Construction method and device of sequence deep learning model for natural population multi-gene risk scoring, scoring method and device and electronic equipment

Abstract

The invention discloses a construction method and device of a sequence deep learning model for natural population multi-gene risk scoring, a scoring method and device and electronic equipment, and relates to the technical field of gene detection and analysis. According to the construction method of the model, upstream and downstream sequence genotypes, distances and correlation coefficients of target variation are innovatively and comprehensively incorporated into the deep learning model, variation information can be effectively enriched, the LD relation can be effectively considered, the accuracy of software can be improved, and application of the model in various natural crowd queues is facilitated. According to the method, genetic variation is screened by using LASSO regression, so that the influence of a strict threshold value is reduced, variation which possibly has effects is brought into the method as much as possible, and the reliability of a subsequent sequence model is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of gene detection and analysis, and in particular to a method for constructing a sequence deep learning model for polygenic risk scoring of a natural population, a scoring method, a device and an electronic device. Background Art

[0002] Polygenic Risk Scores (PRS) play an important role in the early screening of complex diseases and precision medicine. By comprehensively analyzing the numerous tiny variations in an individual's genetic code, a comprehensive and sophisticated risk assessment system is constructed, opening up a new path for the prediction and prevention of complex diseases such as chronic respiratory diseases and lung cancer.

[0003] In the process of constructing polygenic risk scores, the conventional practice is to identify genetic variants associated with chronic respiratory diseases based on the summary data of Genome-Wide Association Study (GWAS). However, the statistical significance threshold usually used in GWAS is very strict, which means that only some variants with strong signals will be included, while ignoring a large number of potential weak effect variants, which are not significant individually, but may have a significant contribution to the disease risk when combined.

[0004] In view of this, how to more comprehensively capture genetic information and improve the predictive ability of risk scores has become a research focus. Accurate polygenic risk scores are helpful for early screening of complex diseases in natural populations and precision medicine. It is crucial to develop an algorithm that can efficiently and accurately analyze polygenic risk scores in natural populations. This method requires powerful bioinformatics analysis capabilities to accurately capture the connection with clinical phenotypes from massive genomic variation data.

[0005] In view of this, the present invention is proposed. Summary of the invention

[0006] The purpose of the present invention is to provide a method for constructing a sequence deep learning model for polygenic risk scoring of a natural population, a scoring method, a device and an electronic device to solve the above-mentioned technical problems. In order to more comprehensively capture genetic information and improve the predictive ability of risk scoring, the present invention adopts the LASSO regression (Least Absolute Shrinkage and Selection Operator) method. LASSO regression is a variable screening technology that can not only process high-dimensional data, but also automatically shrink (or completely eliminate) those coefficients that contribute less to the model through regularization penalty terms, thereby retaining important variations. This method allows more variations that are less statistically significant but may still have certain biological significance to be included while controlling the complexity of the model, thereby improving the sensitivity and accuracy of polygenic risk scoring.

[0007] Compared with the strict threshold of whole genome association study, using LASSO regression to screen genetic variants associated with specific diseases will screen out more genetic variants, most of which are in linkage disequilibrium (LD). When it comes to polygenic risk scores, the effects of genetic variants in linkage disequilibrium (LD) may overlap. In addition, different genetic structures and different LD relationships exist between different natural populations, which is also the main reason why the PRS model cannot be universally used in various populations. Traditional deep neural networks (DNNs) and convolutional neural networks (CNNs) usually ignore the role of LD when dealing with genetic variants.

[0008] The inventors found that the use of sequence neural networks can effectively capture the correlation in sequence data, including LD between multiple variants. By using sequence neural networks, it is possible to better utilize the genome datasets of various natural population cohorts and build more accurate polygenic risk scoring models.

[0009] In addition, the present invention innovatively provides a new genetic variation data frame that integrates multiple SNPs of each genetic variation, each genetic variation data frame includes genotype information of multiple SNPs upstream and downstream of each target genetic variation, LD correlation coefficient of each natural population variation, and distance from the target genetic variation. Comprehensively incorporating the upstream and downstream sequence genotypes, distances, and correlation coefficients of the target variation into the deep learning model can effectively enrich variation information, effectively consider LD relationships, improve the accuracy of the software, and help the model be applied in a variety of natural population cohorts.

[0010] The present invention uses LASSO regression to screen genetic variations, which not only reduces the impact of strict thresholds, but also incorporates as many variations as possible with effects, thereby ensuring the reliability of subsequent sequence models.

[0011] The present invention is achieved in that:

[0012] In a first aspect, the present invention provides a method for constructing a sequence deep learning model for polygenic risk scoring of a natural population, comprising the following steps:

[0013] S1: LASSO regression is used to screen genetic variation data for preprocessed genotype data;

[0014] S2: Establishing a genetic variation window: Calculate the linkage disequilibrium (LD) correlation coefficient of each natural population variation based on the genetic variation data after LASSO regression in step S1; Establish a genetic variation data frame of X SNPs for each genetic variation; Each genetic variation data frame includes the genotype information of X SNPs for each target genetic variation, the distance to the target genetic variation, and the LD correlation coefficient; The obtained n*X*Y data set is the genetic variation window, where n is the number of variations; The number of SNPs in each genetic variation data frame is X≤51; Y≥3;

[0015] X SNPs refer to the target variant, 0.5*(X-1) genetic variants upstream of the target variant, and 0.5*(X-1) genetic variants downstream of the target variant; and X SNPs are located within 250kb upstream and downstream of the target variant, and have a correlation coefficient r2>0.1 with the target variant;

[0016] S3: Construction of sequence deep learning model for polygenic risk score:

[0017] The genetic variation window obtained in step S2 is dimensionless, and then the dimensionless data is input into the sequential neural network for model training until the model reaches the preset convergence condition. The model is used for polygenic risk scoring.

[0018] In a second aspect, the present invention also provides a method for obtaining polygenic risk scores for natural populations based on a deep learning model, comprising: scoring the SNP data to be scored using a sequence deep learning model constructed based on the above-mentioned method for constructing a sequence deep learning model for polygenic risk scoring.

[0019] In a third aspect, the present invention also provides a device for polygenic risk scoring of a natural population, comprising: an input module, a control module and an output module;

[0020] The input module is configured to: input SNP sample data and / or sample genotype data;

[0021] The control module includes: a genotype data preprocessing module, a genetic variation data screening module, a genetic variation window construction module, a model construction module and a data scoring module;

[0022] The genotype data preprocessing module is configured to: perform file conversion on the SNP sample data and / or the sample genotype data;

[0023] The genetic variation data screening module is configured to: use LASSO regression to screen genetic variation data;

[0024] The genetic variation window construction module is configured as follows: for the genetic variation data after LASSO regression, the linkage disequilibrium (LD) correlation coefficient of each natural population variation is calculated; a genetic variation data frame of X SNPs for each genetic variation is established; each genetic variation data frame includes the genotype information of X SNPs for each target genetic variation, the distance to the target genetic variation and the LD correlation coefficient; the n*X*Y data set obtained is the genetic variation window, where n is the number of variations; the number of SNPs in each genetic variation data frame is X≤51; Y≥3; the X SNPs refer to the target variation, 0.5*(X-1) genetic variations upstream of the target variation and 0.5*(X-1) genetic variations downstream of the target variation; and the X SNPs are located within 250kb upstream and downstream of the target variation, and the correlation coefficient with the target variation is r2>0.1;

[0025] The model building module is configured to: non-dimensionalize the acquired genetic variation window, and then input the non-dimensionalized data into the sequence neural network for model training until the model reaches a preset convergence condition;

[0026] The data scoring module is configured to: score the SNP data to be scored based on the sequence deep learning model constructed by the model building module;

[0027] The output module is configured to: output the polygenic risk score result.

[0028] In a fourth aspect, the present invention further provides an electronic device, comprising a memory and a processor, and computer instructions stored in the memory and executed on the processor, wherein when the computer instructions are executed by the processor, the steps of the above method are completed.

[0029] In a fifth aspect, the present invention further provides a computer-readable storage medium for storing computer instructions, which, when executed by a processor, complete the steps of the above method.

[0030] The present invention has the following beneficial effects:

[0031] (1) The present invention innovatively provides a new genetic variation data frame that integrates multiple SNPs of each genetic variation. Each genetic variation data frame includes genotype information of multiple SNPs upstream and downstream of each target genetic variation, LD correlation coefficient of each natural population variation, and distance to the target genetic variation. Incorporating the upstream and downstream sequence genotypes, distances, and correlation coefficients of the target variation into the deep learning model can effectively enrich variation information, effectively consider LD relationships, improve the accuracy of the software, and facilitate the application of the model in a variety of natural population cohorts.

[0032] (2) Sequential neural networks can effectively capture correlations in sequence data, including LD between multiple variants. By using sequential neural networks, we can better utilize genomic datasets from natural population cohorts and build more accurate polygenic risk scoring models.

[0033] (3) The present invention uses LASSO regression to screen genetic variations, which not only reduces the impact of strict thresholds, but also incorporates as many variations as possible that may have effects, thereby ensuring the reliability of subsequent sequence models. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0035] Figure 1 An overall flow chart of the method for constructing a sequence deep learning model for polygenic risk scoring in natural populations;

[0036] Figure 2 Develop schematics for genetic variation window data and models. DETAILED DESCRIPTION

[0037] References to embodiments of the present invention will now be provided in detail, one or more examples of which are described below. Each example is provided as an explanation rather than a limitation of the present invention. In fact, it will be apparent to those skilled in the art that various modifications and variations may be made to the present invention without departing from the scope or spirit of the present invention. For example, a feature illustrated or described as part of one embodiment may be used in another embodiment to produce a further embodiment.

[0038] Complex diseases, also known as polygenic diseases, refer to common chronic diseases that are difficult to cure medically and seriously affect the health of the people, whose phenotypes of related traits do not have an obvious Mendelian inheritance pattern and whose genetic mechanisms are relatively complex. They are not controlled by a pair of genes, but are coordinated by multiple genes and environmental factors. Common chronic diseases mainly include cardiovascular and cerebrovascular diseases, diabetes, malignant tumors, chronic respiratory diseases, arthritis, obesity, etc.

[0039] In a first aspect, the present invention provides a method for constructing a sequence deep learning model for polygenic risk scoring of a natural population, comprising the following steps:

[0040] S1: LASSO regression is used to screen genetic variation data for preprocessed genotype data;

[0041] S2: Establishing a genetic variation window: Calculate the linkage disequilibrium (LD) correlation coefficient of each natural population variation based on the genetic variation data after LASSO regression in step S1; Establish a genetic variation data frame of X SNPs for each genetic variation; Each genetic variation data frame includes the genotype information of X SNPs for each target genetic variation, the distance to the target genetic variation, and the LD correlation coefficient; The obtained n*X*Y data set is the genetic variation window, where n is the number of variations; The number of SNPs in each genetic variation data frame is X≤51; Y≥3;

[0042] X SNPs refer to the target variant, 0.5*(X-1) genetic variants upstream of the target variant, and 0.5*(X-1) genetic variants downstream of the target variant; and X SNPs are located within 250kb upstream and downstream of the target variant, and have a correlation coefficient r2>0.1 with the target variant;

[0043] S3: Construction of sequence deep learning model for polygenic risk score:

[0044] The genetic variation window obtained in step S2 is dimensionless, and then the dimensionless data is input into the sequential neural network for model training until the model reaches the preset convergence condition. The model is used for polygenic risk scoring.

[0045] In the above model construction method, a new genetic variation data frame integrating multiple SNPs of each genetic variation is provided, and each genetic variation data frame includes genotype information of multiple SNPs upstream and downstream of each target genetic variation, LD correlation coefficient of each natural population variation, and distance from the target genetic variation. Therefore, the upstream and downstream sequence genotypes, distances, and correlation coefficients of the target variation are comprehensively incorporated into the deep learning model, which can effectively enrich the variation information, effectively consider the LD relationship, improve the accuracy of the software, and help the model be applied in a variety of natural population cohorts.

[0046] The present invention uses a sequence neural network to effectively capture the correlation in the sequence data, including the LD between multiple variants. By using a sequence neural network, the genome datasets of various natural population cohorts can be better utilized, and a more accurate polygenic risk scoring model can be constructed.

[0047] In addition, the present invention uses LASSO regression to screen genetic variations, which not only reduces the impact of strict thresholds, but also incorporates as many variations as possible that may have effects, thereby ensuring the reliability of subsequent sequence models.

[0048] The size of the genetic variation window of the present invention takes into account a maximum of 25 genetic variation sites and genetic variations upstream and downstream of the target variation, thereby effectively enriching variation information and facilitating accurate analysis of polygenic risks.

[0049] When X SNPs are located within 250kb upstream and downstream of the target variation, and the correlation coefficient with the target variation is r2>0.1, and the number of genetic variations upstream or downstream of the target variation is less than 0.5*(X-1); the actual genetic variations within the upstream or downstream range of 250kb are added to the genetic variation window, and the remaining genetic variations upstream or downstream of the target variation are filled with identifiers that are different from the variant genotype (such as NA).

[0050] When X SNPs are located within 250kb upstream and downstream of the target variation, the correlation coefficient with the target variation r2>0.1, and the number of genetic variations upstream or downstream of the target variation is greater than 0.5*(X-1); then 0.5*(X-1) genetic variations within the upstream and downstream range of 250kb are added to the genetic variation window.

[0051] In a preferred embodiment of the present invention, these flanking variations (upstream and downstream variations) are located within 250 kb upstream and downstream, and have a correlation coefficient r with the target variation. 2 >0.1.

[0052] In a preferred embodiment of the present invention, the number of SNPs in each genetic variation data frame is: 11≤X≤51;

[0053] In a preferred embodiment of the present invention, the number of SNPs in each genetic variation data frame is: 21≤X≤51. For example, when X=31, 15 SNPs upstream or downstream of the target variation are added to the genetic variation window.

[0054] In a preferred embodiment of the present invention, if the number of genetic variants upstream of the target variant is less than 0.5*(X-1), it is filled with an identifier that is different from the variant genotype; if the number of genetic variants downstream of the target variant is less than 0.5*(X-1), it is filled with an identifier that is different from the variant genotype, such as NA. Ensure that all inputs have uniform traits.

[0055] In a preferred embodiment of the present invention, the genotype of each genetic variation data frame is numerically and / or alphabetically labeled;

[0056] In a preferred embodiment of the present invention, the genotype of each genetic variation data frame includes a major allele type, a heterozygote, a minor allele type, and an identifier that is distinguished from the variant genotype;

[0057] In a preferred embodiment of the present invention, when the SNP is located upstream of the target genetic variation, the distance to the target genetic variation is set to a negative value; when the SNP is located downstream of the target genetic variation, the distance to the target genetic variation is set to a positive value; when the SNP is the target genetic variation, the distance to the target genetic variation is set to 0.

[0058] In a preferred embodiment of the present invention, when Y≥4, each genetic variation data frame further includes other variation features of the X SNPs of each target genetic variation, and the other variation features are selected from at least one of variation type and variation length features.

[0059] In a preferred embodiment of the present invention, Y is 3 or 4. Y is the number of columns in the data frame, such as 3 or 4 columns, such as the first column is genotype information, the second column is the distance to the target variation, and the third column is the linkage disequilibrium correlation coefficient, in which case Y is 3. If each genetic variation data frame also includes the variation type of the X SNPs of each target genetic variation, the fourth column can be set as the variation type of the SNP, and Y is 4. If each genetic variation data frame also includes the variation type and variation length feature of the X SNPs of each target genetic variation, the fourth column can be set as the variation type of the SNP, the fifth column can be set as the variation length feature, and Y is 5.

[0060] In a preferred embodiment of the present invention, using LASSO regression to screen genetic variation data includes: using LASSO regression to minimize the following loss function:

[0061]

[0062] Where N is the number of samples; p is the number of genetic variants; y i is the target value of the ith observation; x ij is the value of the jth genetic variant of the ith observation; β iis the coefficient of the jth feature variable; β0 is the intercept term; λ is the regularization parameter;

[0063] In a preferred embodiment of the present invention, PLINK 1.90beta software is used to calculate the linkage disequilibrium (LD) correlation coefficient of each natural population variation;

[0064] In a preferred embodiment of the present invention, the sequence neural network is selected from a recurrent neural network, a Transform, a recursive neural network, a gated recursive unit, and a long short-term memory unit.

[0065] Recurrent Neural Networks (RNNs) include but are not limited to GRU and LSTM.

[0066] Taking the training of the GRU model as an example, after inputting dimensionless data, it passes through 3 layers of GRU plus an activation layer, and then passes through two layers of fully connected layers and activation layers to obtain the prediction result. Activation functions include but are not limited to Sigmoid function, Tanh activation function, and ReLU function. Then, the optimizer is gradually iterated to make the loss function value tend to the global minimum. Optimizers include but are not limited to stochastic gradient descent, SGD with momentum, and adaptive learning rate algorithm.

[0067] The polygenic risk score for natural populations refers to scoring the risk of at least one polygenic disease among cardiovascular and cerebrovascular diseases, diabetes, malignant tumors, chronic respiratory diseases, arthritis and obesity in natural populations;

[0068] In a preferred embodiment of the present invention, the polygenic risk score for a natural population refers to scoring the risk of chronic respiratory diseases in a natural population.

[0069] In a second aspect, the present invention also provides a method for obtaining polygenic risk scores for natural populations based on a deep learning model, comprising: scoring the SNP data to be scored using a sequence deep learning model constructed based on the above-mentioned method for constructing a sequence deep learning model for polygenic risk scoring.

[0070] In a third aspect, the present invention also provides a device for polygenic risk scoring, comprising: an input module, a control module and an output module;

[0071] The input module is configured to: input SNP sample data and / or sample genotype data;

[0072] The control module includes: a genotype data preprocessing module, a genetic variation data screening module, a genetic variation window construction module, a model construction module and a data scoring module;

[0073] The genotype data preprocessing module is configured to: perform file conversion on the SNP sample data and / or the sample genotype data;

[0074] The genetic variation data screening module is configured to: use LASSO regression to screen genetic variation data;

[0075] The genetic variation window construction module is configured as follows: for the genetic variation data after LASSO regression, the linkage disequilibrium (LD) correlation coefficient of each natural population variation is calculated; a genetic variation data frame of X SNPs for each genetic variation is established; each genetic variation data frame includes the genotype information of X SNPs for each target genetic variation, the distance to the target genetic variation and the LD correlation coefficient; the n*X*Y data set obtained is the genetic variation window, where n is the number of variations; the number of SNPs in each genetic variation data frame is X≤51; Y≥3; the X SNPs refer to the target variation, 0.5*(X-1) genetic variations upstream of the target variation and 0.5*(X-1) genetic variations downstream of the target variation; and the X SNPs are located within 250kb upstream and downstream of the target variation, and the correlation coefficient with the target variation is r2>0.1;

[0076] The model building module is configured to: non-dimensionalize the acquired genetic variation window, and then input the non-dimensionalized data into the sequence neural network for model training until the model reaches a preset convergence condition;

[0077] The data scoring module is configured to: score the SNP data to be scored based on the sequence deep learning model constructed by the model building module;

[0078] The output module is configured to: output the polygenic risk score result.

[0079] In a fourth aspect, the present invention further provides an electronic device, comprising a memory and a processor, and computer instructions stored in the memory and executed on the processor, wherein when the computer instructions are executed by the processor, the steps of the above method are completed.

[0080] Specifically, the electronic device may further include a bus and a communication interface, and the memory, processor and communication interface are electrically connected to each other directly or indirectly to achieve data transmission or interaction. For example, these elements may be electrically connected to each other via one or more buses or signal lines. The processor may process information and / or data related to target identification to perform one or more functions described in this application.

[0081] The memory can be, but is not limited to, random access memory (RAM), read only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable read-only memory (EEPROM), etc.

[0082] The processor can be an integrated circuit chip with signal processing capabilities. The processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0083] In a fifth aspect, the present invention further provides a computer-readable storage medium for storing computer instructions, which, when executed by a processor, complete the steps of the above method.

[0084] In order to make the purpose, technical scheme and advantages of the embodiments of the present invention clearer, the technical scheme in the embodiments of the present invention will be described clearly and completely below. If the specific conditions are not specified in the embodiments, they are carried out according to conventional conditions or conditions recommended by the manufacturer. If the manufacturer of the reagents or instruments used is not specified, they are all conventional products that can be purchased commercially.

[0085] The features and performance of the present invention are further described in detail below in conjunction with the embodiments.

[0086] Accurate polygenic risk scores can help early screening of complex diseases and precision medicine, and it is crucial to develop an algorithm that can efficiently and accurately analyze polygenic risk scores. This method requires powerful bioinformatics analysis capabilities to accurately capture the relationship with clinical phenotypes from massive genomic variation data.

[0087] Example 1

[0088] This example provides a method for constructing a sequence deep learning model for polygenic risk scoring of natural populations. Figure 1 The specific steps include:

[0089] 1. Data preprocessing

[0090] The present invention allows vcf files, ped / map files, and raw files when training and using models. vcf is a file format for storing variation information in genome sequences, which contains annotation information as the beginning of the file and variation information as the main body. map / ped files are associated and usually used together. The map file records variation position information, and the ped file records the pedigree and genotype information of the research object. The raw file is converted into a genotype file based on the additive effect. The first line of the file records the variation position information, and the rest is genotype data of 0, 1 and 2. If it is a vcf file and a ped / map file, use PLINK 1.90beta software [1] to convert it into a raw file for subsequent model training and use.

[0091] 2. Using LASSO regression to screen genetic variation data

[0092] This example uses LASSO regression to screen important genetic variations, reduce the complexity of subsequent models, and prevent overfitting. The optimization goal of Lasso regression is to minimize the following loss function:

[0093]

[0094] Where N is the number of samples; p is the number of genetic variants; y i is the target value of the ith observation; x ij is the value of the jth genetic variant of the ith observation; β j is the coefficient of the jth characteristic variable. When it is 0, x ij is deleted to achieve the effect of screening genetic variation; β0 is the intercept term; λ is the regularization parameter, which is used to control the strength of regularization;

[0095] 3. Establish genetic variation window:

[0096] The present invention uses PLINK 1.90beta software to calculate the LD correlation coefficient of each natural population variation based on the genetic variation data after LASSO regression.

[0097] This example proposes a window size of 51 SNPs, that is, considering 25 genetic variant sites and genetic variants upstream of the target variant. These flanking variants are located within 250 kb upstream and downstream and have a correlation coefficient r with the target variant. 2>0.1. If the number of flanking variants is less than 25, NA is filled to ensure that all inputs have uniform traits.

[0098] The genotype of each genetic variation data frame is numerically identified. In this embodiment, each variation has four numerical values, which are 0, 1, 2 and -9, respectively. 0 represents homozygous and major allele genotype, 1 represents heterozygous, 2 represents homozygous and minor allele genotype, and -9 represents NA. In other implementations, the genotype of each genetic variation data frame can also be alphabetically identified.

[0099] In addition, this embodiment also incorporates the LD correlation coefficient as a basis for subsequent models to adapt to various natural populations.

[0100] This embodiment numerically identifies the distance between the SNP and the target genetic variation and the LD correlation coefficient. When the SNP is located upstream of the target genetic variation, the distance to the target genetic variation is set to a negative value; when the SNP is located downstream of the target genetic variation, the distance to the target genetic variation is set to a positive value; when the SNP is the target genetic variation, the distance to the target genetic variation is set to 0. Set the upstream NA to -999999bp and the downstream NA to 999999bp. For the linkage disequilibrium correlation coefficient, NA is 0.

[0101] Finally, a 51*3 data frame was obtained for each variation (refer to Figure 2 ), the first column is the genotype information, the second column is the distance to the target variant, and the third column is the linkage disequilibrium correlation coefficient. Therefore, a data set of n*51*3 will be obtained, where n is the number of variants and 3 is the number of columns.

[0102] 4. Construct a sequence deep learning model for polygenic risk scoring in natural populations.

[0103] The deep learning network architecture is used to process genetic variation window data. First, the genetic variation window data is dimensionless to ensure the reliability of the subsequent model. If the variation n of the training data is greater than 50, the model input data will adopt a batch strategy to reduce the pressure of model training. Sequence models such as RNN (such as GRU and LSTM) and Transform can be used for model training.

[0104] This embodiment takes the training of the GRU model as an example. After inputting the dimensionless data, the present invention obtains the prediction result through three layers of GRU plus an activation layer, and then through two layers of fully connected layers and an activation layer.

[0105] The activation function includes but is not limited to Sigmoid function, Tanh activation function and ReLU function, etc. In this embodiment, the Sigmoid function is used.

[0106] Then, the optimizer is used to gradually iterate and make the loss function value approach the global minimum. The optimizer includes but is not limited to the stochastic gradient descent method, the SGD with momentum method, and the adaptive learning rate algorithm. The present embodiment uses the stochastic gradient descent method.

[0107] Example 2

[0108] This embodiment provides a device for natural population polygenic risk scoring, including: an input module, a control module and an output module;

[0109] The input module is configured to: input SNP sample data and / or sample genotype data;

[0110] The control module includes: a genotype data preprocessing module, a genetic variation data screening module, a genetic variation window construction module, a model construction module and a data scoring module;

[0111] The genotype data preprocessing module is configured to: perform file conversion on the SNP sample data and / or the sample genotype data;

[0112] The genetic variation data screening module is configured to: use LASSO regression to screen genetic variation data;

[0113] The genetic variation window construction module is configured as follows: Calculate the linkage disequilibrium (LD) correlation coefficient of each natural population variation based on the genetic variation data after LASSO regression; Establish a genetic variation data frame of 51 SNPs for each genetic variation; Each genetic variation data frame includes the genotype information of 51 SNPs for each target genetic variation, the distance to the target genetic variation, and the LD correlation coefficient; The n*51*3 data set is the genetic variation window, where n is the number of variations; The 51 SNPs are located within 250kb upstream and downstream of the target variation, and the correlation coefficient with the target variation r2>0.1;

[0114] The model building module is configured to: non-dimensionalize the acquired genetic variation window, and then input the non-dimensionalized data into the sequence neural network for model training until the model reaches a preset convergence condition;

[0115] The data scoring module is configured to: score the SNP data to be scored based on the sequence deep learning model constructed by the model building module;

[0116] The output module is configured to: output the polygenic risk score result.

[0117] Example 3

[0118] This embodiment provides a method for obtaining polygenic risk scores for natural populations based on a deep learning model, which includes: scoring SNP data to be scored based on a sequence deep learning model constructed based on the above-mentioned method for constructing a sequence deep learning model for polygenic risk scoring.

[0119] In summary, the present invention innovatively provides a new genetic variation data frame that integrates multiple SNPs of each genetic variation, each genetic variation data frame includes genotype information of multiple SNPs upstream and downstream of each target genetic variation, LD correlation coefficient of each natural population variation, and distance from the target genetic variation. Comprehensively incorporating the upstream and downstream sequence genotypes, distances, and correlation coefficients of the target variation into the deep learning model can effectively enrich the variation information, effectively consider the LD relationship, improve the accuracy of the software, and help the model be applied in a variety of natural population cohorts.

[0120] The present invention uses a sequence neural network to effectively capture the correlation in the sequence data, including the LD between multiple variants. By using a sequence neural network, the genome datasets of various natural population cohorts can be better utilized, and a more accurate polygenic risk scoring model can be constructed.

[0121] The present invention uses LASSO regression to screen genetic variations, which not only reduces the impact of strict thresholds, but also incorporates as many variations as possible with effects, thereby ensuring the reliability of subsequent sequence models.

[0122] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0123] References:

[0124] [1] Chang CC, Chow CC, Tellier LC, Vattikuti S, Purcell SM, Lee JJ. Second-generation PLINK: rising to the challenge of larger and richer datasets. Gigascience. 2015 Feb 25; 4:7.

Claims

1. A method for constructing a sequence deep learning model for polygenic risk scoring of natural populations, characterized in that: It includes the following steps: S1: LASSO regression is used to screen genetic variation data for preprocessed genotype data; S2: Establishing a genetic variation window: Calculate the linkage disequilibrium (LD) correlation coefficient of each natural population variation based on the genetic variation data after LASSO regression in step S1; Establish a genetic variation data frame of X SNPs for each genetic variation; Each of the genetic variation data frames includes the genotype information of the X SNPs of each target genetic variation, the distance to the target genetic variation and the LD correlation coefficient; The obtained n*X*Y data set is the genetic variation window, where n is the number of variations; The number of SNPs in each genetic variation data frame is X≤51; Y≥3; X SNPs refer to the target variation, 0.5*(X-1) genetic variations upstream of the target variation, and 0.5*(X-1) genetic variations downstream of the target variation; and the X SNPs are located within 250kb upstream and downstream of the target variation, and have a correlation coefficient with the target variation r2>0.1; S3: Construction of sequence deep learning model for polygenic risk score: The genetic variation window obtained in step S2 is dimensionless, and then the dimensionless data is input into a sequence neural network for model training until the model reaches a preset convergence condition. The model is used for polygenic risk scoring.

2. The method for constructing a sequence deep learning model for polygenic risk scoring of natural populations according to claim 1, characterized in that: The number of SNPs in each of the genetic variation data frames: 11≤X≤51; Preferably, the number of SNPs in each of the genetic variation data frames is: 21≤X≤51.

3. The method for constructing a sequence deep learning model for polygenic risk scoring of natural populations according to claim 1, characterized in that: If the number of genetic variants upstream of the target variant is less than 0.5*(X-1), it is filled with an identifier that is different from the variant genotype; if the number of genetic variants downstream of the target variant is less than 0.5*(X-1), it is filled with an identifier that is different from the variant genotype.

4. The method for constructing a sequence deep learning model for polygenic risk scoring of natural populations according to claim 1, characterized in that: Numerically and / or alphabetically label the genotype of each of the genetic variation data frames; Preferably, the genotype of each of the genetic variation data frames includes a major allele type, a heterozygote, a minor allele type, and an identifier that is distinguished from the variant genotype; Preferably, when the SNP is located upstream of the target genetic variation, the distance to the target genetic variation is set to a negative value; when the SNP is located downstream of the target genetic variation, the distance to the target genetic variation is set to a positive value; when the SNP is the target genetic variation, the distance to the target genetic variation is set to 0.

5. The method for constructing a sequence deep learning model for polygenic risk scoring of natural populations according to claim 1, characterized in that: When Y≥4, each of the genetic variation data frames further includes other variation features of the X SNPs of each target genetic variation, and the other variation features are selected from at least one of variation type and variation length features.

6. The method for constructing a sequence deep learning model for polygenic risk scoring of natural populations according to claim 1, characterized in that: Using the LASSO regression to screen genetic variation data includes: using LASSO regression to minimize the following loss function: Where N is the number of samples; p is the number of genetic variants; y i is the target value of the ith observation; x ij is the value of the jth genetic variant of the ith observation; β i is the coefficient of the jth feature variable; β0 is the intercept term; λ is the regularization parameter; Preferably, the linkage disequilibrium (LD) correlation coefficient of each natural population variation is calculated using PLINK 1.90beta software; Preferably, the sequence neural network is selected from a recurrent neural network, a Transform, a recursive neural network, a gated recursive unit, and a long short-term memory unit; Preferably, the polygenic risk score for natural population refers to: scoring the risk of at least one polygenic disease among cardiovascular and cerebrovascular diseases, diabetes, malignant tumors, chronic respiratory diseases, arthritis and obesity in natural population; Preferably, the natural population polygenic risk score refers to: scoring the risk of chronic respiratory diseases in the natural population.

7. A method for obtaining polygenic risk scores of natural populations based on a deep learning model, characterized in that: include: The sequence deep learning model constructed based on the method for constructing a sequence deep learning model for polygenic risk scoring according to any one of claims 1 to 6 scores the SNP data to be scored.

8. A device for polygenic risk scoring of natural populations, characterized in that: include: Input module, control module and output module; The input module is configured to: input SNP sample data and / or sample genotype data; The control module includes: a genotype data preprocessing module, a genetic variation data screening module, a genetic variation window construction module, a model construction module and a data scoring module; The genotype data preprocessing module is configured to: perform file conversion on the SNP sample data and / or the sample genotype data; The genetic variation data screening module is configured to: screen genetic variation data using LASSO regression; The genetic variation window construction module is configured to: calculate the linkage disequilibrium (LD) correlation coefficient of each population variation for the genetic variation data after LASSO regression; establish a genetic variation data frame of X SNPs for each genetic variation; wherein each of the genetic variation data frames includes the genotype information of the X SNPs of each target genetic variation, the distance to the target genetic variation and the LD correlation coefficient; obtain a data set of n*X*Y, which is the genetic variation window, wherein n is the number of variations; the number of SNPs in each of the genetic variation data frames is X≤51; Y≥3; the X SNPs refer to the target variation, 0.5*(X-1) genetic variations upstream of the target variation and 0.5*(X-1) genetic variations downstream of the target variation; and the X SNPs are located within 250kb upstream and downstream of the target variation, and the correlation coefficient with the target variation is r2>0.1; The model building module is configured to: non-dimensionalize the obtained genetic variation window, and then input the non-dimensionalized data into a sequence neural network for model training until the model reaches a preset convergence condition; The data scoring module is configured to: score the SNP data to be scored based on the sequence deep learning model constructed by the model construction module; The output module is configured to output a polygenic risk score result.

9. An electronic device, characterized in that: The method comprises a memory and a processor, and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the steps in the method for constructing a sequence deep learning model for polygenic risk scoring according to any one of claims 1 to 6 are completed.

10. A computer-readable storage medium, characterized in that: Used to store computer instructions, which, when executed by a processor, complete the steps in the method for constructing a sequence deep learning model for polygenic risk scoring according to any one of claims 1 to 6.

Images