JP2023168188A - Risk determination method and risk determination system for arrhythmia or arrhythmia-related disorder - Google Patents

Abstract

To provide a method and a system that can determine the risk of arrhythmia or arrhythmia-related disorder such as atrial fibrillation with high accuracy, which can be applied to a wide range of races.SOLUTION: A method for determining a subject's risk of arrhythmia or arrhythmia-related disorder includes the steps of: preparing a data list including the effect allele and effect size of genetic variations related to arrhythmia; obtaining subject's genetic information; calculating an arrhythmia risk score from the subject's genetic information on the basis of information in the data list on the effect allele and effect size of genetic variations related to arrhythmia; and determining an arrhythmia risk on the basis of the risk score. The data list is a data list in which the results of genome analysis of Western populations and the results of genome analysis of non-Western populations are combined through meta-analysis.SELECTED DRAWING: Figure 1

Description

Application for exception to loss of novelty applied

The present invention relates to a method and system for determining the risk of arrhythmia or arrhythmia-related disorders such as atrial fibrillation using gene sequence analysis.

Arrhythmia is a condition in which the heart beats irregularly, and it can cause symptoms such as synchronization, shortness of breath, and chest pain, as well as fainting, heart failure, and even sudden death. Atrial fibrillation (AF) is one such arrhythmia that affects approximately 46.3 million people worldwide. The global prevalence of atrial fibrillation is increasing due to the rapid aging of the general population and enhanced detection of subclinical atrial fibrillation. Although advances have been made in the diagnosis and treatment of atrial fibrillation, a significant number of patients remain hospitalized with life-threatening complications such as stroke and heart failure, placing a heavy burden on patients and the public healthcare system. In addition to traditional clinical risk factors such as aging, obesity, hypertension, and heart failure, genetic contributions are also widely recognized in the development of atrial fibrillation. Recent genome-wide association studies (GWAS) have identified more than 100 atrial fibrillation-associated loci, some of which have been implicated in cardiac development, electrophysiology, contraction, and structural pathways ( Non-patent documents 1 to 4). However, because most AF-GWAS have been performed primarily in European populations, the genetic pathology of AF in non-European populations is not comprehensively understood, and the multifactorial Applying risk scores (PRS) to non-European populations has been difficult.

Christophersen, I.E. et al. Large-scale analyzes of common and rare variants identify 12 new loci associated with atrial fibrillation. Nat. Genet. 49, 946-952 (2017). Low, S.K. et al. Identification of six new genetic loci associated with atrial fibrillation in the Japanese population. Nat. Genet. 49, 953-958 (2017). Nielsen, J.B. et al. Biobank-driven genomic discovery yields new insights into atrial fibrillation biology. Nat. Genet. 50, 1234-1239 (2018). Roselli, C. et al. Multi-ethnic genome-wide association study for atrial fibrillation. Nat. Genet. 50, 1225-1233 (2018).

Predicting the risk of developing atrial fibrillation and prognosis based on genetic information is expected to play an important role in advancing the medical science and treatment of atrial fibrillation in the future. However, genetic analyzes of atrial fibrillation to date have focused on specific ethnic groups, making it difficult to universally predict risk across ethnic groups. In other words, it is known that there are large ethnic differences in the distribution of genetic variation, and for example, whether research results using European populations can be applied to non-Western populations, such as East Asian populations including Japanese. It was not clear what.

Therefore, an object of the present invention is to provide a method and system that can be applied to a wide range of races and that can determine the risk of arrhythmia such as atrial fibrillation or arrhythmia-related disorders with high accuracy.

First, the research group of the present inventors compared the genome sequences of approximately 9,800 atrial fibrillation patients and 140,000 controls registered in Biobank Japan, and found that We conducted genome-wide association analysis (GWAS) to comprehensively detect genetic variations found in . As a result, disease susceptibility loci and gene mutations related to atrial fibrillation as shown in Table 2 were identified.

Next, we meta-analyzed the GWAS results for the Japanese population (approximately 150,000 people) with the GWAS results for the European population (approximately 1.03 million people in a large-scale analysis in Europe, and approximately 60,000 people in the Finnish Biobank (FinnGen)). We conducted the world's largest cross-ethnic GWAS on atrial fibrillation, involving over 1.2 million people.
As a result, additional disease susceptibility loci and gene mutations associated with atrial fibrillation were identified as shown in Table 3.

Next, we created a polygenic risk score (PRS) using the results of the association analysis between genetic variation and atrial fibrillation obtained this time, and evaluated its predictive performance. The performance of the PRS was found to be significantly higher than that of the PRS based on Japanese data and the PRS based on European population data. Then, when we analyzed Japanese data (BBJ) using PRS created using the results of cross-ethnic GWAS, we found that the age of onset and prognosis of atrial fibrillation, as well as the risk of related disorders such as stroke and cerebral infarction. It has become clear that it is possible to predict efficiently.
The present invention was completed based on the above findings.

One aspect of the present invention is a method for determining the risk of arrhythmia or arrhythmia-related disorder in a subject, the method comprising:
A step of preparing a data list including effect alleles and effect sizes of genetic variations related to arrhythmia, a step of obtaining genetic information of the subject,
a step of calculating an arrhythmia risk score from the subject's genetic information based on information regarding the effect allele and effect size of the arrhythmia-related genetic variation in the data list; and determining the risk of arrhythmia or arrhythmia-related disorder based on the risk score. process, including
The present invention relates to a method, wherein the data list is a data list obtained by integrating genome analysis results of a Western population and genome analysis results of a non-Western population (non-Western population) through meta-analysis.

Another aspect of the present invention is a system for determining a subject's risk of arrhythmia or arrhythmia-related disorder, comprising:
means for storing data including a data list including effect alleles and effect sizes of genetic variants associated with arrhythmia;
means for receiving genetic information of a subject;
Means for calculating an arrhythmia risk score from genetic information of a subject based on data regarding effect alleles and effect sizes of arrhythmia-related genetic variations in the data list, and presenting a risk of arrhythmia or arrhythmia-related disorder based on the risk score. means, including;
The present invention relates to a system, characterized in that the data list is a data list that integrates genome analysis results of Western populations and genome analysis results of non-Western populations through meta-analysis.

Another aspect of the present invention is a method for determining the risk of arrhythmia or arrhythmia-related disorders, comprising:
rs202030113, rs2930856, rs1055894680, rs73205368, rs778479352,
rs4970418, rs9782984, rs75414548, rs1933723, rs12512502, rs6841049, rs17118812, rs7766436, rs12209223, rs4896104, rs2727757, rs17430357, rs1730310 1, rs11527634
, rs76460895, rs1769758, rs7126870, rs10500790, rs517938, rs10845620, rs2629755
, rs1344543, rs11614295, rs11841562, rs1886512, rs9284324, rs8096658, rs11881441, rs3746471, rs5754508, rs139557, rs73205368, rs1891095, one or more single nucleotide polynucleotides The present invention relates to a method comprising the step of analyzing a mold.

According to the method and system of the present invention, the risk of arrhythmia such as atrial fibrillation or arrhythmia-related disorders can be accurately determined by calculating a risk score using a data list that integrates genome analysis results of multiple ethnicities through meta-analysis. can be determined. This will make it possible to prevent arrhythmias such as atrial fibrillation and related diseases, provide guidelines for the prevention and treatment of arrhythmias, and build a foundation for the realization of precision medicine for arrhythmias.

Overview of research design. Top row, overview and reference panel of the Japanese cohort consisting of 9,826 AF cases and 140,446 controls. Middle row: GWAS summary of each study and number of cases and control group in trans-ancestry meta-analysis. Bottom row, summary of downstream analysis of atrial fibrillation-related mutations in the Trans-ancestry meta-analysis. GWAS, genome-wide association analysis; JENGER, Japanese ENcyclopedia GEnetic associations by Riken; MR, Mendelian randomization; PRS, polygenic risk score; TWAS, transcriptome-wide association analysis Manhattan plot of trans ancestry meta-analysis. We present the results of a trans-ancestry meta-analysis (77,690 atrial fibrillation cases and 1,167,040 controls). The log ₁₀ BF is shown on the y-axis with respect to the genomic position (hg19) on the x-axis. Association signals that reached genome-wide significance levels (log ₁₀ BF>6) are shown in blue for previously reported loci and red for new loci. AF is atrial fibrillation, BF is Bayes factor. Performance of polygenic risk scores (PRS) derived from trans-ancestry meta-analyses. Each point represents the median Nagelkerke pseudo ^R2 of polygenic risk scores in (2,953 cases, 21,194 controls). Error bars indicate 95% CI. Median values and CI of pseudo ^R2 were estimated from ^5x104 bootstrap methods. The color of each point indicates the ancestor included in the discovery population. The size of each point indicates the number of cases included in the discovery population. CI: confidence interval. Atrial fibrillation - evaluation of the predictive performance of polygenic risk scores. a, b: Association analysis between atrial fibrillation-polygenic risk score (AF-PRS) and age at onset of atrial fibrillation. a: Each point indicates the median age of onset by polygenic risk score quantile. b: Each point and error bar represent the estimated β and 95% CI obtained from the linear regression model. c: Relationship between atrial fibrillation and stroke subtypes. The data is AF-PRS 1s. d. Expressed as estimated OR and 95% CI for increase. d: Kaplan-Meier estimates and 95% CIs of cumulative events of cardiovascular death (a) and stroke death (b). Individuals are classified into high PRS (top 10%, red), low PRS (bottom 10%, blue), and intermediate (other, green). e: Effect of AF-PRS on long-term mortality. The data is AF-PRS 1s. d. Estimated HRs and 95% CI for the increase are shown. AF: atrial fibrillation, PRS: polygenic risk score, CI: confidence interval, OR: odds ratio, HR: hazard ratio, s. d. : standard deviation, CV: cardiovascular death, HF: death from congestive heart failure, IHD: death from ischemic heart disease.

<Arrhythmia risk determination method of the present invention>
The method for determining the risk of arrhythmia or arrhythmia-related disorders of the present invention includes:
preparing a data list including effect alleles and effect sizes of genetic variations associated with arrhythmia (data list preparation step);
the process of obtaining genetic information of the subject (genetic information obtaining process);
A step of calculating an arrhythmia risk score from the genetic information of the subject based on the arrhythmia-related genetic variation data in the data list (risk score calculation step), and determining the risk of arrhythmia or arrhythmia-related disorder based on the risk score. process (judgment process),
This is a method that includes

Arrhythmia includes tachyarrhythmia, bradyarrhythmia, and premature contractions. Tachyarrhythmias include sinus tachycardia, ventricular tachycardia, atrial fibrillation, atrial flutter, multisource atrial tachycardia, ventricular fibrillation, ventricular flutter, and supraventricular tachycardia. Bradyarrhythmias include sinoatrial block, atrioventricular block, junctional rhythm, sick sinus syndrome, respiratory arrhythmia, and bundle branch block. Premature contractions include atrial premature contractions and ventricular premature contractions. Among these, atrial fibrillation is preferred.
Arrhythmia is diagnosed by electrocardiogram testing, etc.
Arrhythmia-related disorders include cerebral infarction, cerebral hemorrhage, and the like caused by abnormal blood flow caused by arrhythmia.

The risk of arrhythmia includes the risk of developing arrhythmia and the risk of worsening the prognosis of arrhythmia (including the risk of death and the risk of developing arrhythmia-related disorders).

Each step will be explained below.

<Data list preparation process>
A data list is a list of data on effective alleles (risk alleles) and their effect sizes (degree of association with disease) for a large number of variants (genetic variations) related to arrhythmia.
Here, the type of genetic variation is not particularly limited as long as it is a variation different from the wild-type sequence, such as single nucleotide polymorphism (SNP), base deletion, or base insertion.
The total number of types of arrhythmia-related variants included in the data list is preferably 1,000 or more, more preferably 5,000 or more, and even more preferably 10,000 or more.
The effect size of a gene mutation associated with arrhythmia is set depending on the degree of association of each allele with arrhythmia, for example, the frequency of occurrence in arrhythmia patients.

The arrhythmia-related variant is preferably a variant with a minor allele frequency of 1% or more.
Furthermore, the arrhythmia-related variant is preferably a variant that is associated with arrhythmia with a probability of P<0.05.

The method of the present invention is characterized in that the data list is a data list that integrates the genome analysis results of a Western population and the genome analysis results of a non-Western population (non-Western population) through meta-analysis. Here, examples of Westerners include Caucasians,
For example, Europeans. Furthermore, examples of non-Western groups include Asian groups, and examples of Asian groups include, for example, Japanese groups. The genome analysis results are preferably obtained on a large scale, for example, from 10,000 or more subjects. Moreover, it is preferable that it is a result of genome-wide association analysis (GWAS). The genome analysis results of European and American populations and the genome analysis results of non-Western populations may be the results of genome analysis that has already been performed and made available to the public, or the results of new genome analysis. good. Genome analysis results that are publicly available include, for example, Biobank Japan for Japanese people, and the European atrial fibrillation data source (The Nord-trφndelag Health Study, deCODE, the Michigan Examples include, but are not limited to, analysis results from the FinnGen Genomics Initiative, DiscovEHR, UK Biobank, and the AFGen Consortium) and the FinnGen project. European AF GWASs (EUR and FIN)

The meta-analysis method is not particularly limited, and general meta-analysis methods can be used, and a fixed-effect model that does not take into account the heterogeneity between each study may be used, but it is preferable to take into account the heterogeneity between each study. A random effects model is used that takes into account the Meta-analysis is performed, for example, by METASOFT v. This can be done using analysis software such as 2 (http://genetics.cs.ucla.edu/meta).

Note that the data list used in the method of the present invention may include one or more, five or more, ten or more, twenty or more, or thirty or more of the novel arrhythmia-related variants described below. preferable.
rs202030113, rs2930856, rs1055894680, rs73205368, rs778479352, rs4970418, rs9782984, rs75414548, rs1933723, rs12512502, rs6841049, rs17118812, rs7 766436, rs12209223, rs4896104, rs2727757, rs17430357, rs17303101, rs11527634, rs76460895, rs1769758, rs7126870, rs10500790, rs517938, rs10845620, rs2629755, rs1344543, rs11614295, rs11841562, rs1886512, rs9284324, rs8096658, rs11881441, rs3746471, rs5754508, rs139557, rs73205368, rs1891095
These variants will be discussed later.

It is also preferable to include variants listed in Table 1 below (divided into Tables 1-1, 1-2, and 1-3) that are known to be associated with arrhythmia. It is preferable to include 10 or more, 50 or more, or 100 or more of these arrhythmia-related variants. In Table 1, each column, from left to right, is the chromosome where the variant exists (CHR), the chromosomal position (POS) of the reference sequence hg19 (GRCh37), the reference allele (REF), the mutant allele (ALT), and the registration of the dbSNP database. Indicates the number (rsID), neighboring genes, location on the gene, and relationship (log ₁₀ BF). The risk of arrhythmia is higher when the variant is a mutated allele.

<Genetic information acquisition process>
In this step, the genetic information of the subject, that is, information on the genome sequence is obtained. The genome sequence information may be any information that provides sequence information corresponding to each variant to be analyzed included in the data list, but whole genome sequence information is preferable.

The subject's genetic information can be obtained through conventional gene sequence analysis. For example, a subject's genetic information can be obtained using a next-generation sequencer or the like. In addition, in the method of the present invention, it is not essential to perform gene sequence analysis, and it is sufficient to obtain already analyzed data.

In the present invention, the subject of analysis is a human, for example, a human who is suspected of having atrial fibrillation risk. There are no particular restrictions on race, but examples include Asians, including Japanese, and Caucasians, including Europeans. Note that the sample used for genetic information analysis is not particularly limited as long as it contains chromosomal DNA derived from the subject. Examples include body fluids such as blood, urine, and spinal fluid, cells of the cervix and oral mucosa, and body hair such as hair. Although these samples can be used directly, it is preferable to isolate and analyze chromosomal DNA from these samples by conventional methods.

<Risk score calculation process>
In this step, an arrhythmia risk score is calculated from the subject's genetic information based on the risk allele and effect size data of the arrhythmia-related variants in the data list. Here, the term "risk score" refers to a score (value) indicating disease risk calculated from genetic information. It means that. For example, the genetic information of the subject is checked against the variant information in the data list, and the value of each variant, for example, the effect size (degree of disease association) multiplied by the number of risk alleles carried by the subject, is calculated, and the sum of these values is calculated. By calculating the risk score for the subject, it is possible to determine the risk score for the subject.

As the risk score (PRS: polygenic risk score), for example, the following formula is exemplified.
PRS=β ₁ x ₁ +β ₂ x ₂ +. ．． ．． β _k x _k + β _n x _n
Here, β _k is the log odds ratio (OR) for coronary artery disease of the risk allele for SNP _k , x _k is the number (0, 1, or 2) of the risk allele for SNP _k , and n is the This is the total number of analyzed SNPs.
Alternatively, it can be calculated logarithmically as shown below.
log(PRS)=x ₁ *logβ ₁ +x ₂ *logβ ₂ +. ．． ．． x _k *logβ _k +x _n *logβ _n

Preferably, calculation of the risk score is performed using a pruning and thresholding method.
That is, in the above data list, variants that are highly related to the disease, for example, variants that are associated with arrhythmia with a probability of P<0.05, preferably P<0.01, are selected, and then those variants are selected based on linkage disequilibrium. It is possible to classify, select an index variant with a low P value that represents each group, calculate a value for each index variant from the effect size of the risk allele and the number of risk alleles, and then calculate a risk score from that value. . Here, the index variant can be selected, for example, by dividing the variants into groups under the condition that the linkage disequilibrium coefficient r ² <0.5, preferably r ² <0.8.

<Judgment process>
In this step, the risk of arrhythmia or arrhythmia-related disorder is determined based on the calculated risk score. That is, this step is a determination step that is based on an objective index called a risk score and is not based on the judgment of a doctor.
For example, if the risk score exceeds a certain reference value, it can be determined that the arrhythmia risk is high. As the reference value, for example, risk scores can be calculated in advance for a group of healthy people and a group of arrhythmia patients, and a cutoff value can be determined.
Furthermore, according to the magnitude of the risk score, it is also possible to rank the disease risks, for example, in 5 stages or 10 stages. Furthermore, it is also possible to rank the arrhythmia in relation to its severity and prognosis.

The method of the invention can be performed using a computer.
That is, one aspect of the method of the present invention is a computer-implemented method for determining the risk of an arrhythmia or arrhythmia-related disorder, the method being operable in a computer system comprising: a processor and a memory;
receiving genetic information of the subject;
processing the received genetic information to determine an arrhythmia risk score;
Determining and presenting the risk of arrhythmia or arrhythmia-related disorders based on a risk score;
Relating to a method, including.
Here, processing is performed for each variant in the data list of arrhythmia-related genetic variations obtained by integrating the received genetic information with the genome analysis results of Western populations and the genome analysis results of non-Western populations. including cross-checking with information on effect alleles and effect sizes.

The method of the present invention may be systematized as a computer-implemented method. That is, one aspect of the present invention provides a computer system for determining a subject's risk of arrhythmia or an arrhythmia-related disorder.

<Arrhythmia risk determination system>
The arrhythmia or arrhythmia-related disorder risk determination system of the present invention includes:
a data storage means including a data list including effect alleles and effect sizes of genetic variations associated with arrhythmia;
means for receiving genetic information of a subject;
Means for calculating an arrhythmia risk score from genetic information of a subject based on information regarding effect alleles and effect sizes of arrhythmia-related genetic variations in the data list, and presenting a risk of arrhythmia or arrhythmia-related disorder based on the risk score. means, including;
The data list is characterized in that the data list is a data list in which the genome analysis results of a Western population and the genome analysis results of a non-Western population are integrated through meta-analysis.

In one embodiment, a data storage means comprising a data list comprising effect alleles and effect amounts of genetic variations associated with arrhythmia is located in a storage area within the computer. In another embodiment, the data storage means may be located outside the computer, such as an external server, and may be connected to the computer during use.

The genetic information of the subject may be received by being directly input into a computer, or may be received from a user interface connected to the computer. The subject's genetic information may also be received from a remote device via a wireless communication network.

The risk score calculation means compares the received genetic information with a data list in the data storage means, and calculates a risk score based on the result of the comparison.
For example, a risk score is calculated by substituting the matching results regarding risk alleles into a pre-stored formula for calculating a risk score.
The presentation means refers to pre-stored criteria based on the risk score and presents the risk determination result. Here, presenting the determination results includes outputting the information to a user interface coupled to the computer system and communicating the information to a remote device via a wireless communication network.
These means can be programmed and installed in a computer as software or applications, and cause the computer system to perform functions such as processing received genetic information (data collation, risk score calculation, etc.) and presenting results. .

<Risk determination method for arrhythmia or arrhythmia-related disorders using new arrhythmia-related SNPs>
Another aspect of the method for determining the risk of arrhythmia or arrhythmia-related disorder of the present invention includes the steps of analyzing one or more SNPs selected from the following, and determining the risk of arrhythmia or arrhythmia-related disorder based on the analysis results. include.

Here, as above, examples of arrhythmia include tachyarrhythmia, bradyarrhythmia, and premature contraction. Tachyarrhythmias include sinus tachycardia, ventricular tachycardia, atrial fibrillation, atrial flutter, multisource atrial tachycardia, ventricular fibrillation, ventricular flutter, and supraventricular tachycardia. Bradyarrhythmias include sinoatrial block, atrioventricular block, junctional rhythm, sick sinus syndrome, respiratory arrhythmia, and bundle branch block. Premature contractions include atrial premature contractions and ventricular premature contractions. Among these, testing for atrial fibrillation is preferred. Arrhythmia-related disorders include cerebral infarction, cerebral hemorrhage, and the like caused by abnormal blood flow caused by arrhythmia.
Furthermore, the risk of arrhythmia includes the risk of developing arrhythmia and the risk of worsening the prognosis of arrhythmia (including the risk of death and the risk of developing arrhythmia-related disorders).

Note that the rs number indicates the registration number of the dbSNP database (https://www.ncbi.nlm.nih.gov/snp/) of the National Center for Biotechnology Information.

rs202030113 is a SNP located in the intron of the SYNE1 gene on chromosome 6, and refers to the cytosine (C)/thymine (T) SNP at base 152466619 of chromosome 6 in the reference sequence hg19 (GRCh37). If it is C, the risk of arrhythmia is high. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of CC>TC>TT.

rs2930856 is a SNP located in the intron of the HCFC2 gene on chromosome 12, and the thymine (T)/
It refers to SNP of cytosine (C), and when this base is T, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of TT>CT>CC.

rs1055894680 is a SNP located in the intron of the ZNF689 gene on chromosome 16,
Reference sequence Thymine (T
)/cytosine (C) SNP, and when this base is T, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of TT>CT>CC.

rs73205368 is a SNP located in the intron of the PTCHD1 gene on the X chromosome, and means the SNP of cytosine (C)/thymine (T) at base 23399501 of the X chromosome in the reference sequence hg19 (GRCh37), and this base is C If so, the risk of arrhythmia is high. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of CC>TC>TT.

rs778479352 is a SNP located in the intron of the FGF13 gene on the X chromosome, and refers to the cytosine (C)/thymine (T) SNP at base 137790580 of the X chromosome in the reference sequence hg19 (GRCh37), and this base is C If so, the risk of arrhythmia is high. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of CC>TC>TT.

rs4970418 is a SNP located between the PERM1; HES4 genes on chromosome 1, and refers to the adenine (A)/guanine (G) SNP at base 918617 of chromosome 1 of the reference sequence hg19 (GRCh37). If the base is A, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of AA>AG>GG.

rs9782984 is a SNP located in the ncRNA intron of the MIR5096 gene on chromosome 1
This means the SNP of cytosine (C)/thymine (T) at base 16199051 of chromosome 1 of the reference sequence hg19 (GRCh37), and when this base is C, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of CC>TC>TT.

rs75414548 is a SNP located in the intron of the RHBDL2 gene on chromosome 1, and is an adenine (A) at base 39385714 of chromosome 1 of the reference sequence hg19 (GRCh37).
/ refers to SNP of guanine (G), and when this base is A, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of AA>AG>GG.

rs1933723 is a SNP located in the intron of the PALMD gene on chromosome 1, and is an adenine (A) at base 100149308 of chromosome 1 of the reference sequence hg19 (GRCh37).
/ refers to SNP of guanine (G), and when this base is A, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of AA>AG>GG.

rs12512502 is a SNP located in the intron of the MOB1B gene on chromosome 4, and refers to the cytosine (C)/adenine (A) SNP at base 71776935 of chromosome 4 in the reference sequence hg19 (GRCh37). If it is C, the risk of arrhythmia is high. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of CC>CA>AA.

rs6841049 is a SNP located in the intron of the LIN54 gene on chromosome 4, and refers to the guanine (G)/thymine (T) SNP at base 83910712 of chromosome 4 in the reference sequence hg19 (GRCh37). If it is G, the risk of arrhythmia is high. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of GG>GT>TT.

rs17118812 is a SNP located between the PFDN1; HBEGF genes on chromosome 5, and refers to the cytosine (C)/thymine (T) SNP at base 139703286 of chromosome 5 of the reference sequence hg19 (GRCh37). If the base is C, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of CC>TC>TT.

rs7766436 is a SNP located between the HDGFHL1; LOC105374972 gene on chromosome 6, and means the SNP of thymine (T)/cytosine (C) at base 22598259 of chromosome 6 of the reference sequence hg19 (GRCh37). If the base is T, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of TT>TC>CC.

rs12209223 is a SNP located in the intron of the FILIP1 gene on chromosome 6, and refers to the adenine (A)/cytosine (C) SNP at base 76164589 of chromosome 6 in the reference sequence hg19 (GRCh37). If it is A, the risk of arrhythmia is high. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of AA>AC>CC.

rs4896104 is LOC101928304 on chromosome 6; it is a SNP located between ALDH8A1 genes, and it means the SNP of thymine (T)/cytosine (C) at base 135119089 of chromosome 6 of the reference sequence hg19 (GRCh37), and this If the base is T, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of TT>TC>CC.

rs2727757 is a SNP located in the intron of the CDHR3 gene on chromosome 7, and refers to the guanine (G)/adenine (A) SNP at base 105612736 of chromosome 7 in the reference sequence hg19 (GRCh37). If it is G, the risk of arrhythmia is high. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of GG>GA>AA.

rs17430357 is a SNP located in the intron of the EXT1 gene on chromosome 8, and means the SNP of thymine (T)/adenine (A) at the 118863412th base of chromosome 8 of the reference sequence hg19 (GRCh37), and this base If is T, the risk of arrhythmia is high. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of TT>TA>AA.

rs17303101 is a SNP located between the PAPPA; ASTN2 gene on chromosome 9, and refers to the adenine (A)/guanine (G) SNP at base 119181794 of chromosome 9 of the reference sequence hg19 (GRCh37). If the base is A, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of AA>AG>GG.

rs11527634 is a SNP located in the intron of the CCDC7 gene on chromosome 10, and the thymine (T)/
It refers to SNP of cytosine (C), and when this base is T, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of TT>TC>CC.

rs76460895 is a SNP located between the C10orf128; C10arf71-AS1 genes on chromosome 10, and refers to the adenine (A)/guanine (G) SNP at base 50485434 of chromosome 10 of the reference sequence hg19 (GRCh37). , if this base is A, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of AA>AG>GG.

rs1769758 is a SNP located in the intron of the ZMIZ1 gene on chromosome 10, and refers to the thymine (T)/guanine (G) SNP at base 80898969 on chromosome 10 of the reference sequence hg19 (GRCh37). If is T, the risk of arrhythmia is high. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of TT>TG>GG.

rs7126870 is a SNP located in the intron of the STIM1 gene on chromosome 11, and refers to the SNP of thymine (T)/cytosine (C) at base 3890059 of chromosome 11 in the reference sequence hg19 (GRCh37). If is T, the risk of arrhythmia is high. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of TT>TC>CC.

rs10500790 is a SNP located in the intron of the SPON1 gene on chromosome 11, and is an adenine (A) at base 14036189 of chromosome 11 in the reference sequence hg19 (GRCh37).
/ refers to SNP of guanine (G), and when this base is A, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of AA>AG>GG.

rs517938 is a SNP located between the LOC100129203; FAM76B genes on chromosome 11, and refers to the thymine (T)/cytosine (C) SNP at the 95089882nd base on chromosome 11 of the reference sequence hg19 (GRCh37). If the base is T, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of TT>TC>CC.

rs10845620 is a SNP located in the intron of the APOLD1 gene on chromosome 12, and is an adenine (A
)/guanine (G) SNP, and when this base is A, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of AA>GA>GG.

rs2629755 is a SNP located in the intron of the HCFC2 gene on chromosome 12, and is the guanine (G) at base 104492003 of chromosome 12 in the reference sequence hg19 (GRCh37).
/Adenine (A) SNP, and if this base is G, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of GG>GA>AA.

rs1344543 is a SNP located between the MVK;FAM222A genes on chromosome 12, and the cytosine (C
)/thymine (T) SNP, and when this base is T, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of CC>TC>TT.

rs11614295 is a SNP located in the intron of the RPH3A gene on chromosome 12, and the adenine (A
)/guanine (G) SNP, and when this base is A, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of AA>GA>GG.

rs11841562 is a SNP located in the intron of the MICU2 gene on chromosome 13, and is an adenine (A) at base 22111521 of chromosome 13 in the reference sequence hg19 (GRCh37).
/ Cytosine (C) SNP, and when this base is A, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of AA>AC>CC.

rs1886512 is a SNP located in the intron of the KLF12 gene on chromosome 13, and the adenine (A)/
It refers to SNP of thymine (T), and when this base is A, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of AA>AT>TT.

rs9284324 is a SNP located in the intron of the MYH11 gene on chromosome 16, and the adenine (A)/
It refers to SNP of guanine (G), and when this base is A, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of AA>GA>GG.

rs8096658 is a SNP located in the intron of the NFATC1 gene on chromosome 18, and is a guanine (G) at base 77156537 of chromosome 18 in the reference sequence hg19 (GRCh37).
/Cytosine (C) SNP, and when this base is G, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of GG>GC>CC.

rs11881441 is a SNP located in the intron of the BICRA gene on chromosome 19, and is a cytosine (C) at base 43142746 of chromosome 19 in the reference sequence hg19 (GRCh37).
/Adenine (A) SNP, and when this base is C, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of CC>AC>AA.

rs3746471 is a SNP located in the exon of the KIAA1755 gene on chromosome 20, and is an adenine (A
)/guanine (G) SNP, and when this base is A, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of AA>AG>GG.

rs5754508 is a SNP located downstream of the SDF2L1 gene on chromosome 22, and refers to the guanine (G)/cytosine (C) SNP at base 21999229 on chromosome 22 of the reference sequence hg19 (GRCh37). If it is G, the risk of arrhythmia is high. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of GG>GC>CC.

rs139557 is a SNP located in the intron of the MEI1 gene on chromosome 22, and refers to the guanine (G)/thymine (T) SNP at base 42189407 of chromosome 22 in the reference sequence hg19 (GRCh37). If it is G, the risk of arrhythmia is high. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of GG>GT>TT.

rs73205368 is a SNP located in the intron of the PTCHD1 gene on the X chromosome, and means the SNP of cytosine (C)/thymine (T) at the 23399501st base of the X chromosome in the reference sequence hg19 (GRCh37), and this base is T If so, the risk of arrhythmia is high. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of CC>TC>TT.

rs1891095 is a SNP located between the ZIC3; UNCOO889 gene on the X chromosome, and refers to the adenine (A)/cytosine (C) SNP at base 137418967 of the X chromosome in the reference sequence hg19 (GRCh37), and this base If it is A, there is a high risk of arrhythmia. Furthermore, when analyzed in consideration of genotype, the risk of arrhythmia is higher in the order of AA>AC>CC.

Furthermore, the SNPs to be analyzed in the present invention are not limited to those described above, and SNPs that are in linkage disequilibrium with the above-mentioned SNPs may be analyzed. Here, "SNP in linkage disequilibrium with the above SNP" means that the above SNP has r ² >0.5, preferably r ² >0.8, more preferably r ² >0.
．． Refers to SNPs that satisfy the relationship 9. r ² is the linkage disequilibrium coefficient. SN in linkage disequilibrium
P can be identified using, for example, the HapMap database (http://www.hapmap.org/index.html.ja).

Although the above-mentioned SNP may be analyzed alone, it is preferable to analyze a plurality of SNPs, for example, in combination of 5 or more, 10 or more, 20 or more, or 30 or more. The above SNP may be analyzed in combination with known arrhythmia-related SNPs such as those listed in Tables 1-1, 1-2 and 1-3 above. By analyzing multiple SNPs in combination, it is possible to improve the accuracy of arrhythmia risk determination. Analyze multiple SNPs, confirm whether each is a risk allele, and use the total number of risk alleles as an indicator of arrhythmia risk (for example, if you have 1 or more, 2 or more, or 3 or more risk alleles, you will be diagnosed with arrhythmia). However, it is also possible to calculate a risk score from the analysis results of multiple SNPs and use the obtained risk score as an index of arrhythmia risk. For example, if the risk score exceeds a certain reference value, it can be determined that the arrhythmia risk is high.

The risk of arrhythmia can be determined by examining the base types of one or more SNPs as described above and associating them with arrhythmia using the obtained results or indicators such as risk scores based on the results. . That is, the method of the present invention can provide data for arrhythmia diagnosis.

The results determined by the method of the present invention are provided to doctors and the like as necessary. A doctor or the like provided with the results can diagnose arrhythmia after performing necessary tests such as electrocardiogram measurements, blood tests, and ultrasound tests. If doctors, etc. diagnose that the risk of arrhythmia is high, they can take appropriate preventive measures such as administering drugs, and if they diagnose that the patient has developed the disease, they can administer treatments such as administering drugs or surgery. .

Note that SNP analysis can be performed by a normal genetic polymorphism analysis method. Examples include sequence analysis, PCR, hybridization, invader method, etc.
Not limited to these.

The present invention also provides test reagents such as primers and probes for testing arrhythmia or arrhythmia-related disorders. Examples of such probes include probes that include the above-mentioned SNP site and can determine the type of base at the SNP site based on the presence or absence of hybridization. Specifically, probes with a length of 15 bases or more that have a base sequence that includes the polymorphic site of each SNP or its complementary sequence, or a sequence that includes a base that is in linkage disequilibrium with the base or its complementary sequence. Examples include probes having a length of 15 bases or more. The length of the probe is preferably 15 to 35 bases, more preferably 20 to 35 bases.

Furthermore, examples of the primers include primers that can be used in PCR for amplifying the above SNP site, and primers that can be used for sequence analysis (sequencing) of the above SNP site. Specifically, we use primers that can amplify and sequence regions that contain the bases at each SNP site, as well as primers that can amplify and sequence regions that contain bases that are in linkage disequilibrium with the relevant bases. Examples include primers that can be used. The length of such a primer is preferably 15 to 50 bases, more preferably 15 to 35 bases, and even more preferably 20 to 35 bases. Primers for sequencing SNP sites include primers having sequences in the 5' region of the above bases, preferably 30 to 100 bases upstream, and primers having sequences in the 3' region of the above bases, preferably 30 to 100 bases downstream. A primer having a sequence complementary to the region is exemplified. Primers used to determine polymorphisms based on the presence or absence of amplification by PCR include primers that have a sequence containing the above bases and the 3' side of the bases, and primers that have complementary sequences to the sequences containing the above bases; Examples include primers containing a complementary base to the above base on the 3' side.

EXAMPLES Hereinafter, the present invention will be specifically explained with reference to Examples, but the present invention is not limited to the following embodiments.

Method Samples All Japanese GWAS subjects are Japanese people registered with the Biobank Japan (BBJ) project (https://biobankjp.org/). BBJ has 12 collaborating medical institutions nationwide (Osaka Prefectural Adult Disease Center, Japan Cancer Research Foundation Cancer Institute Ariake Hospital, Juntendo University, Tokyo Health Geriatrics Center, Nippon Medical University, Nihon University School of Medicine, Iwate Medical University). This is a hospital-based nationwide biobank project that collects DNA, serum samples, and clinical information from hospitals such as Tokushukai Hospital, Shiga University of Medical Science, Fukujuen, National Hospital Organization Osaka Hospital, and Iizuka Hospital). From 2003 to 2007, approximately 200,000 patients with any of the 47 target diseases were enrolled. All samples were over the age of 18. Informed consent was obtained from all participants, and the study was approved by the relevant ethics committees at each institution.
For quality control of GWAS, samples with call rate < 0.98 and PLINK 2.0 (version August 20, 2018: https://www.cog-genomics.org/plink/2.0/) were used. Relatives with PI_HAT>0.2, an index of kinship based on pairwise ancestry identity performed, were excluded. and heterozygosity rate>+4s. d. samples were excluded. To clarify the stratification of the population, PLINK 2.0. Principal component analysis was performed using , and outliers in the Japanese cluster were excluded. Case samples in GWAS were selected from people with atrial fibrillation or atrial flutter diagnosed by a physician based on common medical practice, or recorded on a 12-lead electrocardiogram.

Genotyping, Imputation, and Quality Control Illumina Human OmniExpress Genotyping BeadChips or a combination of Illumina HumanOmniExpress and HumanExome BeadChips (manufactured by Illumina, San Diego, CA) ) was used to determine the genotype of the GWAS subjects. As a genotype quality control, we excluded variants with (i) SNP call rate <0.99, (ii) MAF <0.01, and (iii) Hardy-Weinberg equilibrium P value ≤1.0×10 ⁻⁶ . did. We also prephased genotypes using EAGLE and included a homemade reference panel of 1,037 Japanese individuals from BBJ using minimac3 (https://genome.sph.umich.edu/wiki/Minimac3). Imputation of variant frequency was performed using the 1,000 Genome Project Phase 3 (1KG Phase 3; May 2012) reference panel (Nature 467, 1061-1073 (2010).). Regarding the X chromosome, prephase was performed for both men and women, and imputation was performed separately for men and women. The frequency of variants in the X chromosome was assigned between 0 and 2 in males.

GWAS
In Japan's GWAS, using PLINK 2.0 (August 20, 2018 version), we assumed an additive model that adjusted for age, ^age2 , gender, and the top 20 principal components, and made associations using logistic regression analysis. Ta. Variants with a minimac3 imputation quality score of 0.3 or higher and a MAF of 0.001 or higher were selected. For the X chromosome, association analyzes were performed separately for males and females, and the results were combined using an inverse variance weighted fixed effects model implemented with METASOFT (v2.0.1). Heterogeneity between studies was calculated using Cochran's Q test. Variants with strong heterogeneity (P _het <1.0×10 ⁻⁴ ) were filtered. Genome-wide significance thresholds were P < 5.0 × 10 ^-8 for variants with MAF ≥ 1% and P < 5.71 × 10 ^-9 for variants with MAF < 1% (0.05/ 8,753,038 variants). The genomic inflation factor (λ _GC ) was 1.12, but LD score regression showed that the inflation was mainly due to polygenic effects (LD score regression intercept = 1.02). Adjacent genome-wide significant SNPs were grouped into one locus if they were within 1 Mb of each other. We defined loci as candidate regions located within 500 Mb upstream and downstream of genome-wide significant variants. If a locus was not within LD (r ² <0.10) of a reported locus, we defined it as a novel locus. Genetic determination or imputed variants were annotated using ANNOVAR (build July 7, 2017) (http://annovar.openbioinformatics.org).

LD score regression and heritability To estimate confounding biases such as cryptic correlations and population stratification, LD-score regression (version 1.0.0) was performed. SNPs with MAF≧0.01 were selected, and variants within the major histocompatibility complex region were excluded. East Asian LD scores were used for regression (https://github.com/bulik/ldsc/).

Trans-ancestry meta-analysis The summary results of the two European atrial fibrillation GWAS (EUR and FIN) are available on the respective previously published websites (http://csg.sph.umich.edu/willer/public/afib2018) (Nat. Genet 50, 1234-1239 (2018).) and FinnG.
obtained from the Finngen Research Project website (https://www.finngen.fi/en). The EUR is sourced from six contributing atrial fibrillation data sources: The Nord-trφndelag Health Study, deCODE, the Michigan Genomics Initiative, DiscovEHR, UK Biobank, and the AFGen Consortium) meta-analysis. FIN consisted of 102,739 Finnish participants, and phenotypes were derived from ICD codes from the Finnish National Hospital Registry and Cause of Death Registry as part of the FinnGen project. To examine the bias of these GWAS, we calculated the LD score regression intercept for each study and found that these two studies were well calibrated (EUR = 1.052 (standard error = 0.012) and LD score regression intercept with FIN=1.033 (standard error=0.010). In addition, when we calculated the genetic correlation, a significant genetic correlation was found between EUR and FIN (r _g = 0.918, standard error = 0.035, P = 3.9 × 10 ^-155 ) .
To account for ancestry heterogeneity among the three studies, we used the MANTRA algorithm, which allows for heterogeneity among diverse ancestry groups and improves performance compared to fixed-effects meta-analyses and random-effects meta-analyses. Applied in trans-ancestry meta-analysis (Genet. Epidemiol.35, 809-822 (2011).). We selected variants with MAF≧1% in both Japanese and European populations and examined the correlation. SNPs with log ₁₀ BF>6 were considered genome-wide significant.

Genetic correlation of trans ancestry To estimate the genetic correlation of trans ancestry, Popcorn software (v.1.0) was used. A 1 KG East Asian or European sample was used as the criterion for LD, and variants within the major histocompatibility complex region were excluded according to the software instructions. To estimate the genetic correlates of the susceptibility scale, we assumed that the prevalence of atrial fibrillation was 0.58% in the Japanese population, 1.38% in the British population, and 0.97% in the Finnish population. (https://vizhub.healthdata.org/gbd-compare/).

Confident set analysis To construct a plausible set of variants including causal variants, we constructed a 99% reliable set at each AF-associated locus using the BF obtained from MANTRA results in trans-ancestry meta-analysis. . For each locus, the posterior probability (PP) of the j-th SNP was calculated from the following formula: PP _j =BF _j /Σ _k BF _k . Here, BF _j represents the BF of the j-th SNP, and BF _k represents all variants contained in that locus. Variants were then added in order of decreasing PP until the cumulative PP reached 0.99 to construct a 99% reliable set. To assess whether the 99% reliable set derived from the trans ancestry meta-analysis narrowed down the causative variants, we performed two additional meta-analyses (EUR+BBJ and EUR+FIN) using the MANTRA algorithm to determine whether the causative variants identified in the trans ancestry meta-analysis We constructed a 99% reliable set of 150 atrial fibrillation-related gene loci. Then, for each locus, the number of variants included in the 99% reliable set was compared. Among them, loci in which at least one variant exceeded genome-wide significance (log ₁₀ BF>6) in all meta-analyses were targeted for testing. Differences in reliable set sizes were tested using a paired Wilcoxon rank sum test.

Derivation and performance of polygenic risk scores
A polygenic risk score (PRS) was derived using a pruning and thresholding method. The polygenic risk score is calculated using the P value threshold (0.5, 5.0×10 ⁻² , 5.0×10 ⁻⁴ , 5.0×10 ⁻⁶ , 5.0×10 ⁻⁸ ) and r ² It was created with a range of threshold values (0.2, 0.5, 0.8). For survival analysis, the dataset was divided into three groups: (i) discovery group (6,890 cases, 49,451 controls) for deriving and validating polygenic risk scores; (ii) polygenic risk (iii) a test group for evaluating the performance of the score (2,953 cases, 21,194 controls); and (iii) a survival time analysis group (70,645 controls). To ensure the independence of derivation and validation of polygenic risk scores, a 10-fold cross-validation method was used. First, the discovery group was randomly divided into 10 subgroups, of which 9 subgroups were used for deriving the polygenic risk score and the remaining one was used for validating the polygenic risk score. To determine the weighting of polygenic risk scores, the GWAS was performed in combination with two European GWAS: (i) Japan subgroup GWAS, (ii) two European GWAS (EUR or FIN), (iii) Japan subgroup GWAS. Group GWAS+EUR (fixed or random effects model), (iv) Japan subgroup GWAS+FIN (fixed or random effects model), (v) EUR+FIN (fixed or random effects model), and (vi) Japan subgroup GWAS+EUR+FIN (fixed or random effects model). model). Meta-analysis was performed using METASOFT software. For meta-analyses using the Japanese subgroup GWAS, LD was calculated separately based on 1KG East Asian or European samples. Thereafter, PLINK 2.0 was used to calculate the polygenic risk score for the retained validation cohort and verify its performance. The validation cohort was changed and these steps were repeated 10 times to obtain performance measurements for each combination. Note that the validation cohort was excluded from the derivation of the GWAS. The performance of the polygenic risk score is defined as (1) the age, sex, and normalized polygenic risk score as the Nagel-Kerge pseudo- ^R2 , and ( ² ) the receiver operating curve in the same model as the Nagel-Kerge's pseudo-R2. was measured as the area under the curve. The optimal model/parameter set for each combination model (BBJ, EUR, FIN, BBJ+EUR, BBJ+FIN, EUR+FIN, BBJ+EUR+FIN) was determined by averaging Nagel-Kerge's pseudo ^R2 . Using the optimal model and parameters determined in the derivation and validation cohorts, we calculated a polygenic risk score and evaluated its performance on an independent test cohort. The bootstrap method was applied to evaluate the performance distribution of polygenic risk scores. Individuals in the test cohort were randomly sampled with replacement, and the performance of that sample was calculated. By repeating this procedure 5.0×10 ⁴ times, the distribution of performance indicators was estimated.

Atrial fibrillation - Association of polygenic risk score with age of onset of atrial fibrillation and stroke subtype Assessing the association between atrial fibrillation - polygenic risk score (AF-PRS) and age of onset of atrial fibrillation For this purpose, we extracted a sample of atrial fibrillation cases (n = 7,458, median age at onset of atrial fibrillation was 63 years (IQR 56-71)) and used a linear regression model that adjusted for gender and the top 10 main factors. We estimated the difference in the effect of age at onset of atrial fibrillation between individuals with the top polygenic risk scores and those with the remaining polygenic risk scores. In the correlation analysis between atrial fibrillation and stroke subtypes, the P value associated with OR was calculated using a logistic regression model that adjusted for age, sex, and the top 10 major factors. Among the control samples included in our dataset (n=140,447), 14,120 cases of stroke phenotype were found. Cerebral infarction was the most common, with 9,177 cases of cerebral infarction, 111 cases of cardioembolism, 1,465 cases of atherothrombosis, 1,266 cases of lacunar infarction, 1,143 cases of cerebral hemorrhage, and 967 cases of subarachnoid hemorrhage.

Survival analysis In the survival analysis, the association between AF-PRS and long-term mortality was evaluated using Cox proportional hazards model. Survival tracking data by ICD-10 codes were obtained for 132,737 individuals from the BBJ dataset. Causes of death were classified according to ICD-10 codes. (i) Cardiovascular death (I00-I99), (ii) Heart failure death (I50), (iii) Ischemic heart disease death (I20-I25), and (iv) Stroke death (I60, I61, I63 and I64). It is. Median follow-up was 8.4 years (IQR 6.8-9.9). The Cox proportional hazards model adjusted for gender, age, top 10 major factors, and disease symptoms. The analysis was performed using the R package survival v. 2.44, and survival curves were generated using the R package survminer v.2.44. Estimation was made using a modified version of 0.4.6.

Results Figure 1 shows an overview of the research design for the five new loci for atrial fibrillation identified in the Japanese genome-wide association analysis. Using 16,394,105 autosomal variants and 423,039 variants with minor allele frequency (MAF) >0.1% on the , 826 cases and 140,446 controls), genome-wide association analysis was performed. Genome-wide association analysis identified 31 genome-wide significant atrial fibrillation-related gene loci, of which 5 were novel loci that had not been previously reported (Table 2). The proportion of variation in atrial fibrillation (AF) explained by genome-wide genetic variations (single nucleotide polymorphism (SNP) heritability: ^h2 ) detected in this Japanese genome-wide association analysis is estimated to be , 6.1% (standard error 1.4%), and the susceptibility scale ^h2 using linkage disequilibrium (LD)-score regression was estimated to be 11.7% (standard error 2.6%). It was done.

Among these, the lead variant rs202030113 on 6q25.1 is located in an intron region 3 bases away from the exon-intron boundary of SYNE1, and was predicted to be a splice donor loss based on the splice AI delta score of ^0.3310 . SYNE1 encodes the nespralin-1 (spectrin repeat) protein, composes a nuclear membrane protein complex together with Sad1p/UNC84 domain-containing protein (SUN1/2), and binds to lamin A/C via its nuclear membrane domain. ing. Mutations in the LMNA and SYNE1 genes have been identified in patients with severe muscular dystrophy and dilated cardiomyopathy (Nat. Genet. 21, 285-288 (1999), Nat. Rev. Genet. 7, 940-952 (2006). ).). SYNE1
Genetic mutations lead to abnormalities in nuclear morphology, myoblast differentiation, cardiac development, and alter nuclear envelope protein complexes that contribute to the structural matrix during atrial arrhythmia (Hum. Mol. Genet. 26 , 2258-2276(2017).)

A strong signal was detected by rs778479352, a lead variant located in the intron of FGF13 (odds ratio [OR] for developing atrial fibrillation = 2.00, 95% confidence interval [CI] = 1.73-2.31, P = 1 .6×10 ⁻²⁰ ) was displayed. FGF13 encodes a member of the fibroblast growth factor family and has a wide range of mitogenic and cell survival effects. FGF13 binds directly to the C-terminus of the major cardiac sodium channel (Na _V 1.5) located in the sarcolemma. Knockdown of FGF13 in rat cardiomyocytes resulted in decreased function of Na _V 1.5-reduced Na ⁺ current density, decreased Na ⁺ channel utilization, and inactivation of Na _V 1.5-reduced Na ⁺ current. A decrease in the recovery rate was observed (Circ. Res. 109, 775-782 (2011)). This evidence of conduction defects in cardiomyocytes suggests that FGF13 is an important target gene associated with atrial fibrillation.

Identification of 33 novel loci for atrial fibrillation by trans-ancestry meta-analysis To improve the statistical power to detect further genetic associations with atrial fibrillation, we analyzed the current Japanese GWAS (BBJ ) and two European GWAS (Large European Population Meta-Analysis (EUR) (Nat. Genet. 50, 1234-1239 (2018)) and FinnGen Data Release 2 (FIN) biobank data). We conducted a trans-ancestry meta-analysis. All three datasets included 77,690 cases (BBJ: 9,826, EUR: 60,620; FIN: 7,244) and control group 1,167,040 (BBJ: 140,446; EUR: 970, 216; FIN: 56,378). We tested a total of 5,158,449 variants with MAF ≥1% and identified 150 genome-wide significant AF-associated loci (log ₁₀ Bayes factor (BF) >6; Figure 2). This trans-ancestry meta-analysis identified a total of 33 novel loci (Table 3).

Allelic effects shared between Japanese and European populations, and fine mapping/confidence set analysis in trans-ancestry meta-analysis 150 lead variants, between BBJ, EUR, and FIN, to calculate variant allele frequencies and allelic effects. compared. The allele frequencies are in good agreement between EUR and FIN (Spearman's ρ = 0.974, P < 2.2 × 10 ^-16 ), but the allele frequencies between BBJ and EUR and between BBJ and FIN are were found to be moderately correlated (ρ=0.592, P=1.5×10 ⁻¹⁵ and ρ=0.632, P<2.2×10 ⁻¹⁶ , respectively). Furthermore, we found a significant positive correlation between the matched allele effects of these variants (ρ = 0.769, P < 2.2 × 10 ⁻¹⁶ for BBJ vs. EUR; , ρ=0.769, P<2.2×10 ⁻¹⁶ ). To further investigate the relationship of allele effects between populations, we performed trans-ancestry genetic correlation analysis, and found that BBJ was strongly correlated with EUR and FIN (BBJ and EUR: r _g =0.990, standard error = 0.097, BBJ and FIN: r _g =0.955, standard error = 0.344).

To evaluate the contribution of trans-ancestry meta-analysis to purification of putative causal variants, we constructed a 99% reliable set of 150 AF-associated loci detected in this trans-ancestry meta-analysis, and used three combinations of meta-analyses ( We compared the number of variants included in the 99% reliable set derived from EUR+BBJ, EUR+FIN, BBJ+EUR+FIN). The 99% confidence set size from EUR+BBJ was significantly reduced compared to the size from EUR+FIN (P=0.004, paired Wilcoxon rank sum test). Additionally, trans-ancestry meta-analysis of three datasets (BBJ+EUR+FIN) resulted in the greatest reduction in variant number of all meta-analysis combinations (median number of variants = 12; interquartile range [IQR] 5-36 ). Notably, only in a trans-ancestry meta-analysis consisting of three datasets, a single variant, rs67329386, at the ZFHX3 locus was identified with a high posterior probability of 0.995. In past atrial fibrillation-GWAS, several variants were found in the ZFHX3 gene locus (Nat. Genet. 41, 876-878 (2009)), and ZFHX3, a large transcription factor, plays a role in DNA binding and transcriptional activity together with PITX2. (J. Biol. Chem. 273, 20066-20072 (1998)).

Predictive power of polygenic risk scores (PRS) derived from single population GWAS and trans-ancestry meta-GWAS Polygenic risk scores (PRS) have the potential to stratify risk of complex constitutions and diseases based on genetic data. . However, it remains difficult to transfer polygenic risk scores from diverse populations to populations with different ancestry. Therefore, we examined the performance of polygenic risk scores derived from various combinations of summary statistics in the Japanese population. The case-control sample was divided into a derivation dataset, a validation dataset, and a test dataset, and summary statistics of three GWAS studies (BBJ, EUR, FIN) and 255 combinations of parameters for deriving polygenic risk scores were calculated. It was constructed. Based on the performance of the polygenic risk score in the validation cohort, we determined the parameters showing the best performance for each combination of summary statistics (BBJ, FIN, EUR, BBJ+FIN, BBJ+EUR, EUR+FIN, BBJ+EUR+FIN) and The performance of the optimal model was evaluated (Figure 3). Consistent with population specificity, polygenic risk scores obtained from BBJ performed significantly better than polygenic risk scores obtained from European studies (BBJ vs. EUR and BBJ vs. FIN). P<2.2×10 ⁻¹⁶ for both). Furthermore, the performance of polygenic risk scores obtained from meta-analyses including different ancestry groups (EUR+BBJ and FIN+BBJ) is not only that from a single study (P<2.2×10 ⁻¹⁶ for all cases); It was also significantly superior to that from a meta-analysis of two European studies (P<2.2×10 ⁻¹⁶ for both EUR+BBJ vs. EUR+FIN and BBJ vs. FIN+EUR). Among all models, the polygenic risk score obtained from the three studies with multi-ancestry and largest sample size (BBJ+EUR+FIN) showed the best performance (pseudo R ² = 0.144, 95% CI = 0.130-0.154, area under the receiver operating characteristic curve = 0.737, 95% CI = 0.726-0.748).

Atrial fibrillation - Effect of polygenic risk score on AF-related phenotypes and cardiovascular outcomes To examine the potential for clinical application of polygenic risk score (PRS), a case sample of BBJ (n = 7,459 ), we investigated the relationship between the polygenic risk score and the age of onset of atrial fibrillation and found that the polygenic risk score was correlated with the age of onset of atrial fibrillation. As a result, the age at onset of atrial fibrillation decreased as the polygenic risk score increased, and those with polygenic risk scores in the top 1% were estimated to have atrial fibrillation onset approximately 4 years younger than the rest of the population. (Figures 4a, 4b).

Next, we investigated whether the phenotype of cerebrovascular disorders could be predicted by the atrial fibrillation-polygenic risk score (AF-PRS). Logistic regression analysis of the 140,446 control samples in our dataset revealed that the polygenic risk score was associated with cerebral infarction (OR [95% CI] = 1.04 [1.02-1.07 ], P = 4.0 × 10 ^-4 ), and cardioembolic stroke (OR [95% CI] = 1.35 [1.13-1.63], P = 1.3 × 10 ^-3 ) ( Figure 4c) was significantly associated with increased risk.
Importantly, polygenic risk scores were observed to have the greatest impact on cardioembolic stroke among those with other stroke phenotypes, and AF-PRS was observed to have the greatest effect on cardioembolic stroke (AF-PRS) in clinically undetectable atrial fibrillation ( This suggests that AF-related symptoms, such as subclinical atrial fibrillation) or thrombotic or hypercoagulable symptoms, may occur even in people without atrial fibrillation.

Atrial fibrillation - To further explore the clinical utility of the polygenic risk score, we evaluated the impact of the polygenic risk score on mortality using long-term follow-up data of BBJ. Kaplan-Meier estimates of cumulative mortality were higher for individuals with higher polygenic risk scores, especially cardiovascular and stroke-related mortality (Figure 4d). In Cox regression analysis, the polygenic risk score was associated with cardiovascular mortality (hazard ratio [HR] per standard deviation [s.d.] of the polygenic risk score = 1.06, 95% CI = 1.02-1. 11, P = 5.0 × 10 ^-3 ) and stroke death (HR [95% CI] = 1.14 [1.04-1.24], P = 4.0 × 10 ^-3 ; Fig. 4e). It was found to be significantly associated with increased risk.

Claims (14)

A method for determining the risk of developing an arrhythmia or an arrhythmia-related disorder in a subject, the method comprising:
A step of preparing a data list including effect alleles and effect sizes of genetic variations related to arrhythmia, a step of obtaining genetic information of the subject,
a step of calculating an arrhythmia risk score from the subject's genetic information based on information regarding the effect allele and effect size of the arrhythmia-related genetic variation in the data list; and determining the risk of arrhythmia or arrhythmia-related disorder based on the risk score. process,
including;
The data list is characterized in that the data list is a data list that integrates genome analysis results of Western populations and genome analysis results of non-Western populations through meta-analysis,
Method. 2. The method of claim 1, wherein the non-Western population is an Asian population. 2. The method of claim 1, wherein the calculation of the risk score is performed using a pruning and thresholding method. 2. The method of claim 1, wherein the meta-analysis is performed using a random effects model. 2. The method of claim 1, wherein the arrhythmia is atrial fibrillation. 2. The method of claim 1, wherein the arrhythmia-related disorder is selected from cerebral infarction, stroke, and cerebral hemorrhage. The genetic variations included in the data list are rs202030113, rs2930856, rs1055894680, rs73205368, rs778479352, rs4970418, rs9782984, rs75414548, rs1933723, rs12512502, rs6841049, rs 17118812, rs7766436, rs12209223, rs4896104, rs2727757, rs17430357, rs17303101, rs11527634, rs76460895 , rs1769758, rs7126870, rs10500790, rs517938, rs10845620, rs2629755, rs1344543, rs11614295, rs11841562, rs1886512, rs9284324, rs8096658, rs11881441 , rs3746471, rs5754508, rs139557, rs73205368, rs1891095 (RS numbers are from the National Center for Biotechnology Information's dbSNP database. The method according to any one of claims 1 to 6, comprising one or more single nucleotide polymorphisms selected from (indicating the registration number of). A system for determining the risk of arrhythmia or arrhythmia-related disorders in a subject, the system comprising:
means for storing data including a data list including effect alleles and effect sizes of genetic variants associated with arrhythmia;
means for receiving genetic information of a subject;
Means for calculating an arrhythmia risk score from genetic information of a subject based on data regarding effect alleles and effect sizes of arrhythmia-related genetic variations in the data list, and presenting a risk of arrhythmia or arrhythmia-related disorder based on the risk score. means, including;
The data list is characterized in that the data list is a data list that integrates genome analysis results of Western populations and genome analysis results of non-Western populations through meta-analysis,
system. The genetic variations included in the data list are rs202030113, rs2930856, rs1055894680, rs73205368, rs778479352, rs4970418, rs9782984, rs75414548, rs1933723, rs12512502, rs6841049, rs 17118812, rs7766436, rs12209223, rs4896104, rs2727757, rs17430357, rs17303101, rs11527634, rs76460895 , rs1769758, rs7126870, rs10500790, rs517938,
rs10845620, rs2629755, rs1344543, rs11614295, rs11841562, rs1886512, rs9284324, rs8096658, rs11881441, rs3746471, rs5754508, rs139557, rs73205368, rs1891095 (the rs number indicates the registration number of the dbSNP database of the National Center for Biotechnology Information). 9. The system according to claim 8, comprising one or more single nucleotide polymorphisms. 10. The system according to claim 8 or 9, wherein the arrhythmia is atrial fibrillation. 10. The system according to claim 8 or 9, wherein the arrhythmia-related disorder is selected from cerebral infarction, stroke and cerebral hemorrhage. A method for determining the risk of arrhythmia or arrhythmia-related disorders, the method comprising:
rs202030113, rs2930856, rs1055894680, rs73205368, rs778479352, rs4970418, rs9782984, rs75414548, rs1933723, rs12512502, rs6841049, rs17118812, rs7 766436, rs12209223, rs4896104, rs2727757, rs17430357, rs17303101, rs11527634, rs76460895, rs1769758, rs7126870, rs10500790, rs517938, rs10845620, rs2629755, rs1344543, rs11614295, rs11841562, rs1886512, rs9284324, rs8096658, rs11881441, rs3746471, rs5754508, rs139557, rs73205368, rs1891095 ( , rs number indicates the accession number of the National Center for Biotechnology Information's dbSNP database)1 A method comprising the step of analyzing single nucleotide polymorphisms of more than one species. 13. The method of claim 12, wherein the arrhythmia is atrial fibrillation. 14. The method according to claim 12 or 13, wherein the arrhythmia-related disorder is selected from cerebral infarction, stroke and cerebral hemorrhage.

Images