Method and apparatus for calculating rare genetic variation polygenic risk scores

Abstract

One embodiment of the present disclosure provides a method for calculating polygenic risk scores. The method comprises the steps of: calculating a phenotype estimation value for genome data of an individual by using a pre-set first algorithm; classifying, into a plurality of groups according to a specific criteria, rare genetic variations for each gene included in the genome data, and estimating the size of the effect of the rare genetic variations by using a pre-set second algorithm based on a genotype matrix for each of the plurality of groups and the phenotype estimation value; and calculating a rare genetic variation polygenic risk score by using a pre-set third algorithm based on the genotype matrix for each of the plurality of groups and the size of the effect of the rare genetic variations.

Description

Method and device for calculating a polygenic risk score for rare genetic variants

[0001] The present invention relates to a method and device for calculating a polygenic risk score for a rare genetic variant, and more particularly, to a method and device for calculating an effect size of a rare genetic variant and calculating a polygenic risk score for a rare genetic variant based on the effect size of the rare genetic variant.

[0002] Genetic variations can be classified into common and rare genetic variations based on their frequency of occurrence within a population.

[0003] Current genome-based disease prediction models only use information on common genetic variants because it is difficult to accurately estimate the effects of rare genetic variants on disease using existing genome-wide association study (GWAS) techniques.

[0004] However, rare genetic variants are known to have a much larger effect size on disease than common genetic variants, and they often have a significant functional impact on disease, so estimating the effect size of rare genetic variants is essential for improving the performance of disease prediction models.

[0005] Burden test, SKAT, and SKAT-O are widely used methods to analyze the association of rare genetic variants.

[0006] In the case of the burden test, it is a method that can statistically test the effect size and whether there is an association under the assumption that all genetic variants within the same gene have the same effect size, but the basic assumption is very far from reality, so there is a problem in that it is difficult to see it as accurately representing the effect size of each genetic variant.

[0007] Additionally, methods such as SKAT and SKAT-O can test whether there is an association under assumptions that are closer to reality by relaxing these assumptions, but they have the limitation of not being able to measure the effect size.

[0008] Therefore, there is a need for a method that can more accurately measure the effect size of rare genetic variants under realistic assumptions and derive a polygenic risk score for rare genetic variants.

[0009] The present invention is intended to solve the problems of the above-mentioned prior art, and to provide a method and device for calculating the effect size of a rare genetic variant and calculating a rare genetic variant polygenic risk score based on the effect size of the rare genetic variant.

[0010] The technical problems to be solved by the present invention are not limited to the technical problems described above, and other technical problems of the present invention can be derived from the following description.

[0011] As a technical means for solving the above-described technical problem, an embodiment according to a first aspect of the present disclosure provides a method for calculating a polygenic risk score. The method includes the steps of calculating an estimate of a phenotype for individual genomic data using a first preset algorithm, classifying rare genetic variants for each gene included in the genomic data into a plurality of groups according to specific criteria, and estimating an effect size of the rare genetic variant using a second preset algorithm based on a genotype matrix for each of the plurality of groups and the estimate of the phenotype, and calculating a polygenic risk score of the rare genetic variant using a third preset algorithm based on the genotype matrix for each of the plurality of groups and the effect size of the rare genetic variant.

[0012] As a technical means for solving the above-described technical problem, an embodiment according to a second aspect of the present disclosure provides a multi-gene risk score calculation device. The device includes a communication module, at least one processor, and a memory electrically connected to the processor and storing at least one code to be executed by the processor, wherein the memory stores a code that, when executed through the processor, causes the processor to calculate an estimate of a phenotype for individual genomic data using a first preset algorithm, classify rare genetic variants for each gene included in the genomic data into a plurality of groups according to a specific criterion, estimate an effect size of the rare genetic variant using a second preset algorithm based on the genotype matrix for each of the plurality of groups and the estimate of the phenotype, and calculate a rare genetic variant multi-gene risk score using a third preset algorithm based on the genotype matrix for each of the plurality of groups and the effect size of the rare genetic variant.

[0013] According to the present invention, the method can reflect actual biological characteristics by excluding extreme assumptions used in the Burden test-based method used to estimate the effect size of existing rare genetic variants.

[0014] In addition, according to the present invention, the effect size can be reliably estimated using the empirical Bayes method, which is a shrinkage method.

[0015] In addition, according to the present invention, Firth bias correction can be additionally performed for binary variables to correct the effect size.

[0016] In addition, according to the present invention, the effect size can be analyzed in a short time even for large-scale genomic data by using the singular value decomposition used in FaST-LMM.

[0017] The effects of the present invention are not limited to the effects described above, and include all effects understood from the following description.

[0018] FIG. 1 is a drawing illustrating a server and a terminal connected to the server in communication with one embodiment of the present invention.

[0019] Figure 2 is a drawing showing the detailed configuration of the server illustrated in Figure 1.

[0020] Figures 3 and 4 are diagrams illustrating an example of the form of genetic data.

[0021] Figures 5 to 7 are diagrams illustrating examples of measuring the effect size of rare genetic variants on Type 2 diabetes.

[0022] Figure 8 is a flowchart illustrating the sequence of a method for estimating the effect size of a rare genetic variant according to another embodiment of the present invention.

[0023] Figure 9 is a flowchart illustrating the order of detailed steps of the method for estimating the effect size of a rare genetic variant illustrated in Figure 8.

[0024] Hereinafter, the present disclosure will be described in detail with reference to the attached drawings. However, the present disclosure may be implemented in various different forms and is not limited to the embodiments described herein. In addition, the attached drawings are only intended to facilitate understanding of the embodiments disclosed in the present specification, and the technical concepts disclosed in the present specification are not limited by the attached drawings. All terms, including technical and scientific terms, used herein should be interpreted as having meanings generally understood by a person of ordinary skill in the technical field to which the present disclosure pertains. Terms defined in the dictionary should be interpreted as having additional meanings consistent with the relevant technical literature and the present disclosure, and shall not be interpreted in an extremely ideal or restrictive sense unless otherwise defined.

[0025] To clearly explain the present disclosure in the drawings, parts irrelevant to the description have been omitted, and the size, shape, and appearance of each component shown in the drawings may vary. Identical / similar parts throughout the specification are designated with identical / similar drawing reference numerals.

[0026] The suffixes "module" and "part" used in the following description for components are assigned or used interchangeably solely for the convenience of writing the specification, and do not in themselves have distinct meanings or roles. Furthermore, in describing the embodiments disclosed herein, detailed descriptions of related known technologies have been omitted if they are deemed to obscure the gist of the embodiments disclosed herein.

[0027] Throughout the specification, when a part is said to be "connected (connected, in contact with, or coupled)" to another part, this includes not only cases where it is "directly connected (connected, in contact with, or coupled)" but also cases where it is "indirectly connected (connected, in contact with, or coupled)" with another member in between. Furthermore, when a part is said to "include (have or provide)" a certain component, this does not mean that it excludes other components, but rather that it may "include (have or provide)" other components, unless otherwise specifically stated.

[0028] As used herein, ordinal terms such as "first," "second," etc., are used solely to distinguish one component from another and do not limit the order or relationship of the components. For example, the first component of the present disclosure may be referred to as the "second component," and similarly, the second component may also be referred to as the "first component." As used herein, singular forms should be construed to include plural forms, unless explicitly stated otherwise.

[0029] FIG. 1 is a drawing illustrating a server and a terminal connected to the server in communication with one embodiment of the present invention.

[0030] Referring to FIG. 1, the server (100) can be connected to a terminal (200) through a preset communication network.

[0031] The server (100) may be a device for calculating a multi-gene risk score of a rare genetic variant.

[0032] The server (100) uses a preset first algorithm to calculate an estimate of a phenotype for an individual's genetic data.

[0033] The server (100) classifies rare genetic variants for each gene included in the genome data into multiple groups according to specific criteria, and estimates the effect size of the rare genetic variants using a preset second algorithm based on the genotype matrix and phenotype estimates for each of the multiple groups.

[0034] The server (100) calculates a rare genetic variant polygenic risk score using a predetermined third algorithm based on a plurality of group-specific genotype matrices and the effect size of rare genetic variants.

[0035] The terminal (200) can transmit individual genetic data including at least one of individual disease diagnosis, age, gender, and genetic factors to the server (100), and receive a rare genetic variant polygenic risk score from the server (100).

[0036] The terminal (200) may refer to any type of handheld-based wireless communication device such as a notebook, desktop, laptop, wireless communication device with portability and mobility, or a smartphone, tablet PC, etc. equipped with a web browser.

[0037] Figure 2 is a drawing showing the detailed configuration of the server illustrated in Figure 1.

[0038] Referring to FIG. 2, the server (100) may include a communication module (110), a processor (120), and a memory (130).

[0039] The communication module (110) may include a device including hardware and software necessary to transmit and receive signals such as control signals or data signals through wired or wireless connections with other network devices.

[0040] The communication module (110) can receive personal genetic data including at least one of individual disease diagnosis, age, gender, and genetic factors from the terminal, and transmit a rare genetic variant polygenic risk score to the terminal.

[0041] The processor (120) may include various types of devices that control and process data. The processor (120) may refer to a data processing device built into hardware that has a physically structured circuit to perform a function expressed in code or instructions included in a program.

[0042] In one example, the processor (120) may be implemented in the form of a microprocessor, a central processing unit (CPU), a processor core, a multiprocessor, an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), etc., but the scope of the present invention is not limited thereto.

[0043] The processor (120) performs operations according to the code stored in the memory (130).

[0044] The memory (130) can store at least one of information and data input to the communication module (110), information and data required for functions performed by the processor (120), and data generated according to the execution of the processor (120).

[0045] Memory (130) should be interpreted as a general term for a non-volatile storage device that maintains stored information even when no power is supplied and a volatile storage device that requires power to maintain stored information. In addition to a volatile storage device that requires power to maintain stored information, memory (130) may include cloud storage, SSD, magnetic storage media, or flash storage media, but the scope of the present invention is not limited thereto.

[0046] The memory (130) is electrically connected to the processor (120) and stores at least one code to be executed by the processor (120). The memory (130) stores code that, when executed by the processor (120), causes the processor (120) to perform the following functions and procedures.

[0047] The memory (130) stores code that causes a phenotype estimate for an individual's genetic data to be calculated using a preset first algorithm. For example, the memory (130) may store code that causes a phenotype estimate to be calculated using a generalized linear mixed model (GLMM).

[0048] Here, the genetic data may include at least one of a personal identification number, age, gender, genetic information, and disease status.

[0049] The memory (130) may store code that causes the estimation of the phenotype to be produced by taking into account covariates, genetic relationships and errors for at least one of age, gender and genetic factors.

[0050] The preset first algorithm can have different forms depending on whether the phenotype is a continuous variable or a binary variable, as in mathematical expression 1.

[0051]

[0052]

[0053] Here, y is a phenotype, X is at least one covariate among age, sex, and genetic factors, a is a coefficient, and b is It is a random effect term that follows the distribution and represents the genetic relationship, and ε is It can mean an error term that follows a distribution.

[0054] For example, continuous variables are variables that can be represented as continuous numbers, such as height, weight, and age, and binary variables are variables that have one of two possible answers, such as gender and presence or absence of a disease.

[0055] In the memory (130), a code is stored that causes rare genetic variations for each gene included in the genomic data to be classified into a plurality of groups according to specific criteria, and to estimate the effect size of the rare genetic variations using a preset second algorithm based on the genotype matrix and phenotype estimates for each of the plurality of groups.

[0056] The memory (130) may store a code that causes rare genetic mutations to be classified into multiple groups according to specific criteria and to form a genotype matrix for each of the multiple groups.

[0057] For example, the memory (130) may store a code that causes a rare genetic variant to be classified into a plurality of groups based on the risk of the rare genetic variant. For example, the memory (130) may store a code that causes a rare genetic variant to be classified into one of a loss-of-function (LoF), a missense variant, and a synonymous variant and to form a genotype matrix. Here, the genotype matrix is , and It could be.

[0058] The memory (130) may store a code that causes genes included in the genomic data to be classified as ultra-rare genetic variants when the number of alleles is less than a preset number, and to group the ultra-rare genetic variants into a single rare genetic variant. For example, ultra-rare genetic variants having an allele count of less than 10 may be grouped into a single genetic variant by applying a technique (collapsing) that causes the output values ​​to converge to a limited range during the model learning process or outputs only limited values.

[0059] The memory (130) may store a code that causes the effect size of a rare genetic variant to be estimated using a second algorithm based on the genotype matrix and the phenotype estimate.

[0060] The second algorithm may be mathematical formula 2.

[0061]

[0062]

[0063]

[0064] Here, for continuous variables And for binary variables and, , and can be random effect terms representing the effect size of each genetic variant.

[0065] , and Each one is It can follow the distribution. Here, ∑ can be any variance-covariance matrix, but in the experiments conducted in this method, it was assumed to be the identity matrix I.

[0066] Accordingly, the effect size of rare genetic mutations can be estimated as in mathematical equation 3.

[0067]

[0068] Here, G is the genotype matrix, G T is the transpose matrix of the genotype matrix, ∑ is the variance-covariance matrix of the prior distribution of the effect size, and τ is the variance component estimated through Equation 2. Is It can mean the dispersion of .

[0069] The memory (130) may store a code that causes the effect size to be re-estimated based on a Firth bias correction method when the effect size is greater than a preset threshold value.

[0070] In the memory (130), a code that causes the effect size to be re-estimated through a Firth bias correction method with an L2 penalty term added as in Equation 4 can be stored when the effect size is greater than a preset threshold value.

[0071]

[0072] In the memory (130), a code is stored that causes a rare variant polygenic risk score (RVPRS) to be calculated according to mathematical formula 5 based on the effect size of the rare genetic variant.

[0073]

[0074] The memory (130) may store a code that causes a final polygenic risk score to be calculated by assigning preset weights to each of a polygenic risk score for a rare genetic variant and a polygenic risk score (PRS) for a common genetic variant, and to create a disease prediction model that predicts a disease based on the final polygenic risk score.

[0075] Figures 3 and 4 are diagrams illustrating examples of the format of genetic data. More specifically, Figure 3 illustrates an example of the format of data corresponding to covariates, and Figure 4 illustrates an example of the format of data corresponding to genetic information.

[0076] Referring to Figures 3 and 4, data corresponding to covariates may include information on individual disease diagnosis, age, gender, and genetic PC. Here, the genetic PC may be a score that roughly represents the individual's race.

[0077] Data corresponding to genetic information may include information about each genetic variant. For example, each column in Figure 4 may represent a genetic variant.

[0078] Figures 5 to 7 are diagrams illustrating examples of measuring the effect size of rare genetic variants on Type 2 diabetes.

[0079] Referring to FIGS. 5 to 7, the multi-gene risk score calculation device can calculate the effect of covariates on phenotypes under a Null model in which only an intercept value is present without any input variables.

[0080] Since this example deals with the presence or absence of a disease diagnosis, which is a binary variable, the effect of the covariate on the phenotype can be calculated through a logistic mixed model as in Equation 6.

[0081]

[0082] Here, g(x) can be a logit function.

[0083] For example, the person with ID 1 shown in Figures 3 and 4 is Y i =1, X i =(63.2, -12.7523, 5.51758, -2.76956) is a vector, and K is a matrix representing the genetic relationships between each person, which can be an nXn-dimensional matrix. Here, Y i is the diagnosis of the disease for the i-th person. If it is 1, it means the person has the disease, and if it is 0, it means the person does not have the disease.

[0084] In general, the diagonal component represents the degree of relationship with oneself, where a close kinship relationship can be represented as 1 and a relationship between people who are not close kinship can be represented as 0. For example, if the ith person and the jth person are close kin, the number representing the degree of relationship between them can be reflected in the (i,j) and (j,i) coordinates of K.

[0085] K can be expressed as in mathematical formula 7 below.

[0086]

[0087] The polygenic risk score calculation device calculates g, Y according to the first algorithm i , X i , perform regression analysis using K, and accordingly a, b i , can be estimated.

[0088] The polygenic risk score calculator estimates a, b i , Using , an estimate of the phenotype for each individual is can be produced.

[0089] Here, the phenotype estimate for each individual can be estimated by excluding genetic factors such as age and gender. The estimated μ i Each represents a different value and can have a form like mathematical expression 8.

[0090]

[0091] The polygenic risk score calculator estimates μ i Using , the effect size of genetic variation is can be produced.

[0092] The effect size of a genetic variant can have a form similar to the mathematical expression 9 below.

[0093]

[0094] Assuming that the effect size of a rare genetic variant for a specific gene j within genomic data is estimated, a polygenic risk score generator can classify rare genetic variants into multiple groups based on specific criteria. For example, genetic variants can be classified into at least one of the loss-of-function, missense, and synonymous groups.

[0095] As shown in Figure 5, in the genotype matrix, assuming that variants 1 and 4 are loss-of-function, variants 2 and 6 are missense, and variants 3 and 5 are synonymous, , and Each can be configured as shown in FIG. 6(a), FIG. 6(b), and FIG. 6(c).

[0096] The multigenic risk score calculation device is in Equation 9. and , and The effect size of each mutation can be estimated according to mathematical formula 2.

[0097] For convenience of explanation, in mathematical expression 2 , and is a prior distribution with a mean of 2 and variances of , and It was assumed that it follows a multivariate normal distribution, and ∑s can be any variance-covariance matrix, but for convenience of explanation, it was assumed to be the identity matrix I.

[0098] The polygenic risk score calculation device can estimate τ and σ by performing regression analysis on three groups. The polygenic risk score calculation device can estimate the effect size of each genetic variant by substituting the estimated τ and σ into Equation 3.

[0099] The effect size of each genetic variant may be as shown in Figure 7(a).

[0100] The polygenic risk score calculation device can be replaced by re-estimating the effect size through Perth bias correction when the absolute value of the effect size exceeds a certain threshold.

[0101] The polygenic risk score calculation device is a critical point. In the case of variant 4 shown in Figure 7(a), since the effect size is greater than the critical point, the multigenic risk score calculation device can re-estimate and update the effect size for variant 4.

[0102] The effect size of each genetic variant reflecting the re-estimated effect size may be as shown in Figure 7(b).

[0103] The polygenic risk score calculator can use the calculated effect sizes to calculate a rare variant polygenic risk score (RVPRS).

[0104] When the set of genes statistically significantly associated with the disease being analyzed is J, the genetic variant polygenic risk score can be calculated by calculating the RVPRS of person i for each gene (j) included in the set and then adding them, as in Equation 5.

[0105] For example, when calculating the RVPRS of gene j for a person with ID=1, the genotype is (0,0,0,0,0,2), and the effect size of each variant is (-0.457, -0.137, 0.283, 1.382, 0.089, -0.096), so the RVPRS i,j It can be calculated as =0x(-0.457)+0x(-0.137)+ 0x(0.283)+ 0x(1.382) +0x(0.089)+ 2x(-0.096)=-0.192.

[0106] Assuming that there are 10 genes significantly associated with diabetes, the multi-gene risk score calculation device can calculate the RVPRS for each gene and then sum them for the 10 genes.

[0107] The polygenic risk score generator can build a disease prediction model by integrating the RVPRS and the PRS calculated using common variants.

[0108] Figure 8 is a flowchart illustrating the sequence of a method for estimating the effect size of a rare genetic variant according to another embodiment of the present invention.

[0109] The method for estimating the effect size of a rare genetic variant to be described below can be performed by the method for estimating the effect size of a rare genetic variant described above with reference to FIGS. 1 to 7 or by the server (100). Therefore, the contents of the embodiments of the present disclosure described above with reference to FIGS. 1 to 7 can be equally applied to the embodiments to be described below, and any duplicated contents with the above description will be omitted below. The steps described below do not necessarily have to be performed in order, and the order of the steps can be set in various ways, and the steps can be performed almost simultaneously.

[0110] Referring to Fig. 8, the method for estimating the effect size of a rare genetic variant includes a step of calculating an estimate of a phenotype (S100), a step of estimating the effect size of a rare genetic variant (S200), and a step of calculating a polygenic risk score of a rare genetic variant (S300).

[0111] The phenotype estimate calculation step (S100) is a step of calculating a phenotype estimate for an individual's genetic data using a preset first algorithm.

[0112] The step of calculating an estimate of a phenotype (S100) may include a step of calculating an estimate of a phenotype by considering at least one covariate among age, sex, and genetic factors, a genetic relationship, and an error.

[0113] The step of estimating the effect size of a rare genetic variant (S200) is a step of classifying rare genetic variants for each gene in the genome data into multiple groups according to specific criteria, and estimating the effect size of the rare genetic variant using a preset second algorithm based on the genotype matrix and phenotype estimates for each of the multiple groups. Here, the specific criteria may be set based on the risk level of the rare genetic variant.

[0114] The step of calculating a rare genetic variant polygenic risk score (S300) may include a step of calculating a rare genetic variant polygenic risk score using a predetermined third algorithm based on a plurality of group-specific genotype matrices and the effect size of the rare genetic variant.

[0115] The method for estimating the effect size of a rare genetic variant may further include a step of assigning preset weights to each of a polygenic risk score for a rare genetic variant and a polygenic risk score for a common genetic variant to produce a final polygenic risk score, and generating a disease prediction model that predicts a disease based on the final polygenic risk score.

[0116] Figure 9 is a flowchart illustrating the order of detailed steps of the method for estimating the effect size of a rare genetic variant illustrated in Figure 8.

[0117] Referring to FIG. 9, the effect size estimation step (S200) of a rare genetic variant may include a genotype matrix construction step (S210), an effect size estimation step of a rare genetic variant (S220), and an effect size re-estimation step of a rare genetic variant (S230).

[0118] The genotype matrix construction step (S210) may be a step of classifying rare genetic variants into multiple groups according to specific criteria and constructing a genotype matrix for each of the multiple groups.

[0119] The genotype matrix construction step (S210) may include a step of classifying genes as ultra-rare genetic variants when the number of alleles of genes included in the genome data is less than a preset number, and grouping ultra-rare genetic variants into one rare genetic variant.

[0120] The step of estimating the effect size of a rare genetic variant (S220) may be a step of estimating the effect size of a rare genetic variant using a second algorithm based on the genotype matrix and phenotype estimates.

[0121] The step of re-estimating the effect size of a rare genetic variant (S230) may be a step of re-estimating the effect size based on the Firth bias correction method when the effect size is greater than a preset threshold value.

[0122] Those skilled in the art will appreciate that the present disclosure can be easily modified into other specific forms based on the above description without changing the technical spirit or essential characteristics of the present disclosure. Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive. The scope of the present disclosure is indicated by the following claims, and all changes or modifications derived from the meaning and scope of the claims and their equivalents should be construed as being included in the scope of the present disclosure. The scope of the present application is indicated by the following claims rather than the above detailed description, and all changes or modifications derived from the meaning and scope of the claims and their equivalents should be construed as being included in the scope of the present application.

[0123] The form for carrying out the invention is substantially the same as the best form for carrying out the invention.

[0124] The present invention can be used to predict diseases, and thus has industrial applicability.

Claims

1. In a method performed by a processor, a) A step of calculating an estimate of a phenotype for an individual's genetic data using a preset first algorithm; b) a step of classifying rare genetic variants for each gene included in the above genome data into multiple groups according to specific criteria, and estimating the effect size of the rare genetic variants using a preset second algorithm based on the genotype matrix for each of the multiple groups and the phenotype estimate; and c) A method for calculating a multi-gene risk score, comprising the step of calculating a multi-gene risk score for the rare genetic variant using a predetermined third algorithm based on the plurality of group-specific genotype matrices and the effect size of the rare genetic variant.

2. In paragraph 1, A method for calculating a polygenic risk score, wherein the above genetic data includes at least one of a personal identification number, age, gender, genetic information, and disease presence.

3. In paragraph 1, A method for calculating a multi-gene risk score, wherein the above-mentioned specific criteria are set based on the risk of the above-mentioned rare genetic mutation.

4. In paragraph 1, Step a) above, A method for calculating a polygenic risk score, comprising the step of calculating an estimate of the phenotype by taking into account at least one covariate of age, sex, and genetic factors, a genetic relationship, and an error.

5. In paragraph 1, The above first algorithm is, If the above phenotype is a continuous variable is in the form of formula 1, If the above phenotype is a binary variable, It is in the form of formula 2, Here, y is the phenotype, X is a covariate, a is a coefficient, b is a term representing a genetic relationship, and ε is an error term, a method for calculating a polygenic risk score.

6. In paragraph 1, Step b) above, b-1) A step of classifying the above rare genetic variants into the plurality of groups according to the specific criteria and constructing a genotype matrix for each of the plurality of groups; b-2) a step of estimating the effect size of the rare genetic variant using the second algorithm based on the genotype matrix and the phenotype estimate; and b-3) A method for calculating a multigenic risk score, comprising a step of re-estimating the effect size based on a Firth bias correction method when the effect size is greater than a preset threshold value.

7. In paragraph 6, The second algorithm above is , , is in the form of formula 3, Here, for continuous variables And for binary variables And, , and A method for calculating a multigenic risk score, wherein each term represents the effect size of the rare genetic variant, and ε is an error term.

8. In paragraph 6, Step b-1) above, A method for calculating a multi-gene risk score, comprising the steps of classifying the genes as ultra-rare genetic variants when the number of alleles of genes included in the above genetic data is less than a preset number, and grouping the ultra-rare genetic variants into one rare genetic variant.

9. In paragraph 1, The third algorithm above is is in the form of formula 4, Here, is the genotype matrix, A method for calculating a multigenic risk score, wherein the effect size of the above rare genetic variant is 10. In paragraph 1, d) A method for calculating a polygenic risk score, further comprising the step of calculating a final polygenic risk score by assigning preset weights to each of the polygenic risk score for the rare genetic variant and the polygenic risk score for the common genetic variant, and generating a disease prediction model that predicts a disease based on the final polygenic risk score.

11. Communication module; at least one processor; and A memory electrically connected to the processor and storing at least one code to be executed by the processor, The above memory, when executed through the processor, causes the processor to: A multi-gene risk score calculation device storing code that causes a phenotype estimate for an individual's genomic data to be calculated using a preset first algorithm, rare genetic variants for each gene included in the genomic data to be classified into a plurality of groups according to specific criteria, an effect size of the rare genetic variant to be estimated using a preset second algorithm based on the genotype matrix for each of the plurality of groups and the phenotype estimate, and a rare genetic variant multi-gene risk score to be calculated using a preset third algorithm based on the genotype matrix for each of the plurality of groups and the effect size of the rare genetic variant.

12. In paragraph 11, A multi-gene risk score calculation device, wherein the above genetic data includes at least one of a personal identification number, age, gender, genetic information, and disease presence.

13. In paragraph 11, A multi-gene risk score calculation device, wherein the above specific criteria are set based on the risk of the rare genetic mutation.

14. In paragraph 11, The above memory causes the processor to: A multigenic risk score calculation device storing code that causes the device to calculate an estimate of the phenotype by taking into account at least one covariate of age, sex, and genetic factors, genetic relationships, and errors.

15. In paragraph 11, The above first algorithm is, If the above phenotype is a continuous variable is in the form of formula 1, If the above phenotype is a binary variable, It is in the form of formula 2, A multigenic risk score calculation device, wherein y is the phenotype, X is a covariate, a is a coefficient, b is a term representing a genetic relationship, and ε is an error term.

16. In paragraph 11, The above memory causes the processor to: A multi-gene risk score calculation device that stores code that classifies the rare genetic variants into the plurality of groups according to the specific criteria, constructs a genotype matrix for each of the plurality of groups, estimates the effect size of the rare genetic variant using the second algorithm based on the genotype matrix and the phenotype estimate, and causes the effect size to be re-estimated based on a Firth bias correction method when the effect size is greater than a preset threshold value.

17. In paragraph 16, The second algorithm above is , , is in the form of formula 3, Here, for continuous variables And for binary variables And, , and A multi-gene risk score calculation device, wherein each of the terms represents the effect size of the rare genetic variant, and the ε is an error term.

18. In paragraph 16, The above memory causes the processor to: A multi-gene risk score calculation device that stores a code that causes the genes included in the above genetic data to be classified as ultra-rare genetic variants when the number of alleles of the genes is less than a preset number and to group the ultra-rare genetic variants into one rare genetic variant.

19. In paragraph 11, The third algorithm above is is in the form of formula 4, Here, is the genotype matrix, A multi-gene risk score calculation device, which is the effect size of the above rare genetic variant.

20. In paragraph 11, The above memory causes the processor to: A polygenic risk score calculation device storing a code that causes a final polygenic risk score to be calculated by assigning preset weights to each of the polygenic risk scores for the rare genetic variants and the polygenic risk scores for the common genetic variants, and to create a disease prediction model that predicts a disease based on the final polygenic risk score.

Images