ADGRV1、PCDHGB6、EPHA5-AS1、USP6、ZNF382、ST8SIA2、UTF1、CDH4、CNTN4、KCNK2、NRG1、PHYHIPL、THCAT155、MYO3A、FAM19A5、PHOX2A GABRG3、IGF2-AS、TCERG1L、ZNF793、COL4A1、SORCS3、11-Mar、PRR35、PLXDC2、TDRD10、LRRK2、ZFP28、NRIP3、ACAN、NLGN1、THY1、AFG3L1P、SFTA3、RAB12、The method comprises: detecting the DNA methylation level of at least one gene selected from the group consisting of TWIST2, FREM2, EBF2, FEZF2, CACNA1E, NTRK3, EMX2OS, HAND1, SNTG1, WNT7A, CCDC140, FGF14, PGR, SATB2-AS1, and GAD2; and predicting the risk of developing gastric cancer by comparing the detected DNA methylation level with at least one threshold value of the methylation level corresponding to the risk of developing gastric cancer. [2] The method according to item 1, wherein the DNA methylation level of the gene is the methylation level in a gene promoter CpG island. [3] The method according to item 1, wherein the gene is a candidate gastric cancer risk marker gene that has been analyzed in a cohort. [4] The method according to item 1, wherein the gastric biopsy sample is from non-cancerous tissue. [5] The method according to item 1, wherein a high methylation level of the gene in normal tissue indicates a high risk of developing gastric cancer in the future, while a low methylation level indicates a low risk of developing gastric cancer in the future. [6] The cohort analyzed candidate gastric cancer risk marker genes are the following: NKX1-1, PABPC4L, SLC32A1, SYT6, TRPC4, WBSCR17, B3GAT1, ST8SIA2, RELN, DRD5, CPNE8, CACNA1B, DPP10, HOXD9, GRIA2, RXFP3, MKX, NKX2-2, LOC101928174, PREX2, GALNT13, CTD-2297D10.2, PTH2R, LYPD5, ADAMTS18, NXPH2, PCSK1, GSX1, GPR146, ZNF578, GUCY1B3, ZNF471, FAM135B, GBX2, PL PPR3, CFAP100, DBX1, NRG1, HMGCLL1, DDX25, ADRA1D, ADCY8, ACTA1, KAZN, SFTA3, ZFP42, NRIP3, NWD2, CADM3, ADCYAP1, MCHR2, HS3ST2, CYYR1, EVX1, TCERG1L, ADAM3 3, NLGN1, EID3, CDH4, EFCC1, ISLR2, PHYHIPL, SLITRK5, FGF3, TMEFF2, SEMA6D, HOXD10, PLXDC2, NKX2-1-AS1, ELAVL2, SYT9, AJAP1, ZSWIM2, SNAP91, GABRG3, PLPPR4,ZNF365, C14orf39, ERBB4, LMX1B, IRX4, FBN2, ACAN, EVC2, TBX5-AS1, DYDC2, HS3ST3A1, IRX6, SLC27A6, PCDH8, CNTN4, BARHL2, SNAP25, RAPGEF4, HOXD11, USP6, LOC392232, HAGLR, PRLHR, TUSC3, SERTM1, VSTM2B, KHDRBS2, FOXI2, TDRP, TDRD10, FGF14, NTM, ST6GAL2, PRMT8, CCDC36, KDR, PGR, KCNIP1, ZNF702P , GABRA5, RNPS1, NTRK3, BOLL, EDNRB, LEPR, RALYL, DSCAM, ROBO3, IGF2-AS, HOXD1, SSTR4, GALR1, POU4F2, MEIOB, INSC, ZNF382, PCSK2, DCLK1, UNCX, LINC01248, DYDC1, OPRM1, COL4A2, LINC01197, ZNF880, CPLX2, GALNT9, GRM7, TRAM1L1, GDNF, SORCS1, ZSCAN1, and EFS. [7] The method according to Item 3, wherein the candidate gastric cancer risk marker gene after cohort analysis is at least one gene selected from the group consisting of: NKX1-1, PABPC4L, SLC32A1, SYT6, TRPC4, WBSCR17, B3GAT1, ST8SIA2, RELN, DRD5, CPNE8, CACNA1B, DPP10, HOXD9, GRIA2, RXFP3, MKX, NKX2 -2, LOC101928174, PREX2, GALNT13, CTD-2297D10.2, PTH2R, LYPD5, ADAMTS18, NXPH2, PCSK1, GSX1, GPR146, ZNF578, GUCY1B3, FAM135B, G BX2, PLPPR3, DBX1, NRG1, HMGCLL1, DDX25, ADRA1D, ADCY8, ACTA1, KAZN, SFTA3, ZFP42, NRIP3, NWD2, CADM3, HS3ST2, CYYR1, EVX1, TCERG1L , ADAM33, NLGN1, EID3, CDH4, EFCC1, ISLR2, PHYHIPL, SLITRK5, FGF3, TMEFF2, NKX2-1-AS1, ELAVL2, SYT9, AJAP1, ZSWIM2, SNAP91, GABRG3,PLPPR4, ZNF365, C14orf39, ERBB4, LMX1B, IRX4, ACAN, EVC2, TBX5-AS1, DYDC2, SLC27A6, PCDH8, CNTN4, LOC392232 , HAGLR, PRLHR, TUSC3, SERTM1, KHDRBS2, FOXI2, TDRP, FGF14, NTM, ST6GAL2, PRMT8, CCDC36, KDR, PGR, KCNIP1, ZNF7 [8] The method according to Item 3, wherein the candidate gastric cancer risk marker gene after cohort analysis is at least one gene selected from the group consisting of 02P, BOLL, EDNRB, RALYL, FOXG1, SLC2A14, PCDHGB7, GRM2, CTNND2, LINC00485, ERVMER34-1, CCDC37-AS1, FAM19A4, TBX18, MIR9-3HG, LOC100289333, LINC00461, TRHDE, FZD10, and FAM19A5. NKX1-1, PABPC4L, SLC32A1, SYT6, TRPC4, WBSCR17, B3GAT1, ST8SIA2, RELN, DRD5, CPNE8, CACNA1B, DPP10, HOXD9, GRIA2, RXFP3, MKX, NKX2-2, LOC101928174, PREX2, GALNT13, CT D-2297D10.2, PTH2R, LYPD5, ADAMTS18, NXPH2, PCSK1, GSX1, GPR146, ZNF578, GUCY1B3, FAM135B, GBX2, PLPPR3, DBX1, NRG1, HMGCLL1, DDX25, ADRA1D, ADCY8, ACTA1, KAZN, SFTA3, ZFP42, NRIP3, NWD2, CADM3, HS3ST2, CYYR1, EVX1, TCERG1L, ADAM33, NLGN1, EID3, EFCC1, SLITRK5, FGF3, TMEFF2, NKX2-1-AS1, ELAVL2, SYT9, AJAP1, ZSWIM2, SNAP91, GABRG3, PLP PR4, ZNF365, C14orf39, ERBB4, LMX1B, IRX4, ACAN, TBX5-AS1, DYDC2, SLC27A6, HAGLR, PRLHR, SERTM1, KHDRBS2, TDRP, FGF14, NTM, PRMT8, CCDC36, KDR, ZNF702P, EDNRB, PCDHGB7,LINC00485, ERVMER34-1, CCDC37-AS1, TBX18, MIR9-3HG, LOC100289333, LINC00461, FZD10, CFAP100, MCHR2, PLXDC2, HS3ST3A1 , BARHL2, SNAP25, RAPGEF4, NTRK3, DSCAM, ROBO3, IGF2-AS, GALR1, INSC, ZNF382, PCSK2, DCLK1, UNCX, DYDC1, OPRM1, COL4A2, GR The method according to Item 3, wherein the gene is at least one selected from the group consisting of M7, SORCS1, SLC18A3, SPAG6, PCDHGB6, ADCY5, ZNF529, THSD7B, COL25A1, PCDHGA11, NPBWR1, VAX1, WTH3DI, ZSCAN18, PCDHAC2, DMRTA2, KCNH7, MOS, TLX3, LAYN, SORCS3, NKAIN4, KIAA1024, CCDC140, NRXN3, and EMX2OS. [9] The cohort analyzed candidate gastric cancer risk marker genes are as follows: NKX1-1, PABPC4L, SLC32A1, SYT6, TRPC4, WBSCR17, B3GAT1, ST8SIA2, RELN, DRD5, CPNE8, CACNA1B, DPP10, HOXD9, GRIA2, RXFP3, MKX, NKX2-2, LOC101928174, PREX2, GALNT13, CTD-2297D10.2, PTH2R, LYPD5, ADAMTS18, NXPH2, PCSK1, GSX1, GPR146, ZNF578, GUCY1B3, FAM135B, GBX2, PLPPR3, DBX1, NRG1, and HMGCLL1. , DDX25, ADRA1D, ADCY8, ACTA1, KAZN, SFTA3, ZFP42, NRIP3, NWD2, CADM3, HS3ST2, CYYR1, EVX1, TCERG1L, ADAM33, NLGN1, EID3, EFCC1, SLITRK5, FGF3, TMEFF2, NKX2-1-AS1, EL AVL2, SYT9, AJAP1, ZSWIM2, SNAP91, GABRG3, PLPPR4, ZNF365, C14orf39, ERBB4, LMX1B, IRX4, ACAN, TBX5-AS1, DYDC2, SLC27A6, HAGLR, PRLHR, SERTM1, KHDRBS2, TDRP, FGF14,

[10] The method according to Item 3, wherein the candidate gastric cancer risk marker gene after cohort analysis is at least one gene selected from the group consisting of NTM, PRMT8, CCDC36, KDR, ZNF702P, and EDNRB.

[11] The method according to Item 3, wherein the candidate gastric cancer risk marker gene after cohort analysis is at least one gene selected from the group consisting of NKX1-1, NKX2-2, DRD5, PABPC4L, ADAMTS18, FAM135B, ZNF578, PTH2R, SLC32A1, HS3ST2, B3GAT1, SLITRK5, SYT6, RXFP3, KHDRBS2, PCSK1, COL4A2, DBX1, EDNRB, ZNF365, HOXD9, MKX, GABRG3, 4. The method of item 3, wherein the gene is at least one selected from the group consisting of ACTA1, SERTM1, FGF14, TRPC4, ZSWIM2, FZD10, PCDHAC2, CACNA1B, PLPPR4, ADRA1D, RELN, GRIA2, DPP10, HMGCLL1, LOC101928174, TBX18, PREX2, CNTN4, CCDC36, WBSCR17, ADCY8, and C14orf39.

[11] The method of Aspect 3, wherein the cohort analyzed candidate gastric cancer risk marker gene is at least one gene selected from the group consisting of PABPC4L, DRD5, NKX1-1, NKX2-2, HS3ST2, SLITRK5, ZNF578, RXFP3, ADAMTS18, SYT6, B3GAT1, FAM135B, PTH2R, ZNF365, ACTA1, PCDHAC2, DBX1, MKX, FZD10, TRPC4, SLC32A1, CACNA1B, COL4A2, PCSK1, KHDRBS2, LOC101928174, DPP10, RELN, CCDC36, WBSCR17, PREX2, PLPPR4, and CNTN4.

[12] The cohort analyzed candidate gastric cancer risk marker genes are any of the following: NKX1-1, SLC32A1, CACNA1B, RXFP3, PABPC4L, PTH2R, B3GAT1, FAM135B, SYT6, ZNF578, ADAMTS18, ADRA1D, RELN, DBX1, DRD5, LOC101928174, DPP10, HS3ST2, NKX2-2, PLPPR4, KHDRBS2, PCSK1, SLITRK5, ZNF365, WBSCR17, MKX, CTD-2297D10.2, PREX2, NXPH2,ADCY8, NRG1, ZFP42, HOXD9, TRPC4, CNTN4, LYPD5, CYYR1, GABRG3, PLPPR3, TCERG1L, C14orf39, ACTA1, ST6GAL2, KAZN, HMGCLL1, SERTM1, ACAN, ZSWIM2, EDN RB, SORCS1, CCDC36, COL4A2, FGF14, MIR9-3HG, FZD10, PCDHGB6, WTH3DI, NKAIN4, TBX18, FOXG1, NKX2-1-AS1, EID3, CADM3, TMEFF2, ZNF702P, PCDHAC2, TRHD 4. The method of claim 3, wherein the gene is at least one selected from the group consisting of E, MOS, LOC100289333, GRIA2, GALNT13, GBX2, ST8SIA2, DDX25, CDH4, AJAP1, LMX1B, FOXI2, PRLHR, SFTA3, TBX5-AS1, HAGLR, ERBB4, UNCX, SLC27A6, EFS, NLGN1, GALR1, ZSCAN1, IRX4, VAX1, FAM19A4, SOX21, COL25A1, PHOX2A, CNTNAP5, PXDN, C8orf34-AS1, WNT16, and HBQ1.

[13] The method according to Item 3, wherein the cohort analyzed candidate gastric cancer risk marker gene is at least one gene selected from the group consisting of NKX1-1, PABPC4L, DRD5, RXFP3, PTH2R, ZNF578, ADAMTS18, HS3ST2, NKX2-2, SLITRK5, ZNF365, SLC32A1, CACNA1B, WBSCR17, B3GAT1, FAM135B, SYT6, PREX2, ADRA1D, CNTN4, PLPPR4, and CCDC36.

[14] The method according to any one of Items 1 to 13, wherein the DNA methylation level is determined using a technique selected from the group consisting of a DNA methylation array, next-generation sequencing, and pyrosequencing.

[15] A method for predicting a risk of developing gastric cancer, comprising measuring a DNA methylation level of at least one gene selected from the group consisting of NKX1-1, TMEM108, PTPN5, DBX1, KAZN, RYR2, CDH11, and KDR in a gastric biopsy sample obtained from the subject,and comparing the detected DNA methylation level with at least one threshold methylation level corresponding to the risk of developing gastric cancer to predict the risk of developing gastric cancer.

[16] A system for determining the risk of developing gastric cancer in a subject, comprising: detecting by pyrosequencing the DNA methylation level; and comparing the detected DNA methylation level with at least one threshold methylation level corresponding to the risk of developing gastric cancer to predict the risk of developing gastric cancer. G1, ZNF578, LRIF1, CADM3, EID3, GRP, KCNA1, SLC2A14, ST6GAL2, DMRTA2, MYOD1, GALR1, ISLR2, SIAH3, IRX4, ISC, DBX1, IRX6, AARD, FAM135B, HBQ1, ADCY8, LHCGR, SLITRK5, DLX5, GFRA1, MIR124-3, PABPC4L, TBX5-AS1, MIR9-3HG, DRD5, PCDH10, TOX2, KNST RN, POU4F3, TMEFF2, CHRM2, ZNF582, CCDC36, CADM1, NXPH2, MKX, GABRA5, TLX1, ZNF418, NKX2-2, COL4A2, OPRM1, TRPA1, FAR2P2 , SLC27A6, CIDEA, CXCL12, PTPRT, PXDN, HIST1H1A, ZFHX4, GSX1, ADCYAP1, SYT6, ZFPM2, DMRT3, KHDRBS2, VSTM2B, LOC100287846 , ELMO1, NPBWR1, DKK2, ERVMER34-1, NKX1-1, C11orf85, CABP7, HELT, LIN7A, ALDH1L1-AS2, NKX2-1, VAX1, SYT9, SUSD5, CCDC8, CRHR2, C8orf34-AS1, CTNND2, LINC01535, FOXI2, PCSK2, LOC101928134, NTMT1, BOLL, RXFP3, HAPLN4, PNPLA5, ZNF492, WT1-AS,<h2 style=";text-align:left;direction:ltr">ZNF529、SERTM1、ERBB4、MDGA2、DPY19L2、IGFBP7-AS1、LOC101928796、USP4 4、SLCO4C1、BHLHE23、PYGO1、PLXNA4、SLC35F4、LBX2、LINC01248、CACNA1B、 RNPS1、CPLX2、DCLK1、NKX2-1-AS1、RSPO3、HOXD1、PLPPR4、C1QL2、CYTL1、OP RK1、ADAMTS18、RAX、PCDHGB7、TMEM108、KAZN、TDRP、NRXN3、FENDRR、ZIC5、ED NRB、RIMS2、LINC00485、EMX2、EVC2、PEAR1、T、ZNF702P、HOXD9、CNTNAP5、CF AP161、ZNF334、ACTA1、CYYR1、COL1A2、DYDC2、FBXL7、GDNF、ELAVL2、HS3ST3 A1、LOC283731、ERICH3、PCDHGA11、EFS、RALYL、EVX1、PCDHGA12、DDX25、DNA JC19、GPR139、SST、NOTO、SYN3、PCDHAC2、DOC2A、PREX2、TSPYL5、ZNF677、CBL N4、GRM2、NEUROD1、MCHR2、C11orf87、LAYN、UNCX、HECW1、PTH2R、DYDC1、SPA G6、KCNK9、HMCN2、SNAP25、KDR、KIAA1024、ARHGEF10、HMGCLL1、SYT1、C3orf 14、BARHL2、GLRA3、NWD2、SNAP91、KCNH7、ZNF582-AS1、ZNF300P1、SLC32A1、 PDE10A、RGS17、CCNA1、PCSK1、B3GAT1、HOXA9、HOXD10、DNAH8、FAM184B、HMX2 、B3GAT2、NXPH1、TRAM1L1、BVES-AS1、DSCAM、GBX2、NTSR2、GRIN2A、EFTUD1P 1、COL25A1、TUSC3、ACSS3、DSC3、NPR3、NEU1、ZIC4、SEMA6D、LINC01197、POU4 F2, GUCY1B3, OLIG2, GALNT9, LOC101928174, NTM, FAM20A, GRIA4, CPXM2, HOXA10, LEPR, PTPN5, PRMT8, CDO1, LHFPL4, SHISA9, ALDH1L1, CPNE8, FAM19A4C8orf88, CALN1, PDPN, SORCS1, CHL1, FSTL1, LYPD5, ADRA1D, C2orf40, SP9, ALK, PDZRN4, ZSCAN1, PRLHR, NRSN1, FBN2, TLX3, ZSWIM2, GPR146, ZIK1, TRHDE-AS1, HOXD11, GRM7, ZNF365, NEFL, C10orf53, ALDH1A2, FGF10-AS1, AJAP1, RELN, KCNIP1, MIR2682, TMEM132E, NECAB1, ADCY5, GJD2, EFCC1, GIPC3 、LINC01231、PLPPR3、RYR3、CTD-2297D10.2、LINC00461、TBX18、CWH43、SELV、CCDC169、NID1、KCND2、ZNF835、DPP10、LOC284395、SSTR4、TRPC4、RNF21 2、GRIA2、EPDR1、CSMD1、RAPGEF4、REC8、HS6ST1、AGTR1、WBSCR17、C14orf39、LRRTM1、THSD7B、LOC100289333、MIR4281、TMEM196、ROBO3、TCF15、GALNT1 3、GUCY1A3、ADAM33、PTH2、CDH7、EPHA7、WNT16、NKAIN4、NGF、ZNF132、PHACTR3、MEIOB、CDH18、SLC18A3、ZNF471、ZFP42、RASGRF1、GALR2、FGF3、DOCK10 、TRHDE、SYT3、VWC2、SOX21、DBX2、FOXG1、ZNF667-AS1、ADGRV1、PCDHGB6、EPHA5-AS1、USP6、ZNF382、ST8SIA2、UTF1、CDH4、CNTN4、KCNK2、NRG1、PHYHIPL 、THCAT155、MYO3A、FAM19A5、PHOX2A、GABRG3、IGF2-AS、TCERG1L、ZNF793、COL4A1、SORCS3、11-Mar、PRR35、PLXDC2、TDRD10、LRRK2、ZFP28、NRIP3、ACA N、NLGN1、THY1、AFG3L1P、SFTA3、RAB12、TWIST2、FREM2、EBF2、FEZF2、CACNA1E、NTRK3、EMX2OS、HAND1、SNTG1、WNT7A、CCDC140、FGF14、PGR、SATB2-AS1、and GAD2, and a gastric cancer risk determination device, wherein the gastric cancer risk determination device comprises: an input unit for inputting the DNA methylation level obtained from the detection device; a memory unit for storing at least one threshold value of the methylation level corresponding to the gastric cancer risk; a processing unit for comparing the DNA methylation level input from the input unit with the at least one threshold value retrieved from the memory unit to determine the gastric cancer risk of the subject; and an output unit for outputting the determined gastric cancer risk of the subject.

[17] The system according to Item 16, wherein the memory unit stores control DNA methylation levels according to at least one risk factor selected from the group consisting of smoking habit, drinking habit, and age.

[18] A system for determining the gastric cancer risk of a subject, comprising: a gastric biopsy sample obtained from the subject, SLC6A3, GFPT2, LOC392232, CCDC37-AS1, LMX1B, ZNF790-AS1, ZSCAN23, SSTR2, FZD10, ZSCAN18, MIR3663HG, MOS, ZNF660, NAALAD2, HAGLR, CFAP100, KCNA7, ZNF549, NCAN, HS3ST2 , SRD5A2, PCDH8, GHSR, WTH3DI, KCNK10, ZNF880, CCDC67, LVRN, COMP, NEUROG1, ZNF578, LRIF1, CADM3, EID3, GRP, KCNA1, SLC2A14, ST6GAL2, DMRTA2, MYOD1, GALR1, ISLR2, SIAH3, I RX4, INSC, DBX1, IRX6, AARD, FAM135B, HBQ1, ADCY8, LHCGR, SLITRK5, DLX5, GFRA1, MIR124-3, PABPC4L, TBX5-AS1, MIR9-3HG, DRD5, PCDH10, TOX2, KNSTRN, POU4F3, TMEFF2, CHRM2, ZNF582, CCDC36, CADM1, NXPH2, MKX, GABRA5, TLX1, ZNF418, NKX2-2, COL4A2, OPRM1, TRPA1, FAR2P2, SLC27A6, CIDEA, CXCL12, PTPRT, PXDN, HIST1H1A, ZFHX4, GSX1, ADCYAP1, SYT6,ZFPM2、DMRT3、KHDRBS2、VSTM2B、LOC100287846、ELMO1、NPBWR1、DKK2、ERVMER34-1、NKX1-1、C11orf85、CABP7、HELT、LIN7A、ALDH1L1-AS2、NKX2-1、VAX1、SYT9、SUSD5、CCDC8、CRHR2、C8orf34-AS1、CTNND2、LINC01535、FOXI2、PCSK2、LOC101928134、NTMT1、BOLL、RXFP3、HAPLN4、PNPLA5、ZNF492、WT1-AS、ZNF529、SERTM1、ERBB4、MDGA2、DPY19L2、IGFBP7-AS1、LOC101928796、USP44、SLCO4C1、BHLHE23、PYGO1、PLXNA4、SLC35F4、LBX2、LINC01248、CACNA1B、RNPS1、CPLX2、DCLK1、NKX2-1-AS1、RSPO3、HOXD1、PLPPR4、C1QL2、CYTL1、OPRK1、ADAMTS18、RAX、PCDHGB7、TMEM108、KAZN、TDRP、NRXN3、FENDRR、ZIC5、EDNRB、RIMS2、LINC00485、EMX2、EVC2、PEAR1、T、ZNF702P、HOXD9、CNTNAP5、CFAP161、ZNF334、ACTA1、CYYR1、COL1A2、DYDC2、FBXL7、GDNF、ELAVL2、HS3ST3A1、LOC283731、ERICH3、PCDHGA11、EFS、RALYL、EVX1、PCDHGA12、DDX25、DNAJC19、GPR139、SST、NOTO、SYN3、PCDHAC2、DOC2A、PREX2、TSPYL5、ZNF677、CBLN4、GRM2、NEUROD1、MCHR2、C11orf87、LAYN、UNCX、HECW1、PTH2R、DYDC1、SPAG6、KCNK9、HMCN2、SNAP25、KDR、KIAA1024、ARHGEF10、HMGCLL1、SYT1、C3orf14、BARHL2、GLRA3、NWD2、SNAP91、KCNH7、ZNF582-AS1、ZNF300P1、SLC32A1、PDE10A、RGS17、CCNA1、PCSK1、B3GAT1、HOXA9、HOXD10、DNAH8、FAM184B、HMX2、B3GAT2、NXPH1、TRAM1L1、BVES-AS1、DSCAM、GBX2、NTSR2、GRIN2A、EFTUD1P1 、COL25A1、TUSC3、ACSS3、DSC3、NPR3、NEU1、ZIC4、SEMA6D、LINC01197、POU4 F2、GUCY1B3、OLIG2、GALNT9、LOC101928174、NTM、FAM20A、GRIA4、CPXM2、HO XA10, PTPN5, PRMT8, CDO1, LHFPL4, SHISA9, ALDH1L1, CPNE8, FAM19A4 C8orf88、CALN1、PDPN、SORCS1、CHL1、FSTL1、LYPD5、ADRA1D、C2orf40、SP9、 ALK、PDZRN4、ZSCAN1、PRLHR、NRSN1、FBN2、TLX3、ZSWIM2、GPR146、ZIK1、TRHD E-AS1、HOXD11、GRM7、ZNF365、NEFL、C10orf53、ALDH1A2、FGF10-AS1、AJAP1 、RELN、KCNIP1、MIR2682、TMEM132E、NECAB1、ADCY5、GJD2、EFCC1、GIPC3、LIN C01231、PLPPR3、RYR3、CTD-2297D10.2、LINC00461、TBX18、CWH43、SELV、CC DC169、NID1、KCND2、ZNF835、DPP10、LOC284395、SSTR4、TRPC4、RNF212、GRIA 2、EPDR1、CSMD1、RAPGEF4、REC8、HS6ST1、AGTR1、WBSCR17、C14orf39、LRRTM 1、THSD7B、LOC100289333、MIR4281、TMEM196、ROBO3、TCF15、GALNT13、GUCY1 A3、ADAM33、PTH2、CDH7、EPHA7、WNT16、NKAIN4、NGF、ZNF132、PHACTR3、MAY B、CDH18、SLC18A3、ZNF471、ZFP42、RASGRF1、GALR2、FGF3、DOCK10、TRHDE、SY T3、VWC2、SOX21、DBX2、FOXG1、ZNF667-AS1、ADGRV1、PCDHGB6、EPHA5-AS1、U SP6、ZNF382、ST8SIA2、UTF1、CDH4、CNTN4、KCNK2、NRG1、PHYHIPL、THCAT155、1. A method for predicting the risk of developing gastric cancer comprising: detecting a DNA methylation level of at least one gene selected from the group consisting of MYO3A, FAM19A5, PHOX2A, GABRG3, IGF2-AS, TCERG1L, ZNF793, COL4A1, SORCS3, 11-Mar, PRR35, PLXDC2, TDRD10, LRRK2, ZFP28, NRIP3, ACAN, NLGN1, THY1, AFG3L1P, SFTA3, RAB12, TWIST2, FREM2, EBF2, FEZF2, CACNA1E, NTRK3, EMX2OS, HAND1, SNTG1, WNT7A, CCDC140, FGF14, PGR, SATB2-AS1, and GAD2; and determining a risk of developing gastric cancer, wherein the determining device for predicting the risk of developing gastric cancer comprises: an input unit for inputting the DNA methylation level obtained from the detecting device;

[19] The system according to Item 18, wherein the learning unit further includes, as the training data, information on at least one risk factor selected from the group consisting of smoking habit, drinking habit, and age, and outputs the gastric cancer risk when input information including the DNA methylation level of a specific gene and information on the at least one risk factor is input.

[0009] In the present invention, the DNA methylation levels of specific genes identified accurately reflect the risk of primary gastric cancer and serve as a useful quantitative marker for estimating the risk of primary gastric cancer. This makes it possible to predict the risk of developing gastric cancer based on the accumulation of DNA methylation in gastric cancer biopsy samples. The accumulation of DNA methylation levels in the gastric mucosa is observed several years before the onset of primary gastric cancer, making it possible to predict the risk of primary gastric cancer and the time to onset. Furthermore, the DNA methylation level in the gastric mucosa is an independent risk factor for primary gastric cancer, and preventative measures can be taken against the increased risk of developing gastric cancer due to Helicobacter pylori infection by making modifiable lifestyle choices in response to the accumulation of DNA methylation levels.

[0010] Figure 1 illustrates the screening cohorts (a, c) and GCEP cohorts (b, d) used in the cohort study. Figure 1 illustrates the screening cohorts (a, c) and GCEP cohorts (b, d) used in the cohort study. Figure 2 shows the DNA methylation epigenotypes in gastric cancer obtained by unsupervised two-way hierarchical clustering of gastric cancer and non-gastric cancer non-H. pylori mucosa. This identified three methylation subtypes in gastric cancer (EBV-CIMP, HME, and LME). Figure 3 shows the results of analyzing DNA methylation epigenotypes in gastric cancer and identifying DNA markers classified as EBV-CIMP and HME. We identified 271 EBV markers methylated in EBV GN (EBV-CIMP), 158 HME markers methylated in EBV GN and HME, and 156 common markers methylated in all gastric cancers but unmethylated in non-gastric cancer, non-H. pylori mucosa. Figure 4 shows the results of a one-way hierarchical clustering analysis of EBV markers, HME markers, and common markers using public data (TCGA) to verify their validity. Figure 5 shows the results of gene ontology analysis of the markers. Figure 6 shows the heatmap results of DNA methylation analysis using gastric cancer tissue and its surrounding mucosa (a). Regarding the gastric mucosa surrounding gastric cancer, biopsies derived from EBV GN, HME, and LME showed higher levels of methylation than non-gastric cancer, non-H. pylori mucosa biopsies (b). Figure 7 shows a heat map of DNA methylation analysis of 156 common marker genes in gastric mucosal biopsies collected during health checkups for subjects screened for future gastric cancer (a). The β values ​​of biopsies from future gastric cancer patients were significantly higher than those of non-H. pylori-infected and non-H. pylori-infected subjects (b). The analysis of methylation data from gastric biopsies collected for GCEP (c) also shows the correlation between β values ​​of biopsies from future gastric cancer patients and the time interval to gastric cancer onset to assess whether differences in β values ​​are associated with time to gastric cancer onset (d). In the gastric mucosa of persistent smokers, β values ​​increased over time, indicating that smoking enhances aberrant DNA methylation (e).We examined DNA methylation changes in H. pylori-infected individuals compared with non-heavy drinkers and non-smokers without gastric cancer to assess whether H. pylori and lifestyle factors play a significant role (f-h). We also assessed the role of lifestyle factors in promoting DNA methylation initiated by H. pylori (i-j). Figure 8 shows the results of a weighted Kaplan-Meier analysis of the relationship between DNA methylation levels and gastric cancer incidence. Figure 9 shows the results of a data analysis demonstrating that common markers are not necessarily methylated equally by each environmental factor. Figure 9 shows the correlation between the initial increase in β levels of common markers due to H. pylori (HP) infection and the accelerated increase in β levels due to the addition of alcohol consumption to H. pylori infection (a). Figure 9 shows the correlation between the initial increase in β levels of common markers due to H. pylori infection and the accelerated increase in β levels due to the addition of H. pylori infection and smoking (b). (e) shows the association between target genes methylated by HP infection (red: 13 + 9 + 8 + 7), genes whose methylation was accelerated by alcohol (green: 17 + 17 + 9 + 7), and genes whose methylation was accelerated by smoking (blue: 20 + 8 + 7 + 17). GO analysis of genes whose methylation was accelerated by alcohol (top) and smoking (bottom) is shown (f). Figure 10 shows Kaplan-Meier curves showing the association between HP status and gastric cancer incidence (a), and the association between HP status, DNA methylation levels of common markers, and gastric cancer incidence (b). Furthermore, weighted Kaplan-Meier curves showing the association between DNA methylation levels of common markers and gastric cancer incidence in (c) drinkers, (d) smokers, (e) subjects aged 60 years or older, and (f) men are shown. ROC curves are also shown (g, h). Furthermore, weighted Kaplan-Meier curves showing the association between DNA methylation levels of common markers and gastric cancer incidence in individuals with a family history of gastric cancer are shown (i). Figure 11-1 shows ROC curves showing the association between DNA methylation levels of NKX1-1, TMEM108, PTPN5, DBX1, KAZN, RYR2, CDH11, and KDR and gastric cancer incidence. The Youden's Index is plotted on each ROC curve, and the threshold values ​​(sensitivity and specificity) at that point are shown.Figure 11-2 shows weighted Kaplan-Meier curves indicating the association between DNA methylation levels of NKX1-1, TMEM108, PTPN5, DBX1, KAZN, RYR2, CDH11, and KDR and gastric cancer incidence. Figure 12 shows Kaplan-Meier curves plotting the association between H. pylori infection and alcohol consumption (a), H. pylori infection and smoking (b), and H. pylori infection, alcohol consumption, and smoking (c) and gastric cancer incidence in a combined analysis using the screening cohort and the GCEP cohort. ROC curves (a) and (b) show the association between the average DNA methylation level of 420 gene markers and gastric cancer incidence. The Youden's Index is plotted on each ROC curve, and the threshold (sensitivity, specificity) at that point is indicated. Fig. 14 is a diagram showing a schematic configuration of a gastric cancer development risk determination device of the present invention, and Fig. 15 is a diagram showing a schematic configuration of a gastric cancer development risk determination system of the present invention.

[0011] The present invention relates to a method for predicting a subject's risk of developing gastric cancer based on the DNA methylation level of a specific gene, and a system for determining the risk of developing gastric cancer. Specifically, the method for predicting the risk of developing gastric cancer according to the present invention comprises the following steps: detecting the DNA methylation level of at least one specific gene; and predicting the risk of developing gastric cancer by comparing the detected DNA methylation level with at least one threshold value of the methylation level corresponding to the risk of developing gastric cancer. This method makes it possible to stratify and predict the risk of developing gastric cancer in a subject. This method for predicting the risk of developing gastric cancer is preferably performed when a gastric biopsy sample is obtained during a health checkup, gastric cancer screening, or the like. Subjects predicted to be at high risk of developing gastric cancer may be provided with lifestyle advice to avoid either or both alcohol and smoking.

[0012] Increased methylation levels in gastric biopsy samples correlate with an increased incidence of future gastric cancer and reflect a shorter time to onset. Therefore, DNA methylation levels can be used as a quantitative marker for stratifying gastric cancer risk. Gastric cancer risk can be stratified into multiple stages, for example, two or three stages.

[0013] The gastric cancer risk determination system 20 according to the present invention includes a system based on a comparison of DNA methylation levels of specific genes with a threshold value, and a system using a machine learning model. These systems 20 are comprised of a detection device 30 that detects DNA methylation levels in a gastric biopsy sample and a gastric cancer risk determination device 10. The detection by the detection device 30 and the risk determination by the determination device 10 may be performed at the same location or at different locations. If performed at different locations, DNA methylation level data is provided to the determination device 10 via the Internet or media. The gastric cancer risk determined by the determination device 10 may further be provided to a terminal via the Internet or media. For example, the methylation level of a gastric biopsy sample may be detected at the hospital or clinic where the gastric biopsy sample was obtained, and the data thus obtained may be transmitted to an analysis center that owns the gastric cancer risk determination device 10. The risk may then be determined at the analysis center, and the results may be transmitted to the hospital or clinic, or to an individual. In another example, a gastric biopsy sample obtained at a hospital or clinic may be sent to an analysis center equipped with the detection device 30 and the determination device 10, where the methylation level is detected and the onset risk is determined, and the results may be transmitted to the hospital or clinic or an individual. In yet another example, the hospital or clinic that obtained the gastric biopsy sample may own the detection device 30 and the determination device 10, and the methylation level may be detected and the onset risk may be determined within the hospital or clinic.

[0014] Specifically, the system 20 based on comparison with a threshold value relates to the system including: a detection device 30 that detects the DNA methylation level of a specific gene in a gastric biopsy sample obtained from a subject; and an apparatus 10 for determining the risk of developing gastric cancer, wherein the apparatus for determining the risk of developing gastric cancer includes: an input unit 11 that inputs the DNA methylation level obtained from the detection device; a memory unit 12 that stores at least one threshold value of the methylation level corresponding to the risk of developing gastric cancer; a processing unit 13 that compares the DNA methylation level input from the input unit 11 with the at least one threshold value retrieved from the memory unit 12 and determines the risk of developing gastric cancer of the subject; and an output unit 14 that outputs the determined risk of developing gastric cancer of the subject.

[0015] An example of a system that also uses a machine learning model is the system comprising: a detection device 30 that detects the DNA methylation level of a specific gene in a gastric biopsy sample obtained from the subject, and a gastric cancer risk determination device 10, wherein the gastric cancer risk prediction determination device 10 comprises: an input unit 11 that inputs the DNA methylation level obtained from the detection device 30; a learning unit 15 that is pre-trained using training data including the DNA methylation level of a specific gene and information related to gastric cancer development, and outputs the gastric cancer development risk when input information including the DNA methylation level of the specific gene is input; and an output unit 14 that outputs the gastric cancer development risk of the subject.

[0016] In the gastric cancer risk prediction method and gastric cancer risk determination system of the present invention, the subjects whose risk is determined may be any subjects. However, from the viewpoint of more accurate risk prediction, subjects who are currently infected with H. pylori and / or have a history of H. pylori infection are preferred. The method of the present invention may be performed simultaneously with H. pylori testing, or even without taking into account the presence or absence of H. pylori infection. In another example, subjects whose risk is determined may have undergone H. pylori testing in parallel. Among subjects who have undergone H. pylori testing in parallel, subjects who have tested positive in the H. pylori test or whose antibody levels are above a predetermined level are preferred. In subjects with a history of H. pylori infection, continuous detection of DNA methylation levels using the method or system of the present invention allows for stratified prediction of gastric cancer risk and recommendations for lifestyle improvements. Furthermore, the frequency of endoscopic examinations can be determined according to the stratified gastric cancer risk. That is, subjects with low DNA methylation levels and a low risk of gastric cancer require less frequent endoscopic examinations, while subjects with high DNA methylation levels and a high risk of gastric cancer require more frequent endoscopic examinations. For example, subjects at low risk of developing gastric cancer may be offered endoscopic examinations every several years, e.g., every five years or every three years, whereas subjects at high risk of developing gastric cancer may be offered endoscopic examinations every year or multiple times per year, e.g., two, three or more times.

[0017] Specific genes whose methylation levels were detected in subjects with a history of H. pylori infection include SLC6A3, GFPT2, LOC392232, CCDC37-AS1, LMX1B, ZNF790-AS1, ZSCAN23, SSTR2, FZD10, ZSCAN18, MIR3663HG, MOS, ZNF660, NAALAD2, HAGLR, CFAP100, KCNA7, ZNF549, NCAN, HS3ST2, SRD5A2, PCDH8, GHSR, WTH3DI, KCNK10, ZNF880, CCDC67, LVRN, COMP, and NEUR. OG1, ZNF578, LRIF1, CADM3, EID3, GRP, KCNA1, SLC2A14, ST6GAL2, DMRTA2, MYOD1, GALR1, ISLR2, SIAH3, IRX4, ISC, DBX1, IRX6, AARD, FAM135B, HBQ1, A DCY8, LHCGR, SLITRK5, DLX5, GFRA1, MIR124-3, PABPC4L, TBX5-AS1, MIR9-3HG, DRD5, PCDH10, TOX2, KNSTRN, POU4F3, TMEFF2, CHRM2, ZNF582, CCDC36, C ADM1, NXPH2, MKX, GABRA5, TLX1, ZNF418, NKX2-2, COL4A2, OPRM1, TRPA1, FAR2P2, SLC27A6, CIDEA, CXCL12, PTPRT, PXDN, HIST1H1A, ZFHX4, GSX1, ADCYA P1, SYT6, ZFPM2, DMRT3, KHDRBS2, VSTM2B, LOC100287846, ELMO1, NPBWR1, DKK2, ERVMER34-1, NKX1-1, C11orf85, CABP7, HELT, LIN7A, ALDH1L1-AS2, NK X2-1, VAX1, SYT9, SUSD5, CCDC8, CRHR2, C8orf34-AS1, CTNND2, LINC01535, FOXI2, PCSK2, LOC101928134, NTMT1, BOLL, RXFP3, HAPLN4, PNPLA5, ZNF492 , WT1-AS, ZNF529, SERTM1, ERBB4, MDGA2, DPY19L2, IGFBP7-AS1, LOC101928796, USP44, SLCO4C1, BHLHE23, PYGO1, PLXNA4, SLC35F4, LBX2, LINC01248,<h2 style=";text-align:left;direction:ltr">CACNA1B、RNPS1、CPLX2、DCLK1、NKX2-1-AS1、RSPO3、HOXD1、PLPPR4、C1QL2、 CYTL1、OPRK1、ADAMTS18、RAX、PCDHGB7、TMEM108、KAZN、TDRP、NRXN3、FENDR R、ZIC5、EDNRB、RIMS2、LINC00485、EMX2、EVC2、PEAR1、T、ZNF702P、HOXD9、C NTNAP5、CFAP161、ZNF334、ACTA1、CYYR1、COL1A2、DYDC2、FBXL7、GDNF、ELAVL 2、HS3ST3A1、LOC283731、ERICH3、PCDHGA11、EFS、RALYL、EVX1、PCDHGA12、D DX25、DNAJC19、GPR139、SST、NOTO、SYN3、PCDHAC2、DOC2A、PREX2、TSPYL5、Z NF677、CBLN4、GRM2、NEUROD1、MCHR2、C11orf87、LAYN、UNCX、HECW1、PTH2R、 DYDC1、SPAG6、KCNK9、HMCN2、SNAP25、KDR、KIAA1024、ARHGEF10、HMGCLL1、SY T1、C3orf14、BARHL2、GLRA3、NWD2、SNAP91、KCNH7、ZNF582-AS1、ZNF300P1、 SLC32A1、PDE10A、RGS17、CCNA1、PCSK1、B3GAT1、HOXA9、HOXD10、DNAH8、FAM 184B、HMX2、B3GAT2、NXPH1、TRAM1L1、BVES-AS1、DSCAM、GBX2、NTSR2、GRIN2 A、EFTUD1P1、COL25A1、TUSC3、ACSS3、DSC3、NPR3、NEU1、ZIC4、SEMA6D、LINC0 1197、POU4F2、GUCY1B3、OLIG2、GALNT9、LOC101928174、NTM、FAM20A、GRIA4 、CPXM2、HOXA10、LEPR、PTPN5、PRMT8、CDO1、LHFPL4、SHISA9、ALDH1L1、CPNE8 、FAM19A4、C8orf88、CALN1、PDPN、SORCS1、CHL1、FSTL1、LYPD5、ADRA1D、C2o rf40、SP9、ALK、PDZRN4、ZSCAN1、PRLHR、NRSN1、FBN2、TLX3、ZSWIM2、GPR146、ZIK1, TRHDE-AS1, HOXD11, GRM7, ZNF365, NEFL, C10orf53, ALDH1A2, FGF10-AS1, AJAP1, RELN, KCNIP1, MIR2682, TMEM132E, NECAB1, ADCY5 , GJD2, EFCC1, GIPC3, LINC01231, PLPPR3, RYR3, CTD-2297D10.2, LINC00461, TBX18, CWH43, SELV, CCDC169, NID1, KCND2, ZNF835, DPP10, LOC284395, SSTR4, TRPC4, RNF212, GRIA2, EPDR1, CSMD1, RAPGEF4, REC8, HS6ST1, AGTR1, WBSCR17, C14orf39, LRRTM1, THSD7B, LOC100289 333, MIR4281, TMEM196, ROBO3, TCF15, GALNT13, GUCY1A3, ADAM33, PTH2, CDH7, EPHA7, WNT16, NKAIN4, NGF, ZNF132, PHACTR3, MEIOB, CDH1 8, SLC18A3, ZNF471, ZFP42, RASGRF1, GALR2, FGF3, DOCK10, TRHDE, SYT3, VWC2, SOX21, DBX2, FOXG1, ZNF667-AS1, ADGRV1, PCDHGB6, EPHA5 -AS1, USP6, ZNF382, ST8SIA2, UTF1, CDH4, CNTN4, KCNK2, NRG1, PHYHIPL, THCAT155, MYO3A, FAM19A5, PHOX2A, GABRG3, IGF2-AS, TCERG1L, and at least one gene selected from the group consisting of ZNF793, COL4A1, SORCS3, 11-Mar, PRR35, PLXDC2, TDRD10, LRRK2, ZFP28, NRIP3, ACAN, NLGN1, THY1, AFG3L1P, SFTA3, RAB12, TWIST2, FREM2, EBF2, FEZF2, CACNA1E, NTRK3, EMX2OS, HAND1, SNTG1, WNT7A, CCDC140, FGF14, PGR, SATB2-AS1, and GAD2. These genes were detected in 21 subjects who subsequently developed gastric cancer, 38 subjects who had H. pylori infection but did not develop gastric cancer, and 21 subjects who subsequently developed gastric cancer.These marker genes were selected by analyzing DNA methylation levels in gastric mucosal biopsies from around gastric cancer and from non-gastric cancer subjects ("non-gastric cancer, non-H. pylori subjects") in a screening cohort of 68 subjects, including 9 subjects with no H. pylori infection and no gastric cancer. The methylation level of at least one gene selected from this group of 420 genes can be used as a marker to distinguish between gastric mucosal biopsies from around gastric cancer and from non-gastric cancer subjects ("non-gastric cancer, non-H. pylori subjects") who are not infected with H. pylori and will not develop gastric cancer in the future. Specifically, if the DNA methylation level of at least one marker gene selected from this group of genes is equal to or greater than a predetermined threshold, it can be determined that a subject is at risk for developing gastric cancer. These specific genes may be used individually or as a group of genes containing the specific genes, which can be referred to as cohort-analyzed candidate gastric cancer risk marker genes. The cohort-analyzed candidate gastric cancer risk marker genes differ from the group of genes simply showing high methylation in the mucosa surrounding gastric cancer and low methylation in the mucosa of non-cancerous tissues, as in Patent Document 1 and Non-Patent Document 5, in that the cohort analysis was conducted by dividing subjects into a group that will develop gastric cancer in the future and a group that will not. While comprehensive DNA methylation analysis of gastric cancer was already identified in Kaneda et al. Cancer Res 2002, simply analyzing methylation in the mucosa surrounding gastric cancer and in the mucosa of non-cancerous tissues alone has not yet led to the identification of risk markers for gastric cancer. The cohort-analyzed candidate gastric cancer risk markers can be accurately referred to as gastric cancer risk markers because methylation changes may occur in normal tissues before the onset of gastric cancer. For these cohort-analyzed candidate gastric cancer risk marker genes, DNA methylation levels before the onset of gastric cancer can be measured, and P values ​​and AUC values ​​for the presence or absence of gastric cancer can be calculated to select cohort-analyzed candidate gastric cancer risk markers. Cohort analyzed gastric cancer risk marker genes in normal tissues before the onset of gastric cancer:These are genes whose DNA methylation levels fluctuate. High methylation levels of these genes in normal tissues indicate a high risk of developing gastric cancer in the future, while low methylation levels indicate a low risk of developing gastric cancer in the future. Therefore, by examining the DNA methylation of these genes in gastric biopsy samples before the onset of gastric cancer, the risk of developing gastric cancer in the future can be determined.

[0018] Among these specific genes, particularly NKX1-1, PABPC4L, SLC32A1, SYT6, TRPC4, WBSCR17, B3GAT1, ST8SIA2, RELN, DRD5, CPNE8, CACNA1B, DPP10, HOXD9, GRIA2, RXFP3, MKX, NKX2-2, LOC101928174, PREX2, GALNT13, CTD-2297D10.2, PTH2R, LYPD5, ADAMTS18, NXPH2, PCSK1, GSX1, GPR146, ZNF578, GUCY1B3, ZNF471, and FAM135B , GBX2, PLPPR3, CFAP100, DBX1, NRG1, HMGCLL1, DDX25, ADRA1D, ADCY8, ACTA1, KAZN, SFTA3, ZFP42, NRIP3, NWD2, CADM3, ADCYAP1, MCHR2, HS3ST2, CYYR1 , EVX1, TCERG1L, ADAM33, NLGN1, EID3, CDH4, EFCC1, ISLR2, PHYHIPL, SLITRK5, FGF3, TMEFF2, SEMA6D, HOXD10, PLXDC2, NKX2-1-AS1, ELAVL2, SYT9, AJA P1, ZSWIM2, SNAP91, GABRG3, PLPPR4, ZNF365, C14orf39, ERBB4, LMX1B, IRX4, FBN2, ACAN, EVC2, TBX5-AS1, DYDC2, HS3ST3A1, IRX6, SLC27A6, PCDH8, CN TN4, BARHL2, SNAP25, RAPGEF4, HOXD11, USP6, LOC392232, HAGLR, PRLHR, TUSC3, SERTM1, VSTM2B, KHDRBS2, FOXI2, TDRP, TDRD10, FGF14, NTM, ST6GAL2, PRMT8, CCDC36, KDR, PGR, KCNIP1, ZNF702P, GABRA5, RNPS1, NTRK3, BOLL, EDNRB, LEPR, RALYL, DSCAM, ROBO3, IGF2-AS, HOXD1, SSTR4, GALR1, POU4F2, ME IOB, ISC, ZNF382, PCSK2, DCLK1, UNCX, LINC01248, DYDC1, OPRM1, COL4A2, LINC01197, ZNF880, CPLX2, GALNT9, GRM7, TRAM1L1, GDNF, SORCS1, ZSCAN1,From the viewpoint of providing a more accurate prediction of the risk of developing gastric cancer, it is preferable to use at least one gene selected from the group consisting of EGFR and EFS. These 149 marker genes were selected by extracting "genes that satisfy p<0.05 and AUC value>0.75 among the top 150 genes with the lowest P values" in a subgroup of H. pylori-infected individuals. Therefore, if the DNA methylation level of at least one marker gene selected from the group consisting of these 149 genes is equal to or greater than a predetermined threshold, it can be determined with higher sensitivity that an H. pylori-infected individual is at risk of developing gastric cancer. These specific genes may be used alone, or may be used as a gene group including the specific genes.

[0019] Furthermore, the specific genes to be used may vary depending on the risk factors of the subject. Risk factors include smoking habits, drinking habits, gender, age, family history, etc. Examples of specific genes include the following: NKX1-1, PABPC4L, SLC32A1, SYT6, TRPC4, WBSCR17, B3GAT1, ST8SIA2, RELN, DRD5, CPNE8, CACNA1B, DPP10, HOXD9, GRIA2, RXFP3, MKX, NKX2-2, LOC101928174, PREX2, GALNT13, CTD-2297D10.2, PTH2R, LYPD5, ADAMTS18, NXPH2, PCSK1, GSX1, GPR146, and ZNF5. 78, GUCY1B3, FAM135B, GBX2, PLPPR3, DBX1, NRG1, HMGCLL1, DDX25, ADRA1D, ADCY8, ACTA1, KAZN, SFTA3, ZFP42, NRIP3, NWD2, C ADM3, HS3ST2, CYYR1, EVX1, TCERG1L, ADAM33, NLGN1, EID3, CDH4, EFCC1, ISLR2, PHYHIPL, SLITRK5, FGF3, TMEFF2, NKX2-1-AS 1, ELAVL2, SYT9, AJAP1, ZSWIM2, SNAP91, GABRG3, PLPPR4, ZNF365, C14orf39, ERBB4, LMX1B, IRX4, ACAN, EVC2, TBX5-AS1, DYD C2, SLC27A6, PCDH8, CNTN4, LOC392232, HAGLR, PRLHR, TUSC3, SERTM1, KHDRBS2, FOXI2, TDRP, FGF14, NTM, ST6GAL2, PRMT8, CC At least one gene selected from the group consisting of DC36, KDR, PGR, KCNIP1, ZNF702P, BOLL, EDNRB, RALYL, FOXG1, SLC2A14, PCDHGB7, GRM2, CTNND2, LINC00485, ERVMER34-1, CCDC37-AS1, FAM19A4, TBX18, MIR9-3HG, LOC100289333, LINC00461, TRHDE, FZD10, and FAM19A5 is included.These 117 marker genes were selected by extracting "genes that satisfy p<0.05 and AUC>0.75 among the top 150 genes with the lowest P values" in a subgroup of subjects with alcohol consumption, and are particularly preferred for use in subjects with alcohol consumption. Therefore, in H. pylori-infected subjects with alcohol consumption, if the DNA methylation level of at least one marker gene selected from the group consisting of these 117 genes is equal to or greater than a predetermined threshold, it can be determined with high sensitivity that the subject is at risk of developing gastric cancer. These specific genes may be used individually, or as a gene group including the specific genes.

[0020] Specific genes include:NKX1-1, PABPC4L, SLC32A1, SYT6, TRPC4, WBSCR17, B3GAT1, ST8SIA2, RELN, DRD5, CPNE8, CACNA1B, DPP10, HOXD9, GRIA2, RXFP3, MKX, NKX2-2, LOC10192 8174, PREX2, GALNT13, CTD-2297D10.2, PTH2R, LYPD5, ADAMTS18, NXPH2, PCSK1, GSX1, GPR146, ZNF578, GUCY1B3, FAM135B, GBX2, PLPPR3, DBX1, NRG1, HM GCLL1, DDX25, ADRA1D, ADCY8, ACTA1, KAZN, SFTA3, ZFP42, NRIP3, NWD2, CADM3, HS3ST2, CYYR1, EVX1, TCERG1L, ADAM33, NLGN1, EID3, EFCC1, SLITRK5, F GF3, TMEFF2, NKX2-1-AS1, ELAVL2, SYT9, AJAP1, ZSWIM2, SNAP91, GABRG3, PLPPR4, ZNF365, C14orf39, ERBB4, LMX1B, IRX4, ACAN, TBX5-AS1, DYDC2, SLC2 7A6, HAGLR, PRLHR, SERTM1, KHDRBS2, TDRP, FGF14, NTM, PRMT8, CCDC36, KDR, ZNF702P, EDNRB, PCDHGB7, LINC00485, ERVMER34-1, CCDC37-AS1, TBX18, M IR9-3HG, LOC100289333, LINC00461, FZD10, CFAP100, MCHR2, PLXDC2, HS3ST3A1, BARHL2, SNAP25, RAPGEF4, NTRK3, DSCAM, ROBO3, IGF2-AS, GALR1, INSC , ZNF382, PCSK2, DCLK1, UNCX, DYDC1, OPRM1, COL4A2, GRM7, SORCS1, SLC18A3, SPAG6, PCDHGB6, ADCY5, ZNF529, THSD7B, COL25A1, PCDHGA11, NPBWR1, VAX1, WTH3DI, ZSCAN18, PCDHAC2, DMRTA2, KCNH7, MOS, TLX3, LAYN, SORCS3, NKAIN4, KIAA1024, CCDC140, NRXN3, and EMX2OS.These 142 marker genes were selected by extracting "genes satisfying p<0.05 and AUC>0.75 among the top 150 genes with the lowest P values" in a subgroup of subjects with smoking habits, and are particularly suitable for use in subjects with smoking habits. Therefore, in H. pylori-infected subjects with smoking habits, if the DNA methylation level of at least one marker gene selected from the group consisting of these 142 genes is equal to or greater than a predetermined threshold, the risk of developing gastric cancer can be determined with high sensitivity. These specific genes may be used alone or as part of a gene group containing the specific genes. Use of such specific genes allows for more accurate determination of the risk of developing gastric cancer in subjects with smoking habits.

[0021] Specific genes include NKX1-1, PABPC4L, SLC32A1, SYT6, TRPC4, WBSCR17, B3GAT1, ST8SIA2, RELN, DRD5, CPNE8, CACNA1B, DPP10, HOXD9, GRIA2, RXFP3, MKX, NKX2-2, LOC101928174, PREX2, GALNT13, CTD-2297D10.2, PTH2R, LYPD5, ADAMTS18, NXPH2, PCSK1, GSX1, GPR146, ZNF578, GUCY1B3, FAM135B, GBX2, PLPPR3, DBX1, NRG1, HMGCLL1, DDX25, ADRA1D, ADCY8, ACTA1, KAZN, SFTA3, ZFP42, and NRIP. It is more preferred to use at least one gene selected from the group consisting of 3, NWD2, CADM3, HS3ST2, CYYR1, EVX1, TCERG1L, ADAM33, NLGN1, EID3, EFCC1, SLITRK5, FGF3, TMEFF2, NKX2-1-AS1, ELAVL2, SYT9, AJAP1, ZSWIM2, SNAP91, GABRG3, PLPPR4, ZNF365, C14orf39, ERBB4, LMX1B, IRX4, ACAN, TBX5-AS1, DYDC2, SLC27A6, HAGLR, PRLHR, SERTM1, KHDRBS2, TDRP, FGF14, NTM, PRMT8, CCDC36, KDR, ZNF702P, and EDNRB. Such specific genes are commonly selected in subjects with smoking habits and subjects with drinking habits. Therefore, by using these genes, the risk of developing gastric cancer can be determined regardless of smoking and drinking habits. Furthermore, even in subjects with a family history of gastric cancer, these common markers can stratify the risk of developing gastric cancer (Figure 10i). Therefore, in subjects with a family history, the use of these common markers can more accurately determine the risk of developing cancer.

[0022] Specific genes include NKX1-1, NKX2-2, DRD5, PABPC4L, ADAMTS18, FAM135B, ZNF578, PTH2R, SLC32A1, HS3ST2, B3GAT1, SLITRK5, SYT6, RXFP3, KHDRBS2, PCSK1, COL4A2, DBX1, EDNRB, ZNF365, HOXD9, MKX, GABRG3, ACTA1, and SERTM1. It is preferable to use at least one gene selected from the group consisting of FGF14, TRPC4, ZSWIM2, FZD10, PCDHAC2, CACNA1B, PLPPR4, ADRA1D, RELN, GRIA2, DPP10, HMGCLL1, LOC101928174, TBX18, PREX2, CNTN4, CCDC36, WBSCR17, ADCY8, and C14orf39. These genes are related to the purpose of listing all of the top 30 genes with high AUC values ​​in the groups with Helicobacter pylori infection, drinking habit, and smoking habit, respectively. In particular, at least one gene selected from the group consisting of NKX1-1, NKX2-2, DRD5, PABPC4L, ADAMTS18, FAM135B, ZNF578, PTH2R, SLC32A1, HS3ST2, B3GAT1, SLITRK5, SYT6, RXFP3, KHDRBS2, and PCSK1 is a selection of overlapping genes with the top 30 AUCs in subject groups with Helicobacter pylori, drinking habits, and smoking habits, respectively, and can stratify the risk of developing gastric cancer with high accuracy, regardless of the subject group's Helicobacter pylori, drinking habits, and smoking habits.

[0023] The specific gene is preferably at least one gene selected from the group consisting of PABPC4L, DRD5, NKX1-1, NKX2-2, HS3ST2, SLITRK5, ZNF578, RXFP3, ADAMTS18, SYT6, B3GAT1, FAM135B, PTH2R, ZNF365, ACTA1, PCDHAC2, DBX1, MKX, FZD10, TRPC4, SLC32A1, CACNA1B, COL4A2, PCSK1, KHDRBS2, LOC101928174, DPP10, RELN, CCDC36, WBSCR17, PREX2, PLPPR4, and CNTN4. These genes are listed as the top 20 genes with the highest AUC values ​​in the groups with Helicobacter pylori infection, drinking habits, and smoking habits, respectively. In particular, at least one gene selected from the group consisting of PABPC4L, DRD5, NKX1-1, NKX2-2, HS3ST2, SLITRK5, ZNF578, RXFP3, ADAMTS18, SYT6, B3GAT1, and FAM135B is a selection of overlapping genes with the top 20 AUCs in subject groups with Helicobacter pylori, drinking habits, and smoking habits, respectively, and can stratify the risk of developing gastric cancer with high accuracy, regardless of subject groups with Helicobacter pylori, drinking habits, and smoking habits.

[0024] The cohort analyzed candidate gastric cancer risk marker genes are as follows: NKX1-1, SLC32A1, CACNA1B, RXFP3, PABPC4L, PTH2R, B3GAT1, FAM135B, SYT6, ZNF578, ADAMTS18, ADRA1D, RELN, DBX1, DRD5, LOC101928174, DPP10, HS3ST2, NKX2-2, PLPPR4, KHDRBS2, PCSK1, SLITRK5, ZNF365, WBSCR17, MKX, CTD-2297D10.2, PREX2, NXPH2, ADCY8, NRG1, ZFP42, HOXD9, TRPC4, CNTN4, LYPD5 , CYYR1, GABRG3, PLPPR3, TCERG1L, C14orf39, ACTA1, ST6GAL2, KAZN, HMGCLL1, SERTM1, ACAN, ZSWIM2, EDNRB, SORC S1, CCDC36, COL4A2, FGF14, MIR9-3HG, FZD10, PCDHGB6, WTH3DI, NKAIN4, TBX18, FOXG1, NKX2-1-AS1, EID3, CADM3 , TMEFF2, ZNF702P, PCDHAC2, TRHDE, MOS, LOC100289333, GRIA2, GALNT13, GBX2, ST8SIA2, DDX25, CDH4, AJAP1, LMX It is preferable to use at least one gene selected from the group consisting of 1B, FOXI2, PRLHR, SFTA3, TBX5-AS1, HAGLR, ERBB4, UNCX, SLC27A6, EFS, NLGN1, GALR1, ZSCAN1, IRX4, VAX1, FAM19A4, SOX21, COL25A1, PHOX2A, CNTNAP5, PXDN, C8orf34-AS1, WNT16, and HBQ1. These genes are a list of all genes that have an AUC value of more than 0.85 and a p-value of less than 0.01 in the groups with H. pylori, drinking habit, and smoking habit, respectively.In particular, at least one gene selected from the group consisting of NKX1-1, SLC32A1, CACNA1B, RXFP3, PABPC4L, PTH2R, B3GAT1, FAM135B, SYT6, ZNF578, ADAMTS18, ADRA1D, RELN, DBX1, DRD5, LOC101928174, DPP10, HS3ST2, NKX2-2, PLPPR4, KHDRBS2, PCSK1, and SLITRK5 was selected based on overlapping genes in subject groups with Helicobacter pylori, drinking habits, and smoking habits, respectively, and can stratify the risk of developing gastric cancer with high accuracy regardless of the subject group's Helicobacter pylori, drinking habits, and smoking habits.

[0025] The cohort analyzed candidate gastric cancer risk marker gene is preferably at least one gene selected from the group consisting of NKX1-1, PABPC4L, DRD5, RXFP3, PTH2R, ZNF578, ADAMTS18, HS3ST2, NKX2-2, SLITRK5, ZNF365, SLC32A1, CACNA1B, WBSCR17, B3GAT1, FAM135B, SYT6, PREX2, ADRA1D, CNTN4, PLPPR4, and CCDC36. These genes are a list of all genes with an AUC value of more than 0.90 and a p-value of less than 0.01 in the groups with Helicobacter pylori, drinking habit, and smoking habit, respectively. In particular, at least one gene selected from the group consisting of NKX1-1, PABPC4L, DRD5, RXFP3, PTH2R, ZNF578, ADAMTS18, HS3ST2, NKX2-2, and SLITRK5 was selected based on overlapping genes in subject groups with Helicobacter pylori, drinking habits, and smoking habits, respectively, and can stratify the risk of developing gastric cancer with high accuracy regardless of the subject group's Helicobacter pylori, drinking habits, and smoking habits.

[0026] The specific gene may be selected from the viewpoint of ease of designing primers for pyrosequencing in addition to its ability to predict cancer risk, and examples of the specific gene selected from this viewpoint include at least one gene selected from the group consisting of NKX1-1, TMEM108, PTPN5, DBX1, KAZN, RYR2, CDH11, and KDR.

[0027] Drinking habits, smoking habits, and family history can be determined by interviews or questionnaires. Smoking habits can be classified into groups based on smoking frequency and number of cigarettes smoked. For example, groups may be classified based solely on whether or not a person smokes, or on the number of cigarettes smoked per day. Drinking habits can be classified into groups based on drinking frequency and amount. Family history can be classified based on whether or not a close relative, particularly a first-degree relative, has developed stomach cancer.

[0028] In the present invention, the methylation level of a specific gene refers to the presence or absence of methylation at any methylation site (CpG) contained in a specific gene region. Among the regions of a specific gene, the presence or absence of methylation at a methylation site in a CpG island within or near the promoter region is particularly preferred. A CpG island refers to a region containing highly conserved CpGs in the genome, and is generally defined as having a length of at least 200 base pairs, a GC content of 50% or more, and a proportion of CpGs present that is 60% or more of the amount expected from the GC content, but is not intended to be limited to such a definition. In the present invention, the methylation level may be a percentage based on the presence or absence of methylation at one or more methylation sites (CpGs) at a specific position, or a percentage based on the presence or absence of methylation at multiple methylation sites (CpGs). The methylation level can be calculated for multiple cells contained in a sample as a percentage of the presence or absence of methylation at one or more methylation sites. When performing Infinium analysis using Infinium's DNA methylation array, the methylation level in the methylation region (cgID) set by Infinium for a specific gene can be used.

[0029] The methylation level can be determined using a methylation level determination technique known in the art. For example, the methylation level can be determined using a DNA methylation array, next-generation sequencing, and pyrosequencing. From the viewpoint of detecting the methylation levels of a large number of specific genes, it is preferable to use a DNA methylation array or next-generation sequencing, but from the viewpoint of detecting the methylation levels of a limited number of specific genes, it is preferable to use pyrosequencing.

[0030] Gastric biopsy samples are obtained by endoscopic examination, and are typically tissue fragments of approximately 2 mm in diameter from a portion of the stomach mucosa. While biopsy samples can be obtained from any location in the stomach, they may be obtained from areas with a high likelihood of developing gastric cancer, such as areas with redness, erosion, or ulceration, or from asymptomatic mucosa.

[0031] Gastric cancer whose risk of onset is predicted by the present invention includes all primary gastric cancers, such as adenocarcinoma, mucinous carcinoma, endocrine cell carcinoma, signet ring cell carcinoma, and lymphocytic infiltrative carcinoma, and further includes dysplasia and intraepithelial carcinoma.

[0032] [Threshold] The threshold used to determine the risk of developing gastric cancer can be determined using statistical methods. As an example, the DNA methylation level of a specific gene is determined for gastric biopsy samples obtained from a group of asymptomatic subjects with no history of gastric cancer. A follow-up period is then set, and subjects who develop primary gastric cancer during the follow-up period are identified, thereby obtaining a cohort. Risk factors for each subject, such as drinking habits, smoking habits, age, sex, and family history, as well as a history of Helicobacter pylori infection, are also determined. The follow-up period may be set as appropriate. In this cohort, a threshold can be determined using an ROC curve, with or without considering each risk factor. An example of an index used as a reference when testing the threshold is Youden's index. If the DNA methylation level of a specific gene is higher than the threshold, it can be determined that the risk of developing primary gastric cancer in the future is high. The threshold can also be set in multiple stages depending on the risk.

[0033] [Detection Device] The detection device 30 may be a DNA methylation array analysis device, a next-generation sequencer, a pyrosequencer, or any other device capable of detecting the methylation level of a specific gene. As the DNA methylation array analysis device, the Infinium Methylation Analysis Kit from Illumina may be used.

[0034] [Device for Determining Risk Prediction] The device 10 for determining a risk of developing stomach cancer in the system 20 based on comparison with a threshold value includes an input unit 11, a memory unit 12, a processing unit 13, and an output unit 14. A DNA methylation level obtained from a detection device 30 is input to the device 10 via the input unit 11. The memory unit 12 stores at least one threshold value for the methylation level corresponding to one or more specific genes used for determining the risk of developing stomach cancer. The processing unit 13 retrieves the threshold value stored in the memory unit 12 and compares it with the DNA methylation level input from the input unit 11 to determine the subject's risk of developing stomach cancer. The device 10 for determining a risk of developing stomach cancer in the system 12 using a machine learning model further includes a learning unit 15. The learning unit 15 determines the risk of developing stomach cancer instead of or in addition to the processing unit 13.

[0035] The input unit 11 relates to a communication interface such as a LAN or a port for connecting the determination device to a network, an operation unit such as a keyboard or a mouse, or any interface connected to an external storage device such as a CD-ROM, a DVD-ROM, a BD-ROM, a memory stick, etc. The input unit 11 provides data received from an external device or the like to the processing unit 13.

[0036] The output unit 14 may be a display device such as a liquid crystal display, a printer, or other output means, or may be an interface for output to an external terminal or output via a network. Such an interface may be a communication interface such as a LAN or a port. The output unit 14 displays information about the subject's gastric cancer risk transmitted by the determination device. The determination device may simultaneously display the determined gastric cancer risk, as well as DNA methylation levels and lifestyle advice.

[0037] The storage unit 12 includes a memory device such as RAM, ROM, or flash memory, a fixed disk device such as a hard disk drive, or a portable storage device such as a flexible disk or optical disk. The storage unit 12 stores information input from the input unit 11, the risk of developing stomach cancer determined by the processing unit 13 and the learning unit 15, as well as programs, databases, output forms, and the like used for various computer processes. The computer program may be installed from a computer-readable recording medium such as a CD-ROM or DVD-ROM, or via the Internet. The computer program is installed in the storage unit using a known setup program or the like.

[0038] The processing unit 13 executes various arithmetic processes according to the programs stored in the memory unit 12. The arithmetic processes are performed by a processor included in the processing unit 13. This processor includes functional modules that control the memory unit 12, the input unit 11, the learning unit 15, and / or the output unit 14, and is capable of performing various controls. Each of these units may be composed of an independent integrated circuit, microprocessor, firmware, etc. The processing unit 13 reads information about DNA methylation levels for specific genes input from the input unit 11 and information about thresholds pre-stored in the memory unit, and compares the information with the DNA methylation level to determine the subject's risk of developing gastric cancer. The processing unit 13 controls the output unit 14 to transmit the determined information about the subject's risk of developing gastric cancer.

[0039] The learning unit 15 is pre-trained using a known machine learning technique such as deep learning to determine the relationship between the DNA methylation level of a specific gene and training data containing information on the onset of gastric cancer. Thus, the learning unit 15 is trained to output a risk of developing gastric cancer when input information containing the DNA methylation level of a specific gene is input. In the pre-training, further target information, such as at least one piece of information selected from the group consisting of smoking habit, drinking habit, age, gender, and family history, may be input and trained. Thus, the learning unit 15 is trained to output a risk of developing gastric cancer when input information includes the DNA methylation level of a specific gene and at least one piece of information selected from the group consisting of smoking habit, drinking habit, age, gender, and family history. Deep learning is machine learning using a multilayer neural network consisting of an input layer, an intermediate layer, and an output layer. A feature vector of the DNA methylation level of a specific gene is input to each node of the input layer. Each node in the intermediate layer outputs the sum of values ​​obtained by multiplying each feature vector output from each node in the input layer by a weight, and the output layer outputs the sum of values ​​obtained by multiplying each feature vector output from each node in the intermediate layer by a weight. The learning unit 15 adjusts each weight while learning so as to reduce the difference between the output value from the output layer and the information related to the onset of stomach cancer.

[0040] All documents mentioned herein are incorporated by reference in their entirety.

[0041] The following examples of the present invention are for illustrative purposes only and do not limit the technical scope of the present invention. The technical scope of the present invention is limited only by the claims. The present invention may be modified, for example, by adding, deleting, or substituting components of the present invention, provided that the modifications do not depart from the spirit of the present invention.

[0042] Example 1: Cohort Study. A cohort study was conducted on the Check-up cohort (Fig. 1a and c) and the Gastric Cancer Epidemiology Program (GCEP) cohort (Fig. 1b and d) at the Preventive Medicine Center of NTT East Kanto Hospital in Tokyo. (1) Check-up Cohort. A longitudinal cohort study was conducted at the Preventive Medicine Center of NTT East Kanto Hospital in Tokyo. The hospital conducts early disease detection and prevention research among healthy individuals. This health check program included a medical interview, physical examination, physiological tests, blood tests, upper endoscopy, and low-dose computed tomography for cancer screening. Healthy individuals who participated in the health check program and had gastric mucosal tissue collected by upper endoscopy between March 2006 and November 2018 were included in this study. Follow-up of the subjects ended in March 2019. The exclusion criteria were as follows: (i) a history of gastric cancer, (ii) a diagnosis of gastric cancer at the time of the health check or within 6 months after the health check, and (iii) no follow-up upper endoscopy examination more than 6 months after the initial health check. In this study, the start date of follow-up was the date on which a gastric mucosal biopsy was performed during the health check. The end date of follow-up was the date of gastric cancer diagnosis. If endoscopic follow-up was not performed, the patient was considered censored, and the date of the last endoscopy examination was defined as the censoring date.

[0043] Clinical Data Collected from the Screening Cohort During the screening program, participants were interviewed by a specialist regarding their age, biological sex, lifestyle habits, and family, medical, and treatment history for various diseases. Alcohol consumption was assessed based on the amount of ethanol consumed per week. Participants were categorized into four groups: non-drinkers (<40 g / week), light drinkers (40-140 g / week), moderate drinkers (140-280 g / week), and heavy drinkers (≥280 g / week). Smoking habits were assessed based on pack-years (pack-years of cigarettes consumed per day multiplied by the number of years of smoking). Participants were categorized into three groups: non-smokers (<1 pack-year), light smokers (1-40 pack-years), and heavy smokers (≥40 pack-years). Smokers were also asked whether they had quit smoking and were classified into four groups: former light smokers, current light smokers, former heavy smokers, and current heavy smokers.

[0044] Gastric atrophy was assessed by endoscopists certified by the Japan Gastroenterological Endoscopy Society. Patients were classified into three groups based on the degree of atrophy: non-atrophic, closed atrophic, and open atrophic. H. pylori serum antibody titers were measured using an enzyme-linked immunosorbent assay (ELISA) according to the manufacturer's instructions (Eiken Chemical, Tokyo, Japan). The antibody titer threshold for H. pylori seropositivity was 10 U / mL. Patients with a positive H. pylori serum antibody test or confirmed gastric atrophy by endoscopic examination were considered to be H. pylori infected. Patients with a negative H. pylori serum antibody test and no endoscopic atrophy were considered to be uninfected with H. pylori. H. pylori eradication was assessed by a gastroenterologist certified by the Japan Gastroenterological Association based on the eradication history obtained through interview, endoscopic findings, and the results of a urea breath test.

[0045] DNA methylation analysis was performed on the following biopsy samples from the screening cohort: (i) gastric biopsies (n = 21) from subjects who subsequently developed gastric cancer (future gastric cancer patients); (ii) gastric biopsies (n = 38) from H. pylori-infected subjects who did not subsequently develop gastric cancer (non-H. pylori-infected subjects); (iii) gastric biopsies (n = 9) from non-H. pylori-infected, non-gastric cancer subjects (non-H. pylori-infected subjects); and (iv) gastric cancer tissues (n = 28) and surrounding mucosa (n = 28) using the Infinium MethylationEPIC BeadChip Kit (Illumina, San Diego, CA, USA). A family history of gastric cancer was considered positive if a first-degree relative had a history of gastric cancer.

[0046] Additional biopsy specimens were also collected from three of the 21 prospective gastric cancer patients during separate health checkups (resulting in a total of 24 biopsy specimens). To avoid bias from multiple specimens from the same patient, the second gastric mucosal biopsies (n = 3) were used only for the comparative analysis between continuing and former smokers. Gastric cancer tissues (EBV CpG island methylator phenotype (EBV-CIMP) or high methylation epigenotype (HME)) (n = 20) and their surrounding mucosa (n = 17) were collected at the University of Tokyo Hospital to select marker genes for each epigenotype.

[0047] (2) Gastric Cancer Epidemiology Program Cohort. The Gastric Cancer Epidemiology Program (GCEP) is a large-scale prospective cohort study of 2,980 Chinese subjects aged 50 years or older in Singapore. In the GCEP cohort, subjects were longitudinally followed, and 21 of them developed early gastric neoplasia (EGN) (high-grade dysplasia (n = 13) or early gastric cancer (n = 8)). Clinical data, including alcohol intake, smoking habits, and endoscopic data on H. pylori status, were collected from GCEP subjects. Alcohol intake was assessed in terms of ethanol intake per week, and subjects were divided into four groups, similar to the health checkup cohort. Smoking habits were assessed based on pack-years (number of packs of cigarettes consumed per day multiplied by the number of years of smoking), and subjects were divided into three groups, similar to the health checkup cohort. H. pylori infection was assessed using enzyme-linked immunosorbent assay (ELISA). Subjects with negative H. pylori serum antibody tests, no history of eradication, and no atrophic gastritis were considered H. pylori-uninfected. In the GCEP cohort, six gastric mucosal biopsy specimens were collected and evaluated for the presence or absence of pathological atrophy and the degree of intestinal metaplasia by two independent, experienced pathologists. In addition, three biopsies were taken from the pylorus and stored in liquid nitrogen until molecular analysis (including infinium methylation assay).

[0048] Clinical Data and Sample Collection from the Gastric Cancer Epidemiology Program (GCEP) Cohort Among these subjects, 2,565 were selected from the GCEP, a large-scale prospective cohort study in Singapore (Fig. 1b). Infinium Human Methylation 450 BeadChip data from frozen gastric mucosa (n = 137) biopsied from the pylorus collected from the GCEP cohort (accession number: GSE 103186) were analyzed (Fig. 1d).

[0049] Infinium Methylation Assay. DNA methylation analysis was performed using the Infinium MethylationEPIC BeadChip Kit (Illumina, San Diego, CA, USA) on 164 extracted DNA samples. The Infinium MethylationEPIC BeadChip contains probes designed for approximately 867,000 CpG sites. The ratio of methylated probes to the sum of methylated and unmethylated probes at each CpG site was measured as the β value. β values ​​range from 0.00 to 1.00 and reflect the methylation level of the CpG site. Bisulfite conversion was performed using the Zymo EZ DNA methylation kit (Zymo Research, Irvine, CA, USA), and whole-genome amplification, labeling, hybridization, and scanning were performed according to the manufacturer's instructions. Infinium probes were annotated to the nearest gene using annotatePeaks.pl in HOMER (http: / / homer.salk.edu / homer / index.html), and the distance to the transcription start site (TSS) was calculated. The promoter region was defined as the region from -1,000 to +100 bp surrounding the transcription start site. The CpG score for each probe was calculated based on the CpG ratio within a 500-bp region centered on the probe site. Probes were classified into three categories based on their CpG score: high-CpG probes, intermediate-CpG probes, and low-CpG probes. In this analysis, high-CpG probes and intermediate-CpG probes with a CpG score of >0.48 were used. Based on the β values ​​of five control samples methylated at 0%, 25%, 50%, 75%, and 100%, (i) the β value was <0.2 for the 0% control sample, (ii) the β value was >0.4 for the 100% control sample, and (iii) the coefficient of determination (R 2Non-quantitative probes that did not meet a p-value >0.8 were excluded. Probes with a p-value >0.01 were treated as missing beta values, and probes with missing values ​​in multiple samples were excluded from the analysis. For each gene, except for genes on the X and Y chromosomes, the probe closest to the TSS in the promoter region was selected. The Infinium data for the stomach tissue collected in this study are available in the Gene Expression Omnibus (GEO) dataset (accession number: GSE199898).

[0050] Example 2: Identification of epigenotypes in gastric cancer cases and identification of epigenotype classification markers To identify the epigenotypes of gastric cancer cases in this study, probes with a standard deviation of 0.1 or greater were extracted, and then unsupervised two-way hierarchical clustering analysis was performed using the Infinium data of gastric cancer cases (n = 48) and non-gastric cancer, non-H. pylori mucosa (n = 9). Next, marker genes were identified for each epigenotype using the Infinium data of gastric cancer cases (n = 48) and non-gastric cancer, non-H. pylori mucosa (n = 9). EBV markers were defined as genes that met the following criteria: (i) a mean β value of greater than 0.3 in EBV gastric cancer patients, (ii) a β value of less than 0.2 in fewer than three EBV gastric cancer patients, and (iii) a mean β value of less than 0.15 in other samples. HME markers were defined as genes that met the following criteria: (i) mean beta values ​​greater than 0.3 in EBV gastric cancer or HME patients; (ii) beta values ​​less than 0.2 in fewer than three EBV gastric cancer or HME patients; and (iii) mean beta values ​​less than 0.15 in other samples. Common markers were defined as genes with the following criteria: (i) mean beta values ​​greater than 0.3 in patients with all epigenotypes; (ii) beta values ​​less than 0.2 in fewer than three patients with all epigenotypes; and (iii) mean beta values ​​less than 0.15 in non-gastric cancer, non-H. pylori mucosa.

[0051] Identification of DNA methylation epigenotypes and marker genes. Gastric cancer methylomes were analyzed to identify DNA methylation epigenotypes reported in gastric cancer, particularly EBV-CIMP, HME, and LME. Unsupervised two-way hierarchical clustering of gastric cancer and non-gastric cancer non-H. pylori mucosa (Figure 2) was performed, and three methylation subtypes were identified: EBV-CIMP, HME, and LME. As previously reported, all gastric cancer cases showing EBER-ISH positivity were classified as EBV-CIMP, and gastric cancer cases showing MSI-high or MLH1 promoter methylation were classified as HME. As a result, 271 EBV markers were identified that were methylated in EBV gastric cancer (EBV-CIMP), 158 HME markers that were methylated in EBV gastric cancer and HME, and 156 common markers that were methylated in all gastric cancers but not in non-gastric cancer, non-H. pylori mucosa (Figure 3, Table 1). The 156 genes selected as common markers are hereinafter referred to as the 156 common marker genes. The relationship between cgID and genes can be determined by referring to the publicly available information from Infinium.

[0052] To verify whether these markers are suitable for epigenotyping, we analyzed Infinium HumanMethylation450 BeadChip data from gastric cancer (frozen samples, n=395) registered in The Cancer Genome Atlas Program (TCGA, https: / / cancergenome.nih.gov / ). For this analysis, we pre-ordered genes and performed one-way hierarchical clustering analysis, specifically targeting markers present on both the Infinium HumanMethylation450 and Infinium MethylationEPIC platforms (Figure 4). Gene ontology (GO) terms enriched in the shared markers were involved in pattern specification, regionalization, cell morphogenesis, cell fate, and stem cell development (Figure 5). The shared markers were also enriched for genes previously reported to function as tumor suppressors, including USP6, NRG1, and MDGA2 (Table 2).

[0053] Example 3: DNA methylation analysis using gastric cancer tissue and its surrounding mucosa We investigated whether EBV, HME, and common markers were also methylated in the surrounding mucosa present alongside gastric cancer tissue (Fig. 6a). In EBV gastric cancer patients, EBV, HME, and common markers were highly methylated in gastric cancer tissue, whereas only common markers were lightly methylated in the gastric mucosa surrounding EBV gastric cancer. Notably, the methylation β value of common markers in the mucosa surrounding EBV gastric cancer was higher than that in non-gastric cancer non-H. pylori mucosa (mean: 0.24 vs. 0.10, P = 3 × 10). -12 ) (Fig. 6b). In HME patients, HME markers were highly methylated only in gastric cancer tissue, whereas common markers were highly methylated in both gastric cancer tissue and the surrounding mucosa (Fig. 6a). In LME patients, common markers were highly methylated in the surrounding mucosa, with the exception of a few rare subtypes, such as hepatoid adenocarcinoma and Helicobacter pylori-negative gastric cancer. These findings suggest that common markers show conserved methylation in the non-malignant gastric mucosa adjacent to gastric cancer, regardless of the epigenotype of the gastric cancer.

[0054] Example 4: DNA methylation analysis using biopsied mucosal tissues. Next, we analyzed the methylation levels of 156 common marker genes in gastric mucosal biopsies taken from subjects during health checkups. Specifically, we compared future gastric cancer patients with non-gastric cancer patients, regardless of whether they had H. pylori infection (Figure 7a). In mucosal samples from future gastric cancer patients, common markers were highly methylated, while EBV markers and HME markers showed low methylation levels. As shown in Figure 7b, the β values ​​of biopsies from future GN patients were significantly higher than those of non-gastric cancer H. pylori-infected patients (mean: 0.38 vs. 0.22, P = 8 × 10). -9 ) and non-gastric cancer non-H. pylori infected individuals (mean: 0.38 vs. 0.10, P = 5 × 10 -35 Of note, the β values ​​of biopsies from non-GN H. pylori-infected individuals were significantly higher than those from non-GN non-H. pylori-infected individuals (mean: 0.22 vs. 0.10, P = 2 × 10). -9 ), the former showed a large variability in β values, possibly reflecting the influence of lifestyle and other clinicopathological factors (see later section). To examine these differences, we analyzed methylation data from gastric biopsies taken for GCEP, a cohort study of subjects at risk for gastric cancer in Singapore. Similar to the Japanese cohort, the β values ​​of common markers in gastric biopsies from H. pylori-infected individuals who subsequently developed early gastric neoplasia (EGN) were significantly higher than those from H. pylori-infected individuals who did not develop EGN (mean: 0.45 vs. 0.32, P = 5 × 10). -9 ) (Fig. 7c). To assess whether differences in β values ​​were associated with the time to gastric cancer onset, we compared the correlation between β values ​​of biopsies from prospective gastric cancer patients and the time interval to gastric cancer onset (Fig. 7d). The correlation coefficient between these two variables was −0.42, indicating a negative correlation between methylation levels in gastric mucosal biopsies and the time to subsequent gastric cancer onset. Taken together, these results suggest that DNA methylation changes in non-neoplastic gastric mucosa predict future gastric cancer onset and are inversely correlated with the time to tumor onset.

[0055] DNA methylation is an independent risk factor for gastric cancer. Table 3 shows that DNA methylation was significantly associated with gastric cancer in a multivariate analysis adjusted for endoscopic and pathological morphological abnormalities and biopsy site (adjusted OR: 2.14, 95% CI: 1.67-2.73, P<0.001). Furthermore, DNA methylation was significantly associated with gastric cancer in a multivariate analysis adjusted for endoscopic and pathological morphological abnormalities, biopsy site, age, sex, alcohol consumption, smoking, Helicobacter pylori infection / eradication status, and family history of gastric cancer (adjusted OR: 1.96, 95% CI: 1.48-2.60). Figure 8 shows the weighted Kaplan-Meier curves showing the association between DNA methylation levels and gastric cancer. It was shown that those with low DNA methylation did not develop gastric cancer in the future, while those with high DNA methylation did develop gastric cancer in the future, and it was shown that the DNA methylation levels of 156 common marker genes enable prediction of gastric cancer risk. Therefore, the 156 common marker genes can be used as specific genes.

[0056] Unfavorable lifestyle habits promote DNA methylation changes. In biopsies of non-H. pylori-infected non-cancer patients, there was considerable variability in β values ​​among cases, highlighting the possibility that factors other than H. pylori infection may influence DNA methylation levels. Initial analyses revealed that these β values ​​were associated with lifestyle factors such as smoking (P = 0.04, Kolmogorov-Smirnov test) (Figure 7A). Lifestyle was also significantly associated with higher rates of morphological abnormalities, such as atrophy and intestinal metaplasia (data not shown). In GCEP, alcohol drinkers and smokers also tended to have higher rates of open atrophy, severe pathological atrophy, and severe intestinal metaplasia (data not shown). To examine the association between environmental and lifestyle factors and DNA methylation changes, we stratified the data by age, H. pylori infection, alcohol consumption, and smoking habits. The mean beta value of the gastric mucosa was related to age (50 years: 0.16; 51-60 years: 0.20; 60 years or older: 0.24), presence or absence of H. pylori (non-infected: 0.10; infected: 0.22), alcohol intake (non-drinkers: 0.16 non-drinkers; light drinkers; 0.19 moderate drinkers; 0.27 heavy drinkers), and smoking amount (0.16 non-smokers; 0.21 light smokers; 0.30 heavy smokers). Multivariate linear regression models showed that heavy drinking (adjusted regression coefficient: 0.10, 95% CI: 0.02-0.17, P = 0.009, heavy vs. non-drinkers) and heavy smoking (adjusted regression coefficient: 0.09, 95% CI: 0.002-0.18, P = 0.046, heavy vs. non-smokers) were associated with DNA methylation, even after adjusting for clinical factors such as age, H. pylori infection status and eradication status, and biopsy site (Table 5). GCEP confirmed a significant association between heavy drinking and DNA methylation (adjusted regression coefficient: 0.12, 95% CI: 0.01-0.23, P = 0.04, heavy vs. non-drinkers) (Table 4).

[0057] These results suggest that unfavorable lifestyle habits may increase gastric cancer risk by increasing DNA methylation levels in the gastric mucosa. Of note, repeated longitudinal sampling of gastric mucosal tissues from selected subjects undergoing H. pylori eradication therapy allowed us to assess the relationship between smoking cessation and DNA methylation changes in smokers (n = 14). In the gastric mucosa of persistent smokers, β values ​​increased over time (mean β change: +0.12), consistent with smoking enhancing aberrant DNA methylation (Fig. 7E, Fig. 5f). However, in subjects who quit smoking before the additional sampling, the increase in DNA methylation was significantly suppressed (P = 0.01, mean β change: -0.001).

[0058] The Role of H. pylori and Environmental Factors in DNA Methylation To assess whether H. pylori and lifestyle factors play an important role in driving these methylation changes, we examined DNA methylation changes in H. pylori-infected subjects compared with those without heavy drinking or smoking (Fig. 7f-h). In subjects with H. pylori infection but without heavy drinking or smoking (n = 156), 37 genes among the common markers showed an increase in β value of 0.1 or more, whereas no increase in β value was observed in heavy drinking subjects without H. pylori infection, and only six genes were observed in heavy smoking subjects without H. pylori infection. Next, we evaluated the role of lifestyle factors in promoting DNA methylation induced by H. pylori infection. We examined DNA methylation changes in H. pylori-infected subjects compared with those without heavy drinking or smoking (Fig. 7i-j). Heavy drinking subjects with H. pylori infection showed an increase in β value of 0.1 or more for 50 genes, while heavy smoking subjects with H. pylori infection showed an increase in β value of 0.1 or more for 52 genes. These results are consistent with a model in which H. pylori infection is essential for inducing methylation changes that are driven by lifestyle choices.

[0059] Next, we evaluated the correlation between H. pylori-targeted methylation and lifestyle factors (Fig. 9a-c). The initial increase in β levels due to H. pylori infection alone was not correlated with the accelerated increase in β levels due to the addition of alcohol consumption (P = 0.9, t-test) or smoking (P = 0.1, t-test). However, a positive correlation was observed between the accelerated increase in β levels due to alcohol consumption and smoking (R = 0.27, P = 7 × 10 -４ , t-test). To investigate whether alcohol consumption and smoking target similar or distinct genes for methylation, we examined the overlap of genes methylated by each factor (Fig. 9d-e). Genes whose methylation was enhanced by alcohol consumption and smoking significantly overlapped in 24 genes (P = 0.01, Fisher's exact test). On the other hand, genes whose methylation was induced by H. pylori did not significantly overlap with genes whose methylation accumulation was enhanced by lifestyle factors (P = 0.1 for alcohol consumption and P = 0.3 for smoking, Fisher's exact test). Gene ontology (GO) enrichment analysis revealed distinct GO terms for the gene sets whose methylation accumulation was enhanced by alcohol consumption and smoking (Fig. 9f). These results suggest that each lifestyle factor may target different sets of genes and pathways for methylation, albeit with partial and significant overlap.

[0060] Example 5: Analysis of Clinical Data of All Subjects in the Screening Cohort Clinical data from the screening cohort (n = 4,234) were combined with data from the GCEP cohort (n = 2,565) to investigate the relationship between environmental factors and gastric cancer incidence. Gastric cancer incidence probability was estimated using the Kaplan-Meier method and compared using the log-rank test based on the combination of H. pylori presence / absence, alcohol consumption, and smoking.

[0061] Data analysis focused on methylation analysis subjects in the screening cohort. The relationship between clinicopathological factors or DNA methylation levels and gastric cancer incidence was analyzed using data from 68 subjects who underwent methylation analysis of gastric mucosal biopsies collected during health examinations. Gastric cancer incidence probabilities were estimated using the weighted Kaplan-Meier method and compared using a robust score test to identify differences in gastric cancer incidence according to clinicopathological factors and H. pylori status (Fig. 10a). Gastric cancer incidence probabilities were also compared by the average β values ​​of common markers in all subjects (Fig. 8) and selected subjects with high-risk factors (Fig. 10b-f). Furthermore, gastric cancer incidence probabilities were compared by the β values ​​of each candidate marker in selected subjects with high-risk factors (Fig. 11(a)-(h)). The candidate markers used were NKX1-1, TMEM108, PTPN5, DBX1, KAZN, RYR2, CDH11, and KDR, for which primers could be designed by pyrosequencing. Accurate prediction of gastric cancer risk was demonstrated using DNA methylation levels of all genes. Weighted analyses were performed to account for the sampling probability of future gastric cancer patients, non-H. pylori-infected individuals, and non-H. pylori-infected individuals. To compare the predictive ability of age, sex, alcohol consumption, smoking, H. pylori status, family history, and DNA methylation (average beta value of common markers) for gastric cancer development, receiver operating characteristic (ROC) curves were constructed using a univariate logistic regression model with each factor as an explanatory variable and gastric cancer development as the outcome variable (Figure 10g). To further explore the possibility that incorporating DNA methylation together with clinical factors could improve the predictive ability of primary gastric cancer development, ROC curves were generated using a multivariate logistic regression model (Fig. 10g-h).

[0062] We investigated whether DNA methylation could be used to stratify primary GN risk beyond environmental and lifestyle factors (Fig. 10). H. pylori eradication status significantly stratified GN risk in H. pylori-infected individuals (P = 0.01, robust score test) (Fig. 10a). Of note, H. pylori-infected individuals were divided into a hypermethylated group (average β value of 156 common markers 0.25) and a hypomethylated group (P = 2 × 10 -４We observed significant differences in gastric cancer incidence when subjects were classified into groups based on their lifestyle (e.g., robust score test) (Fig. 10b). Similarly, DNA methylation levels significantly stratified the risk of gastric cancer among subjects with high-risk lifestyles, such as alcohol drinkers, smokers, those aged 60 years or older, and men (Fig. 10c-f).

[0063] To evaluate the predictive ability of DNA methylation and clinical factors for gastric cancer, receiver operating characteristic curves (ROC) were constructed using a univariate logistic regression model (Figure 10). The univariate model incorporating DNA methylation as an explanatory variable showed the highest area under the curve (AUC). To further explore the possibility that incorporating DNA methylation along with clinical factors could improve the predictive performance of gastric cancer, a multivariate logistic regression model was used to construct an ROC curve (Figure 10h). Adding lifestyle factors as explanatory variables to a logistic regression model that already included H. pylori infection and eradication status improved the AUC from 0.73 to 0.81. Furthermore, when clinical factors (age and sex) were incorporated as explanatory variables, the AUC further increased to 0.87. Furthermore, when DNA methylation was incorporated as an additional explanatory variable, the AUC increased to 0.92. These findings suggest that integrating DNA methylation data with clinical factors can improve the predictive ability of primary gastric cancer. To identify the best candidate markers for stratifying primary gastric cancer risk in each subgroup with high-risk lifestyle / environmental factors, we compared the gastric cancer incidence rates according to the β value of each common marker (n = 156) and selected three markers with the lowest P values ​​in each subgroup, such as NKX1-1 and SLC32A1. Weighted Kaplan-Meier curves showed that the DNA methylation levels of these three candidate markers were strongly associated with a higher gastric cancer incidence rate in subjects at high risk of developing GN (Figure 12).

[0064] We combined clinical data from the screening cohort (n = 4,234) with data from the GCEP cohort (n = 2,565) to examine the relationship between environmental factors (H. pylori infection, alcohol consumption, and smoking) and gastric cancer incidence. Gastric cancer incidence was estimated using the Kaplan-Meier method and compared using the log-rank test according to the combination of H. pylori infection status and alcohol consumption (Figure 12), the combination of H. pylori infection status and smoking (Figure 12), and the combination of H. pylori infection status, alcohol consumption, and smoking (Figure 12). These results indicate that unfavorable lifestyle habits promote the accumulation of DNA methylation, which synergistically accelerates the gastric cancer risk in H. pylori-infected individuals. Although primary gastric cancer risk increases as a result of the combination of multiple environmental factors, DNA methylation is considered a simple and useful quantitative marker for estimating gastric cancer risk. We also found that the target genes methylated by each environmental factor may partially overlap. However, common markers are not necessarily methylated in the same way, suggesting that the optimal markers for stratifying gastric cancer risk may differ for each subgroup classified by exposure to environmental factors. Indeed, although candidate markers for stratifying primary gastric cancer risk overlapped among subgroups, candidate markers specific to each subgroup were also identified. These results highlight the potential for integrated gastric cancer risk estimation based on lifestyle and DNA methylation.

[0065] Example 6: Identification of genes exhibiting DNA methylation levels that enable prediction of gastric cancer risk (1) Identification of genes that enable prediction of gastric cancer risk in H. pylori-infected individuals Using Infinium assay data for (iv) gastric mucosal biopsies (n=45) surrounding gastric cancer and (iii) gastric mucosal biopsies (n=9) from non-H. pylori-infected, non-gastric cancer subjects ("non-GN, non-H. pylori subjects") from the screening cohort, genes with "average beta value of gastric mucosa surrounding cancer > 0.25" and "average beta value of gastric mucosal biopsy specimens from subjects who were not H. pylori-infected and did not develop gastric cancer < 0.15" were determined. Such genes include the following: SLC6A3, GFPT2, LOC392232, CCDC37-AS1, LMX1B, ZNF790-AS1, ZSCAN23, SSTR2, FZD10, ZSCAN18, MIR3663HG, MOS, ZNF660, NAALAD2, HAGLR, CFAP100, KCNA7, ZNF549, NCAN, HS3ST2, SRD5A2, PCDH8, GHSR, and WT. H3DI, KCNK10, ZNF880, CCDC67, LVRN, COMP, NEUROG1, ZNF578, LRIF1, CADM3, EID3, GRP, KCNA1, SLC2A14, ST6GAL 2, DMRTA2, MYOD1, GALR1, ISLR2, SIAH3, IRX4, INSC, DBX1, IRX6, AARD, FAM135B, HBQ1, ADCY8, LHCGR, SLITRK5, D LX5, GFRA1, MIR124-3, PABPC4L, TBX5-AS1, MIR9-3HG, DRD5, PCDH10, TOX2, KNSTRN, POU4F3, TMEFF2, CHRM2, ZN F582, CCDC36, CADM1, NXPH2, MKX, GABRA5, TLX1, ZNF418, NKX2-2, COL4A2, OPRM1, TRPA1, FAR2P2, SLC27A6, CIDE A, CXCL12, PTPRT, PXDN, HIST1H1A, ZFHX4, GSX1, ADCYAP1, SYT6, ZFPM2, DMRT3, KHDRBS2, VSTM2B, LOC100287846 , ELMO1, NPBWR1, DKK2, ERVMER34-1, NKX1-1, C11orf85, CABP7, HELT, LIN7A, ALDH1L1-AS2, NKX2-1, VAX1, SYT9,SUSD5、CCDC8、CRHR2、C8orf34-AS1、CTNND2、LINC01535、FOXI2、PCSK2、LOC101928134、NTMT1、BOLL、RXFP3、HAPLN4、PNPLA5、ZNF492、WT1-AS、ZNF529、SERTM1、ERBB4、MDGA2、DPY19L2、IGFBP7-AS1、LOC101928796、USP44、SLCO4C1、BHLHE23、PYGO1、PLXNA4、SLC35F4、LBX2、LINC01248、CACNA1B、RNPS1、CPLX2、DCLK1、NKX2-1-AS1、RSPO3、HOXD1、PLPPR4、C1QL2、CYTL1、OPRK1、ADAMTS18、RAX、PCDHGB7、TMEM108、KAZN、TDRP、NRXN3、FENDRR、ZIC5、EDNRB、RIMS2、LINC00485、EMX2、EVC2、PEAR1、T、ZNF702P、HOXD9、CNTNAP5、CFAP161、ZNF334、ACTA1、CYYR1、COL1A2、DYDC2、FBXL7、GDNF、ELAVL2、HS3ST3A1、LOC283731、ERICH3、PCDHGA11、EFS、RALYL、EVX1、PCDHGA12、DDX25、DNAJC19、GPR139、SST、NOTO、SYN3、PCDHAC2、DOC2A、PREX2、TSPYL5、ZNF677、CBLN4、GRM2、NEUROD1、MCHR2、C11orf87、LAYN、UNCX、HECW1、PTH2R、DYDC1、SPAG6、KCNK9、HMCN2、SNAP25、KDR、KIAA1024、ARHGEF10、HMGCLL1、SYT1、C3orf14、BARHL2、GLRA3、NWD2、SNAP91、KCNH7、ZNF582-AS1、ZNF300P1、SLC32A1、PDE10A、RGS17、CCNA1、PCSK1、B3GAT1、HOXA9、HOXD10、DNAH8、FAM184B、HMX2、B3GAT2、NXPH1、TRAM1L1、BVES-AS1、DSCAM、GBX2、NTSR2、GRIN2A、EFTUD1P1、COL25A1、TUSC3、ACSS3、DSC3、NPR3、NEU1、ZIC4、SEMA6D、LINC01197、POU4F2、GUCY1B3、OLIG2, GALNT9, LOC101928174, NTM, FAM20A, GRIA4, CPXM2, HOXA10, LEPR, PTPN5, PRMT8, CDO1, LHFPL4, SHISA9, ALDH1L1, CPNE8, FAM19A4, C8orf88, CALN1, PDPN, SORCS1, CHL1, FSTL1, LYPD5, ADRA1D, C2orf40, SP9, ALK, PDZRN4, ZSCAN1, PRLHR, NRSN1, FBN2, TLX3, ZSWIM2, GPR146, ZIK1, TRHDE-AS1, HOXD 11、GRM7、ZNF365、NEFL、C10orf53、ALDH1A2、FGF10-AS1、AJAP1、RELN、KCNIP1、MIR2682、TMEM132E、NECAB1、ADCY5、GJD2、EFCC1、GIPC3、LINC01231、PLP PR3、RYR3、CTD-2297D10.2、LINC00461、TBX18、CWH43、SELV、CCDC169、NID1、KCND2、ZNF835、DPP10、LOC284395、SSTR4、TRPC4、RNF212、GRIA2、EPDR1、CS MD1、RAPGEF4、REC8、HS6ST1、AGTR1、WBSCR17、C14orf39、LRRTM1、THSD7B、LOC100289333、MIR4281、TMEM196、ROBO3、TCF15、GALNT13、GUCY1A3、ADAM33 、PTH2、CDH7、EPHA7、WNT16、NKAIN4、NGF、ZNF132、PHACTR3、MEIOB、CDH18、SLC18A3、ZNF471、ZFP42、RASGRF1、GALR2、FGF3、DOCK10、TRHDE、SYT3、VWC2、S OX21、DBX2、FOXG1、ZNF667-AS1、ADGRV1、PCDHGB6、EPHA5-AS1、USP6、ZNF382、ST8SIA2、UTF1、CDH4、CNTN4、KCNK2、NRG1、PHYHIPL、THCAT155、MYO3A、FAM 19A5、PHOX2A、GABRG3、IGF2-AS、TCERG1L、ZNF793、COL4A1、SORCS3、11-Mar、PRR35、PLXDC2、TDRD10、LRRK2、ZFP28、NRIP3、ACAN、NLGN1、THY1、AFG3L1P、A total of 420 genes were selected, including SFTA3, RAB12, TWIST2, FREM2, EBF2, FEZF2, CACNA1E, NTRK3, EMX2OS, HAND1, SNTG1, WNT7A, CCDC140, FGF14, PGR, SATB2-AS1, and GAD2. GN incidence was estimated using the weighted Kaplan-Meier method and compared using a robust score test ( Figure 13 ). It was shown that DNA methylation levels in these 420 genes can be used to predict the risk of developing gastric cancer. Therefore, the methylation level of at least one gene selected from the group consisting of the above 420 genes can be used as a marker to distinguish gastric mucosal biopsies surrounding GN from gastric mucosal biopsies of non-GN subjects not infected with H. pylori ("non-GN non-H. pylori subjects").

[0066] (2) Identification of genes that can predict the risk of developing gastric cancer in H. pylori-infected individuals according to lifestyle habits (smoking and drinking) Using Infinium assay data for (i) gastric biopsies (n=21) of subjects who subsequently developed GN ("future GN patients"), (ii) gastric biopsies (n=38) of H. pylori-infected individuals who did not subsequently develop GN ("non-GN H. pylori-infected individuals"), and (iii) gastric biopsies (n=9) of non-GN subjects not infected with H. pylori ("non-GN non-H. pylori subjects"), only methylation level information for the 420 genes selected in (1) was extracted. Three subgroups were extracted from the 68 subjects: i. H. pylori-infected individuals (n=59), ii. alcohol drinkers (n=40), and iii. smokers (n=39) (with some overlap). The following two indices were calculated for each group to determine their performance as markers: (i) A p-value was calculated using a robust score test for the beta value of each marker. (ii) An ROC curve was created using a logistic regression model with the beta value of each marker as the explanatory variable and the presence or absence of cancer onset as the outcome variable, and the AUC value was calculated. After that, for each group, "genes that satisfy p<0.05, AUC value>0.75 among the top 150 genes with the lowest p-values" were extracted, and the following marker genes were extracted.

[0067] (I) Marker genes in H. pylori infected individuals (n=149)NKX1-1, PABPC4L, SLC32A1, SYT6, TRPC4, WBSCR17, B3GAT1, ST8SIA2, RELN, DRD5, CPNE8, CACNA1B, DPP10, HOXD9, GRIA2, RXFP3, MKX, NKX2-2, LOC101928174, PREX2, GALNT13, CTD-2297D10.2, PTH2R, LYPD5, ADAMTS18, NXPH2, PCSK1, GSX1, GPR146, ZNF578, GUCY1B3, ZNF471, FAM135B, GBX2, PLPPR3, CFAP100, DBX1, NRG1, HMGCLL1, DDX25, ADRA1D, ADCY8, ACTA1, KAZN, SFTA3, ZFP42, NRIP3, NWD2, CADM3, ADCYAP1, MCHR2, HS3ST2, CYYR1, EVX1, TCERG1L, ADAM33, NLGN1, EID3, CDH4, EFCC1, ISLR2, PHYHIPL, SLITRK5, FGF3, TMEFF2, SEMA6D, HOXD10, PLXDC2, NKX2-1-AS1, ELAVL2, SYT9, AJAP1, ZSWIM2, SNAP91, GABRG3, PLPPR4, ZNF365, C14orf39, ERBB4, LMX1B, IRX4, FBN2, ACAN, EVC2, TBX5-AS1, DYDC2, HS3ST3A1, IRX6, SLC27A6, PCDH8, CNTN4, BARHL2, SNAP25, RAPGEF4, HOXD11, USP6, LOC392232, HAGLR, PRLHR, TUSC3, SERTM1, VSTM2B, KHDRBS2, FOXI2, TDRP, TDRD10, FGF14, NTM, ST6GAL2, PRMT8, CCDC36, KDR, PGR, KCNIP1, ZNF702P, GABRA5, RNPS1, NTRK3, BOLL, EDNRB, LEPR, RALYL, DSCAM, ROBO3, IGF2-AS, HOXD1, SSTR4, GALR1, POU4F2, MEIOB, INSC, ZNF382, PCSK2, DCLK1, UNCX, LINC01248, DYDC1, OPRM1, COL4A2, LINC01197, ZNF880, CPLX2, GALNT9, GRM7, TRAM1L1, GDNF, SORCS1, ZSCAN1, and EFSThe methylation levels of these genes can be used as markers that can distinguish between those who have developed gastric cancer and those who have not. The 149 genes selected here are markers that can be used in H. pylori-infected individuals, further selected from 420 genes that can be used to predict the risk of gastric cancer, and can be used to predict the risk of gastric cancer, particularly in subjects with a history of H. pylori infection.

[0068] (II) Marker genes in alcohol drinkers (n=117): NKX1-1, PABPC4L, SLC32A1, SYT6, TRPC4, WBSCR17, B3GAT1, ST8SIA2, RELN, DRD5, CPNE8, CACNA1B, DPP10, HOXD9, GRIA2, RXFP3, MKX, NKX2-2, LOC101928174, PREX2, GALNT13, CTD-2297D10.2, PTH2R, LYPD5, ADAMTS18, NXPH2, PCSK1, GSX1, GPR1 46, ZNF578, GUCY1B3, FAM135B, GBX2, PLPPR3, DBX1, NRG1, HMGCLL1, DDX25, ADRA1D, ADCY8, ACTA1, KAZN, SFTA3, ZFP42, N RIP3, NWD2, CADM3, HS3ST2, CYYR1, EVX1, TCERG1L, ADAM33, NLGN1, EID3, CDH4, EFCC1, ISLR2, PHYHIPL, SLITRK5, FGF3, TM EFF2, NKX2-1-AS1, ELAVL2, SYT9, AJAP1, ZSWIM2, SNAP91, GABRG3, PLPPR4, ZNF365, C14orf39, ERBB4, LMX1B, IRX4, ACAN , EVC2, TBX5-AS1, DYDC2, SLC27A6, PCDH8, CNTN4, LOC392232, HAGLR, PRLHR, TUSC3, SERTM1, KHDRBS2, FOXI2, TDRP, FGF14 , NTM, ST6GAL2, PRMT8, CCDC36, KDR, PGR, KCNIP1, ZNF702P, BOLL, EDNRB, RALYL, FOXG1, SLC2A14, PCDHGB7, GRM2, CTNND2, LINC00485, ERVMER34-1, CCDC37-AS1, FAM19A4, TBX18, MIR9-3HG, LOC100289333, LINC00461, TRHDE, FZD10, and FAM19A5. Methylation levels in these genes can be used as more sensitive markers to distinguish between gastric cancer and non-cancer in alcohol drinkers.The 117 genes selected here are markers that can be used in alcohol drinkers, selected from 420 genes that can be used to predict the risk of developing stomach cancer, and can be used to predict the risk of developing stomach cancer, particularly in subjects with a drinking habit.

[0069] (III) Marker genes in smokers (n=142) NKX1-1, PABPC4L, SLC32A1, SYT6, TRPC4, WBSCR17, B3GAT1, ST8SIA2, RELN, DRD5, CPNE8, CACNA1B, DPP10, HOXD9, GRIA2, RXFP3, MKX, NKX2-2, LOC101928174, PREX2, GALNT13, CTD-2297D10.2, PTH2R, LYPD5, ADAMTS18, NXPH2, PCSK1, GSX1, GPR146, ZNF578, GUCY1B3, FAM135B, GBX2, PLPPR3, DBX1, NRG1, HMGCLL1, DDX25, ADRA1D, ADCY8, ACTA1, KAZN, SFTA3, ZFP42, NRIP3, NWD2, CADM3, HS3ST2, CYYR1, EVX1, TCERG1L, ADAM33, NLGN1, EID3, EFCC1 , SLITRK5, FGF3, TMEFF2, NKX2-1-AS1, ELAVL2, SYT9, AJAP1, ZSWIM2, SNAP91, GABRG3, PLPPR4, ZNF365, C14orf39, ERBB4, LMX1B, IRX4, ACAN, TBX5- AS1, DYDC2, SLC27A6, HAGLR, PRLHR, SERTM1, KHDRBS2, TDRP, FGF14, NTM, PRMT8, CCDC36, KDR, ZNF702P, EDNRB, PCDHGB7, LINC00485, ERVMER34-1, C CDC37-AS1, TBX18, MIR9-3HG, LOC100289333, LINC00461, FZD10, CFAP100, MCHR2, PLXDC2, HS3ST3A1, BARHL2, SNAP25, RAPGEF4, NTRK3, DSCAM, ROB O3, IGF2-AS, GALR1, INSC, ZNF382, PCSK2, DCLK1, UNCX, DYDC1, OPRM1, COL4A2, GRM7, SORCS1, SLC18A3, SPAG6, PCDHGB6, ADCY5, ZNF529, THSD7B, CO L25A1, PCDHGA11, NPBWR1, VAX1, WTH3DI, ZSCAN18, PCDHAC2, DMRTA2, KCNH7, MOS, TLX3, LAYN, SORCS3, NKAIN4, KIAA1024, CCDC140, NRXN3, and EMX2OSThe methylation levels of these genes can be used as more sensitive markers to distinguish between those who develop gastric cancer and those who do not in smokers. The 142 genes selected here are markers that can be used in smokers, further selected from 420 genes that can be used to predict the risk of gastric cancer, and can be used to predict the risk of gastric cancer, particularly in subjects with a smoking habit.

[0070] (IV) Marker genes overlapping all of (I) to (III) (n = 87) NKX1-1, PABPC4L, SLC32A1, SYT6, TRPC4, WBSCR17, B3GAT1, ST8SIA2, RELN, DRD5, CPNE8, CACNA1B, DPP10, HOXD9, GRIA2, RXFP3, MKX, NKX2-2, LOC101928174, PREX2, GALNT13, C TD-2297D10.2, PTH2R, LYPD5, ADAMTS18, NXPH2, PCSK1, GSX1, GPR146, ZNF578, GUCY1B3, FAM135B, GBX2, PLPPR3, DBX1, NRG1, HMGCLL1, DDX25, ADRA1D, ADCY8, ACTA1, KAZN, SFT A3, ZFP42, NRIP3, NWD2, CADM3, HS3ST2, CYYR1, EVX1, TCERG1L, ADAM33, NLGN1, EID3, EFCC1, SLITRK5, FGF3, TMEFF2, NKX2-1-AS1, ELAVL2, SYT9, AJAP1, ZSWIM2, SNAP91, GABRG3, PLPPR4, ZNF365, C14orf39, ERBB4, LMX1B, IRX4, ACAN, TBX5-AS1, DYDC2, SLC27A6, HAGLR, PRLHR, SERTM1, KHDRBS2, TDRP, FGF14, NTM, PRMT8, CCDC36, KDR, ZNF702P, and EDNRB. The above genes are overlapping marker genes in all of (I) to (III). The methylation levels of these marker genes can be used as markers that can be used more accurately, regardless of the lifestyle of the subject.

[0071] 10 Gastric cancer risk determination device 11 Input unit 12 Memory unit 13 Processing unit 14 Output unit 15 Learning unit 20 System for determining gastric cancer risk 30 Detection device for detecting DNA methylation level

Claims

1. A method for predicting a risk of developing gastric cancer in a subject, comprising detecting any of the following in a gastric biopsy sample obtained from the subject: SLC6A3, GFPT2, LOC392232, CCDC37-AS1, LMX1B, ZNF790-AS1, ZSCAN23, SSTR2, FZD10, ZSCAN18, MIR3663HG, MOS, ZNF660, NAALAD2, HAGLR, CFAP100, KCNA7, ZNF549, NCAN, HS3ST2, SRD5A2, PCDH8, GHSR, WTH3DI, KCNK10, ZNF880, CCDC67, LVRN, COMP, NEUROG1, ZNF578, LRIF1, CADM3, EID3, GRP, KCNA1, SLC2A14, ST6GAL2, DMRTA2, MYOD1, GALR1, ISLR2, SIAH3, IRX4, INSC, DBX1, IRX6, AARD, FAM135B, HBQ1, ADCY8, LHCGR, SLITRK5, DL X5, GFRA1, MIR124-3, PABPC4L, TBX5-AS1, MIR9-3HG, DRD5, PCDH10, TOX2, KNSTRN, POU4F3, TMEFF2, CHRM2, ZNF582, CCDC36, CADM1, NXPH2, MKX, GABRA5, TLX1, ZNF418, NKX2-2, COL4A2, OPRM1, TRPA1, FAR2P2, SLC27A6, CIDEA, CXCL12, PTPRT, PXDN, HIST1H1A, ZFHX4, GSX1, ADCYAP1, SYT6, ZFPM2, DMRT3, KHDRBS2, VSTM2B, LOC100287846, ELMO1, NPBWR1, DKK2, ERVMER34-1, NKX1-1, C11orf85, CABP7, HELT, LIN7A, ALDH1L1-AS2, NKX2 -1, VAX1, SYT9, SUSD5, CCDC8, CRHR2, C8orf34-AS1, CTNND2, LINC01535, FOXI2, PCSK2, LOC101928134, NTMT1, BOLL, RXFP3, HAPLN4, PNPLA5, ZN F492, WT1-AS, ZNF529, SERTM1, ERBB4, MDGA2, DPY19L2, IGFBP7-AS1, LOC101928796, USP44, SLCO4C1, BHLHE23, PYGO1, PLXNA4, SLC35F4, LBX2,<h2 style=";text-align:left;direction:ltr">LINC01248、CACNA1B、RNPS1、CPLX2、DCLK1、NKX2-1-AS1、RSPO3、HOXD1、PLP PR4、C1QL2、CYTL1、OPRK1、ADAMTS18、RAX、PCDHGB7、TMEM108、KAZN、TDRP、NR Xn3 、 Fendrr 、 zic5 、 Ednrb 、 rims2 、 LINC00485 、 EMX2 、 EVC2 、 PEAR1 、 T 、 znf702 P、HOXD9、CNTNAP5、CFAP161、ZNF334、ACTA1、CYYR1、COL1A2、DYDC2、FBXL7、G DNF、ELAVL2、HS3ST3A1、LOC283731、ERICH3、PCDHGA11、EFS、RALYL、EVX1、P CDHGA12、DDX25、DNAJC19、GPR139、SST、NOTO、SYN3、PCDHAC2、DOC2A、PREX2、 TSPYL5、ZNF677、CBLN4、GRM2、NEUROD1、MCHR2、C11orf87、LAYN、UNCX、HECW 1、PTH2R、DYDC1、SPAG6、KCNK9、HMCN2、SNAP25、KDR、KIAA1024、ARHGEF10、HM GCLL1、SYT1、C3orf14、BARHL2、GLRA3、NWD2、SNAP91、KCNH7、ZNF582-AS1、Z NF300P1、SLC32A1、PDE10A、RGS17、CCNA1、PCSK1、B3GAT1、HOXA9、HOXD10、DN AH8、FAM184B、HMX2、B3GAT2、NXPH1、TRAM1L1、BVES-AS1、DSCAM、GBX2、NTSR 2、GRIN2A、EFTUD1P1、COL25A1、TUSC3、ACSS3、DSC3、NPR3、NEU1、ZIC4、SEMA6 D、LINC01197、POU4F2、GUCY1B3、OLIG2、GALNT9、LOC101928174、NTM、FAM20 A、GRIA4、CPXM2、HOXA10、LEPR、PTPN5、PRMT8、CDO1、LHFPL4、SHISA9、ALDH1L 1、CPNE8、FAM19A4、C8orf88、CALN1、PDPN、SORCS1、CHL1、FSTL1、LYPD5、ADR A1D、C2orf40、SP9、ALK、PDZRN4、ZSCAN1、PRLHR、NRSN1、FBN2、TLX3、ZSWIM2、GPR146, ZIK1, TRHDE-AS1, HOXD11, GRM7, ZNF365, NEFL, C10orf53, ALDH1A2, FGF10-AS1, AJAP1, RELN, KCNIP1, MIR2682, TMEM132E, NECAB1, ADCY5, GJD2, EFCC1, GIPC3, LINC01231, PLPPR3, RYR3, CTD-2297D10.2, LINC00461, TBX18, CWH43, SELV, CCDC169, NID1, KCND2, ZNF835, DPP1 0, LOC284395, SSTR4, TRPC4, RNF212, GRIA2, EPDR1, CSMD1, RAPGEF4, REC8, HS6ST1, AGTR1, WBSCR17, C14orf39, LRRTM1, THSD7B, LOC1002893 33, MIR4281, TMEM196, ROBO3, TCF15, GALNT13, GUCY1A3, ADAM33, PTH2, CDH7, EPHA7, WNT16, NKAIN4, NGF, ZNF132, PHACTR3, MEIOB, CDH18, SL C18A3, ZNF471, ZFP42, RASGRF1, GALR2, FGF3, DOCK10, TRHDE, SYT3, VWC2, SOX21, DBX2, FOXG1, ZNF667-AS1, ADGRV1, PCDHGB6, EPHA5-AS1, U SP6, ZNF382, ST8SIA2, UTF1, CDH4, CNTN4, KCNK2, NRG1, PHYHIPL, THCAT155, MYO3A, FAM19A5, PHOX2A, GABRG3, IGF2-AS, TCERG1L, ZNF793, CO detecting a DNA methylation level of at least one gene selected from the group consisting of L4A1, SORCS3, 11-Mar, PRR35, PLXDC2, TDRD10, LRRK2, ZFP28, NRIP3, ACAN, NLGN1, THY1, AFG3L1P, SFTA3, RAB12, TWIST2, FREM2, EBF2, FEZF2, CACNA1E, NTRK3, EMX2OS, HAND1, SNTG1, WNT7A, CCDC140, FGF14, PGR, SATB2-AS1, and GAD2; and comparing the detected DNA methylation level with at least one threshold value of the methylation level corresponding to the risk of developing gastric cancer to predict the risk of developing gastric cancer.The method.

2. The method of claim 1, wherein the DNA methylation level of the gene is the methylation level in a gene promoter CpG island.

3. The method according to claim 1, wherein the gene is a candidate gastric cancer risk marker gene based on a cohort analysis.

4. The method of claim 1, wherein the gastric biopsy sample is non-cancerous tissue.

5. The method of claim 1, wherein a high methylation level of the gene in normal tissue indicates a high risk of developing gastric cancer in the future, while a low methylation level indicates a low risk of developing gastric cancer in the future.

6. The cohort analyzed candidate gastric cancer risk marker genes are as follows: NKX1-1, PABPC4L, SLC32A1, SYT6, TRPC4, WBSCR17, B3GAT1, ST8SIA2, RELN, DRD5, CPNE8, CACNA1B, DPP10, HOXD9, GRIA2, RXFP3, MKX, NKX2-2, LOC101928174, PREX2, GALNT13, CTD-2297D10.2, PTH2R, LYPD5, ADAMTS18, NXPH2, PCSK1, GSX1, GPR146, ZNF578, GUCY1B3, ZNF471, FAM135B, GBX2, PLPPR3, CFAP100, DBX1, NRG1, HMGCLL1, DDX25, ADRA1D, ADCY8, ACTA1, KAZN, SFTA3, ZFP42, NRIP3, NWD2, CADM3, ADCYAP1, MCHR2, HS3ST2, CYYR1, EVX1, TCERG1L, A DAM33, NLGN1, EID3, CDH4, EFCC1, ISLR2, PHYHIPL, SLITRK5, FGF3, TMEFF2, SEMA6D, HOXD10, PLXDC2, NKX2-1-AS1, ELAVL2, SYT9, A JAP1, ZSWIM2, SNAP91, GABRG3, PLPPR4, ZNF365, C14orf39, ERBB4, LMX1B, IRX4, FBN2, ACAN, EVC2, TBX5-AS1, DYDC2, HS3ST3A1, IRX 6, SLC27A6, PCDH8, CNTN4, BARHL2, SNAP25, RAPGEF4, HOXD11, USP6, LOC392232, HAGLR, PRLHR, TUSC3, SERTM1, VSTM2B, KHDRBS2, F OXI2, TDRP, TDRD10, FGF14, NTM, ST6GAL2, PRMT8, CCDC36, KDR, PGR, KCNIP1, ZNF702P, GABRA5, RNPS1, NTRK3, BOLL, EDNRB, LEPR, RA The method according to claim 3, wherein the gene is at least one selected from the group consisting of LYL, DSCAM, ROBO3, IGF2-AS, HOXD1, SSTR4, GALR1, POU4F2, MEIOB, INSC, ZNF382, PCSK2, DCLK1, UNCX, LINC01248, DYDC1, OPRM1, COL4A2, LINC01197, ZNF880, CPLX2, GALNT9, GRM7, TRAM1L1, GDNF, SORCS1, ZSCAN1, and EFS.

7. The cohort analyzed candidate gastric cancer risk marker genes are the following: NKX1-1, PABPC4L, SLC32A1, SYT6, TRPC4, WBSCR17, B3GAT1, ST8SIA2, RELN, DRD5, CPNE8, CACNA1B, DPP10, HOXD9, GRIA2, RXFP3, MKX, NKX2-2, LOC101928174, PREX2, GALNT13, CTD-2297D10.2, PTH2R, LYPD5, ADAMTS18, NXPH2, PCSK1, GSX1, GPR146, ZNF578 , GUCY1B3, FAM135B, GBX2, PLPPR3, DBX1, NRG1, HMGCLL1, DDX25, ADRA1D, ADCY8, ACTA1, KAZN, SFTA3, ZFP42, NRIP3, NWD2, CADM 3, HS3ST2, CYYR1, EVX1, TCERG1L, ADAM33, NLGN1, EID3, CDH4, EFCC1, ISLR2, PHYHIPL, SLITRK5, FGF3, TMEFF2, NKX2-1-AS1, EL AVL2, SYT9, AJAP1, ZSWIM2, SNAP91, GABRG3, PLPPR4, ZNF365, C14orf39, ERBB4, LMX1B, IRX4, ACAN, EVC2, TBX5-AS1, DYDC2, SL C27A6, PCDH8, CNTN4, LOC392232, HAGLR, PRLHR, TUSC3, SERTM1, KHDRBS2, FOXI2, TDRP, FGF14, NTM, ST6GAL2, PRMT8, CCDC36, K The method of claim 3, wherein the gene is at least one selected from the group consisting of DR, PGR, KCNIP1, ZNF702P, BOLL, EDNRB, RALYL, FOXG1, SLC2A14, PCDHGB7, GRM2, CTNND2, LINC00485, ERVMER34-1, CCDC37-AS1, FAM19A4, TBX18, MIR9-3HG, LOC100289333, LINC00461, TRHDE, FZD10, and FAM19A5.

8. The following gastric cancer risk marker genes were analyzed in the cohort: NKX1-1, PABPC4L, SLC32A1, SYT6, TRPC4, WBSCR17, B3GAT1, ST8SIA2, RELN, DRD5, CPNE8, CACNA1B, DPP10, HOXD9, GRIA2, RXFP3, MKX, NKX2-2, LOC101928174, PREX2, GALNT13, CTD-2297D10.2, PTH2R, LYPD5, ADAMTS18, NXPH2, PCSK1, GSX1, GPR146, ZNF578, GUCY1B3, FAM135B, GBX2, PLPPR3, DBX1, NRG1, HMGCLL1, DDX25, ADRA1D, ADCY8, ACTA1, KAZN, SFTA3, ZFP42, NRIP3, NWD2, CADM3, HS3ST2, CYYR1, EVX1, TCERG1L, ADAM33, NLGN1, EID3, EFCC1, SLI TRK5, FGF3, TMEFF2, NKX2-1-AS1, ELAVL2, SYT9, AJAP1, ZSWIM2, SNAP91, GABRG3, PLPPR4, ZNF365, C14orf39, ERBB4, LMX1B, IRX4 , ACAN, TBX5-AS1, DYDC2, SLC27A6, HAGLR, PRLHR, SERTM1, KHDRBS2, TDRP, FGF14, NTM, PRMT8, CCDC36, KDR, ZNF702P, EDNRB, PCDHG B7, LINC00485, ERVMER34-1, CCDC37-AS1, TBX18, MIR9-3HG, LOC100289333, LINC00461, FZD10, CFAP100, MCHR2, PLXDC2, HS3ST3 A1, BARHL2, SNAP25, RAPGEF4, NTRK3, DSCAM, ROBO3, IGF2-AS, GALR1, INSC, ZNF382, PCSK2, DCLK1, UNCX, DYDC1, OPRM1, COL4A2, G The method according to claim 3, wherein the gene is at least one selected from the group consisting of RM7, SORCS1, SLC18A3, SPAG6, PCDHGB6, ADCY5, ZNF529, THSD7B, COL25A1, PCDHGA11, NPBWR1, VAX1, WTH3DI, ZSCAN18, PCDHAC2, DMRTA2, KCNH7, MOS, TLX3, LAYN, SORCS3, NKAIN4, KIAA1024, CCDC140, NRXN3, and EMX2OS.

9. The cohort analyzed candidate gastric cancer risk marker genes are the following: NKX1-1, PABPC4L, SLC32A1, SYT6, TRPC4, WBSCR17, B3GAT1, ST8SIA2, RELN, DRD5, CPNE8, CACNA1B, DPP10, HOXD9, GRIA2, RXFP3, MKX, NKX2-2, LOC101928174, PREX2, GALNT13, CTD-2297D10.2, PTH2R, LYPD5, ADAMTS18, NXPH2, PCSK1, GSX1, GPR146, ZNF578, GUCY1B3, FAM135B, GBX2, PLPPR3, DBX1, NRG1, HMGCLL1, DDX25, ADRA1D, ADCY 8, ACTA1, KAZN, SFTA3, ZFP42, NRIP3, NWD2, CADM3, HS3ST2, CYYR1, EVX1, TCERG1L, ADAM33, NLGN1, EID3, EFCC1, SLITRK5, FGF3, TMEFF2, NKX2-1-AS1, ELAVL2, SYT9, AJAP1, ZSWIM2, SNAP91, GABRG3, PLPPR4, ZNF365, C14orf39, ERBB4, LMX1B, IRX4, ACAN, TBX5-AS1, DYDC2, SLC27A6, HAGLR, PRLHR, SERTM1, KHDRBS2, TDRP, FGF14, NTM, PRMT8, CCDC36, KDR, ZNF702P, and EDNRB.

10. The cohort analyzed candidate gastric cancer risk marker genes are the following: NKX1-1, NKX2-2, DRD5, PABPC4L, ADAMTS18, FAM135B, ZNF578, PTH2R, SLC32A1, HS3ST2, B3GAT1, SLITRK5, SYT6, RXFP3, KHDRBS2, PCSK1, COL4A2, DBX1, EDNRB, ZNF365, HOXD9, MKX, GABRG3, AC The method of claim 3, wherein the gene is at least one selected from the group consisting of TA1, SERTM1, FGF14, TRPC4, ZSWIM2, FZD10, PCDHAC2, CACNA1B, PLPPR4, ADRA1D, RELN, GRIA2, DPP10, HMGCLL1, LOC101928174, TBX18, PREX2, CNTN4, CCDC36, WBSCR17, ADCY8, and C14orf39.

11. The method of claim 3, wherein the cohort analyzed candidate gastric cancer risk marker gene is at least one gene selected from the group consisting of PABPC4L, DRD5, NKX1-1, NKX2-2, HS3ST2, SLITRK5, ZNF578, RXFP3, ADAMTS18, SYT6, B3GAT1, FAM135B, PTH2R, ZNF365, ACTA1, PCDHAC2, DBX1, MKX, FZD10, TRPC4, SLC32A1, CACNA1B, COL4A2, PCSK1, KHDRBS2, LOC101928174, DPP10, RELN, CCDC36, WBSCR17, PREX2, PLPPR4, and CNTN4.

12. The cohort analyzed gastric cancer risk marker candidate genes are the following: NKX1-1, SLC32A1, CACNA1B, RXFP3, PABPC4L, PTH2R, B3GAT1, FAM135B, SYT6, ZNF578, ADAMTS18, ADRA1D, RELN, DBX1, DRD5, LOC101928174, DPP10, HS3ST2, NKX2-2, PLPPR4, KHDRBS2, PCSK1, SLITRK5, ZNF365, WBSCR17, MKX, CTD-2297D10.2, PREX2, NXPH2, ADCY8, NRG1, ZFP42, HOXD9, TRPC4, CNTN4, LYPD5, CYYR1, GABRG3, PLPPR3, TCERG1L, C14orf39, ACTA1, ST6GAL2, KAZN, HMGCLL1, SERTM1, ACAN, ZSWIM2, EDNRB, SORCS 1, CCDC36, COL4A2, FGF14, MIR9-3HG, FZD10, PCDHGB6, WTH3DI, NKAIN4, TBX18, FOXG1, NKX2-1-AS1, EID3, CADM3, T MEFF2, ZNF702P, PCDHAC2, TRHDE, MOS, LOC100289333, GRIA2, GALNT13, GBX2, ST8SIA2, DDX25, CDH4, AJAP1, LMX1B 4. The method of claim 3, wherein the gene is at least one selected from the group consisting of FOXI2, PRLHR, SFTA3, TBX5-AS1, HAGLR, ERBB4, UNCX, SLC27A6, EFS, NLGN1, GALR1, ZSCAN1, IRX4, VAX1, FAM19A4, SOX21, COL25A1, PHOX2A, CNTNAP5, PXDN, C8orf34-AS1, WNT16, and HBQ1.

13. The method of claim 3, wherein the cohort analyzed candidate gastric cancer risk marker gene is at least one gene selected from the group consisting of NKX1-1, PABPC4L, DRD5, RXFP3, PTH2R, ZNF578, ADAMTS18, HS3ST2, NKX2-2, SLITRK5, ZNF365, SLC32A1, CACNA1B, WBSCR17, B3GAT1, FAM135B, SYT6, PREX2, ADRA1D, CNTN4, PLPPR4, and CCDC36.

14. The method according to any one of claims 1 to 13, wherein the DNA methylation level is determined using a technique selected from the group consisting of DNA methylation array, next generation sequencing, and pyrosequencing.

15. A method for predicting a risk of developing gastric cancer, comprising the steps of: detecting the DNA methylation level of at least one gene selected from the group consisting of NKX1-1, TMEM108, PTPN5, DBX1, KAZN, RYR2, CDH11, and KDR by pyrosequencing in a gastric biopsy sample obtained from the subject; and comparing the detected DNA methylation level with at least one threshold value of the methylation level corresponding to the risk of developing gastric cancer to predict the risk of developing gastric cancer.

16. A system for determining a risk of developing gastric cancer in a subject, comprising detecting any of the following in a gastric biopsy sample obtained from the subject: SLC6A3, GFPT2, LOC392232, CCDC37-AS1, LMX1B, ZNF790-AS1, ZSCAN23, SSTR2, FZD10, ZSCAN18, MIR3663HG, MOS, ZNF660, NAALAD2, HAGLR, CFAP100, KCNA7, ZNF549, NCAN, HS3ST2, SRD5A2, PCDH8, GHSR, WTH3DI, KCNK10, ZNF880, CCDC67, LVRN, COMP, NEUROG1, ZNF578, LRIF1, CADM3, EID3 , GRP, KCNA1, SLC2A14, ST6GAL2, DMRTA2, MYOD1, GALR1, ISLR2, SIAH3, IRX4, INSC, DBX1, IRX6, AARD, FAM135B, HBQ1, ADCY8, LHCGR, SLITRK5, D LX5, GFRA1, MIR124-3, PABPC4L, TBX5-AS1, MIR9-3HG, DRD5, PCDH10, TOX2, KNSTRN, POU4F3, TMEFF2, CHRM2, ZNF582, CCDC36, CADM1, NXPH2, MK X, GABRA5, TLX1, ZNF418, NKX2-2, COL4A2, OPRM1, TRPA1, FAR2P2, SLC27A6, CIDEA, CXCL12, PTPRT, PXDN, HIST1H1A, ZFHX4, GSX1, ADCYAP1, SYT 6, ZFPM2, DMRT3, KHDRBS2, VSTM2B, LOC100287846, ELMO1, NPBWR1, DKK2, ERVMER34-1, NKX1-1, C11orf85, CABP7, HELT, LIN7A, ALDH1L1-AS2, N KX2-1, VAX1, SYT9, SUSD5, CCDC8, CRHR2, C8orf34-AS1, CTNND2, LINC01535, FOXI2, PCSK2, LOC101928134, NTMT1, BOLL, RXFP3, HAPLN4, PNPLA 5, ZNF492, WT1-AS, ZNF529, SERTM1, ERBB4, MDGA2, DPY19L2, IGFBP7-AS1, LOC101928796, USP44, SLCO4C1, BHLHE23, PYGO1, PLXNA4, SLC35F4,<h2 style=";text-align:left;direction:ltr">LBX2、LINC01248、CACNA1B、RNPS1、CPLX2、DCLK1、NKX2-1-AS1、RSPO3、HOXD 1、PLPPR4、C1QL2、CYTL1、OPRK1、ADAMTS18、RAX、PCDHGB7、TMEM108、KAZN、T DRP、NRXN3、FENDRR、ZIC5、EDNRB、RIMS2、LINC00485、EMX2、EVC2、PEAR1、T、 ZNF702P、HOXD9、CNTNAP5、CFAP161、ZNF334、ACTA1、CYYR1、COL1A2、DYDC2、F BXL7、GDNF、ELAVL2、HS3ST3A1、LOC283731、ERICH3、PCDHGA11、EFS、RALYL、 EVX1、PCDHGA12、DDX25、DNAJC19、GPR139、SST、NOTO、SYN3、PCDHAC2、DOC2A、 PREX2、TSPYL5、ZNF677、CBLN4、GRM2、NEUROD1、MCHR2、C11orf87、LAYN、UNC X,HECW1,PTH2R,DYDC1,SPAG6,KCNK9,HMCN2,SNAP25,KDR,KIAA1024,ARHGE F10、HMGCLL1、SYT1、C3orf14、BARHL2、GLRA3、NWD2、SNAP91、KCNH7、ZNF582 -AS1、ZNF300P1、SLC32A1、PDE10A、RGS17、CCNA1、PCSK1、B3GAT1、HOXA9、HO XD10, DNAH8, FAM184B, HMX2, B3GAT2, NXPH1, TRAM1L1, BVES-AS1, DSCAM, GB X2 、 ntsr2 、 Grin2a 、 Eftud1p1 、 col25a1 、 tusc3 、 acss3 、 dsc3 、 nPR3 、 neu1 、 zic 4、SEMA6D、LINC01197、POU4F2、GUCY1B3、OLIG2、GALNT9、LOC101928174、NT M、FAM20A、GRIA4、CPXM2、HOXA10、LEPR、PTPN5、PRMT8、CDO1、LHFPL4、SHISA9 、ALDH1L1、CPNE8、FAM19A4、C8orf88、CALN1、PDPN、SORCS1、CHL1、FSTL1、LY PD5、ADRA1D、C2orf40、SP9、ALK、PDZRN4、ZSCAN1、PRLHR、NRSN1、FBN2、TLX3、ZSWIM2, GPR146, ZIK1, TRHDE-AS1, HOXD11, GRM7, ZNF365, NEFL, C10orf53, ALDH1A2, FGF10-AS1, AJAP1, RELN, KCNIP1, MIR2682, TMEM132E, NEC AB1, ADCY5, GJD2, EFCC1, GIPC3, LINC01231, PLPPR3, RYR3, CTD-2297D10.2, LINC00461, TBX18, CWH43, SELV, CCDC169, NID1, KCND2, ZNF835, DPP 10, LOC284395, SSTR4, TRPC4, RNF212, GRIA2, EPDR1, CSMD1, RAPGEF4, REC8, HS6ST1, AGTR1, WBSCR17, C14orf39, LRRTM1, THSD7B, LOC100289333 , MIR4281, TMEM196, ROBO3, TCF15, GALNT13, GUCY1A3, ADAM33, PTH2, CDH7, EPHA7, WNT16, NKAIN4, NGF, ZNF132, PHACTR3, MEIOB, CDH18, SLC18A3 , ZNF471, ZFP42, RASGRF1, GALR2, FGF3, DOCK10, TRHDE, SYT3, VWC2, SOX21, DBX2, FOXG1, ZNF667-AS1, ADGRV1, PCDHGB6, EPHA5-AS1, USP6, ZNF38 2, ST8SIA2, UTF1, CDH4, CNTN4, KCNK2, NRG1, PHYHIPL, THCAT155, MYO3A, FAM19A5, PHOX2A, GABRG3, IGF2-AS, TCERG1L, ZNF793, COL4A1, SORCS3, a detection device for detecting a DNA methylation level of at least one gene selected from the group consisting of 11-Mar, PRR35, PLXDC2, TDRD10, LRRK2, ZFP28, NRIP3, ACAN, NLGN1, THY1, AFG3L1P, SFTA3, RAB12, TWIST2, FREM2, EBF2, FEZF2, CACNA1E, NTRK3, EMX2OS, HAND1, SNTG1, WNT7A, CCDC140, FGF14, PGR, SATB2-AS1, and GAD2; and a device for determining a risk of developing gastric cancer, wherein the device for determining the risk of developing gastric cancer comprises:The system includes the following: an input unit for inputting the DNA methylation level obtained from the detection device; a memory unit for storing at least one threshold value of the methylation level corresponding to the risk of developing stomach cancer; a processing unit for comparing the DNA methylation level input from the input unit with the at least one threshold value retrieved from the memory unit to determine the risk of developing stomach cancer of the subject; and an output unit for outputting the determined risk of developing stomach cancer of the subject.

17. The system according to claim 16, wherein the memory unit stores control DNA methylation levels according to at least one risk factor selected from the group consisting of smoking habits, drinking habits, and age.

18. A system for determining a risk of developing gastric cancer in a subject, comprising detecting any of the following in a gastric biopsy sample obtained from the subject: SLC6A3, GFPT2, LOC392232, CCDC37-AS1, LMX1B, ZNF790-AS1, ZSCAN23, SSTR2, FZD10, ZSCAN18, MIR3663HG, MOS, ZNF660, NAALAD2, HAGLR, CFAP100, KCNA7, ZNF549, NCAN, HS3ST2, SRD5A2, PCDH8, GHSR, WTH3DI, KCNK10, ZNF880, CCDC67, LVRN, COMP, NEUROG1, ZNF578, LRIF1, CADM3, EID3, GRP, KCNA1, SLC2A14, ST6GAL2, DMRTA2, MYOD1, GALR1, ISLR2, SIAH3, IRX4, INSC, DBX1, IRX6, AARD, FAM135B, HBQ1, ADCY8, LHCGR, SLITRK5, DL X5, GFRA1, MIR124-3, PABPC4L, TBX5-AS1, MIR9-3HG, DRD5, PCDH10, TOX2, KNSTRN, POU4F3, TMEFF2, CHRM2, ZNF582, CCDC36, CADM1, NXPH2, MKX, GABRA5, TLX1, ZNF418, NKX2-2, COL4A2, OPRM1, TRPA1, FAR2P2, SLC27A6, CIDEA, CXCL12, PTPRT, PXDN, HIST1H1A, ZFHX4, GSX1, ADCYAP1, SYT6, ZFPM2, DMRT3, KHDRBS2, VSTM2B, LOC100287846, ELMO1, NPBWR1, DKK2, ERVMER34-1, NKX1-1, C11orf85, CABP7, HELT, LIN7A, ALDH1L1-AS2, NKX2 -1, VAX1, SYT9, SUSD5, CCDC8, CRHR2, C8orf34-AS1, CTNND2, LINC01535, FOXI2, PCSK2, LOC101928134, NTMT1, BOLL, RXFP3, HAPLN4, PNPLA5, ZN F492, WT1-AS, ZNF529, SERTM1, ERBB4, MDGA2, DPY19L2, IGFBP7-AS1, LOC101928796, USP44, SLCO4C1, BHLHE23, PYGO1, PLXNA4, SLC35F4, LBX2,<h2 style=";text-align:left;direction:ltr">LINC01248、CACNA1B、RNPS1、CPLX2、DCLK1、NKX2-1-AS1、RSPO3、HOXD1、PLP PR4、C1QL2、CYTL1、OPRK1、ADAMTS18、RAX、PCDHGB7、TMEM108、KAZN、TDRP、NR Xn3 、 Fendrr 、 zic5 、 Ednrb 、 rims2 、 LINC00485 、 EMX2 、 EVC2 、 PEAR1 、 T 、 znf702 P、HOXD9、CNTNAP5、CFAP161、ZNF334、ACTA1、CYYR1、COL1A2、DYDC2、FBXL7、G DNF、ELAVL2、HS3ST3A1、LOC283731、ERICH3、PCDHGA11、EFS、RALYL、EVX1、P CDHGA12、DDX25、DNAJC19、GPR139、SST、NOTO、SYN3、PCDHAC2、DOC2A、PREX2、 TSPYL5、ZNF677、CBLN4、GRM2、NEUROD1、MCHR2、C11orf87、LAYN、UNCX、HECW 1、PTH2R、DYDC1、SPAG6、KCNK9、HMCN2、SNAP25、KDR、KIAA1024、ARHGEF10、HM GCLL1、SYT1、C3orf14、BARHL2、GLRA3、NWD2、SNAP91、KCNH7、ZNF582-AS1、Z NF300P1、SLC32A1、PDE10A、RGS17、CCNA1、PCSK1、B3GAT1、HOXA9、HOXD10、DN AH8、FAM184B、HMX2、B3GAT2、NXPH1、TRAM1L1、BVES-AS1、DSCAM、GBX2、NTSR 2、GRIN2A、EFTUD1P1、COL25A1、TUSC3、ACSS3、DSC3、NPR3、NEU1、ZIC4、SEMA6 D、LINC01197、POU4F2、GUCY1B3、OLIG2、GALNT9、LOC101928174、NTM、FAM20 A、GRIA4、CPXM2、HOXA10、LEPR、PTPN5、PRMT8、CDO1、LHFPL4、SHISA9、ALDH1L 1、CPNE8、FAM19A4、C8orf88、CALN1、PDPN、SORCS1、CHL1、FSTL1、LYPD5、ADR A1D、C2orf40、SP9、ALK、PDZRN4、ZSCAN1、PRLHR、NRSN1、FBN2、TLX3、ZSWIM2、GPR146, ZIK1, TRHDE-AS1, HOXD11, GRM7, ZNF365, NEFL, C10orf53, ALDH1A2, FGF10-AS1, AJAP1, RELN, KCNIP1, MIR2682, TMEM132E, NECAB1, ADC Y5, GJD2, EFCC1, GIPC3, LINC01231, PLPPR3, RYR3, CTD-2297D10.2, LINC00461, TBX18, CWH43, SELV, CCDC169, NID1, KCND2, ZNF835, DPP10, LOC 284395, SSTR4, TRPC4, RNF212, GRIA2, EPDR1, CSMD1, RAPGEF4, REC8, HS6ST1, AGTR1, WBSCR17, C14orf39, LRRTM1, THSD7B, LOC100289333, MIR4 281, TMEM196, ROBO3, TCF15, GALNT13, GUCY1A3, ADAM33, PTH2, CDH7, EPHA7, WNT16, NKAIN4, NGF, ZNF132, PHACTR3, MEIOB, CDH18, SLC18A3, ZNF 471, ZFP42, RASGRF1, GALR2, FGF3, DOCK10, TRHDE, SYT3, VWC2, SOX21, DBX2, FOXG1, ZNF667-AS1, ADGRV1, PCDHGB6, EPHA5-AS1, USP6, ZNF382, S T8SIA2, UTF1, CDH4, CNTN4, KCNK2, NRG1, PHYHIPL, THCAT155, MYO3A, FAM19A5, PHOX2A, GABRG3, IGF2-AS, TCERG1L, ZNF793, COL4A1, SORCS3, 11 a detection device for detecting a DNA methylation level of at least one gene selected from the group consisting of -Mar, PRR35, PLXDC2, TDRD10, LRRK2, ZFP28, NRIP3, ACAN, NLGN1, THY1, AFG3L1P, SFTA3, RAB12, TWIST2, FREM2, EBF2, FEZF2, CACNA1E, NTRK3, EMX2OS, HAND1, SNTG1, WNT7A, CCDC140, FGF14, PGR, SATB2-AS1, and GAD2; and a device for determining a risk of developing gastric cancer, wherein the device for determining a prediction of the risk of developing gastric cancer comprises:The system includes the following: an input unit for inputting the DNA methylation level obtained from the detection device; a learning unit that is pre-trained using teacher data including the DNA methylation level of a specific gene and information regarding the onset of gastric cancer, and outputs a risk of developing gastric cancer when input information including the DNA methylation level of a specific gene is input; and an output unit that outputs the risk of developing gastric cancer of a subject.

19. The system described in claim 18, wherein the learning unit further includes, as teacher data, information on at least one risk factor selected from the group consisting of smoking habits, drinking habits, and age, and outputs a risk of developing gastric cancer when input information including a DNA methylation level of a specific gene and information on the at least one risk factor is input.

Citation Information

Patent Citations

  • Carcinogenesis risk estimation using measurement of quantitative methylation of rasgrf1

    JP2012090555A

  • Method and kit for detection of cancer, and therapeutic agent for cancer

    JP2014140377A

  • Method for obtaining the index of detection of gastric cancer or risk assessment of gastric cancer onset

    JP2014161308A

  • GENE COMPOSITIONS FOR DETECTION OF CELL PROLIFERATIONAL ABNORMALITIES OR FOR SEVERITY GRADE AND USES THEREOF

    JP2017525977A

  • Method of assisting diagnosis of risk of onset of gastric cancer and artificial DNA used in method and diagnostic kit for diagnosing risk of onset of gastric cancer

    JP2018057362A