Prediction method and device for predicting cell carcinogenicity based on genotype data and storage medium

The Bayesian model addresses the inefficiencies of existing cell carcinogenicity prediction algorithms by utilizing linkage disequilibrium data to select relevant SNP sites, resulting in accurate and efficient predictions.

CN119091956APending Publication Date: 2024-12-06ALLIFE REGENERATIVE MEDICINE TECH BEIJING CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411271721.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-11
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

Current algorithms for predicting cell carcinogenicity based on polygenic scores suffer from high misjudgment rates and calculation times, and fail to consider linkage disequilibrium effects.

Method used

A Bayesian model is used to predict cell carcinogenicity by considering all SNP sites, selecting relevant sites for prediction, and employing linkage disequilibrium data to construct a prediction model with high accuracy and efficiency.

Benefits of technology

The Bayesian model achieves highly accurate and efficient prediction of cell carcinogenicity by identifying specific SNP sites, reducing misjudgment rates and improving analysis speed.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119091956A_ABST
    Figure CN119091956A_ABST
Patent Text Reader

Abstract

The invention discloses a prediction method and device for predicting cell carcinogenicity based on genotype data and a storage medium, and provides a brand-new prediction method for predicting cell carcinogenicity based on genotype data, which is high in accuracy and analysis efficiency, for the field of human cell carcinogenicity prediction. The specific SNP sites related to the cell carcinogenicity are found in all the SNP sites by using the Bayesian model, and cell carcinogenicity prediction is realized by detecting the genotypes of the specific SNP sites of the to-be-detected cells, so that the method has important significance on human cell carcinogenicity prediction and has a wide application prospect.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of bioinformatics, and in particular relates to a prediction method, device and storage medium for predicting cell carcinogenicity based on genotype data. Background Art

[0002] With the popularization of second-generation sequencing technology, polygenic scores have received widespread attention. That is, the risk effects of multiple variant sites of an individual organism are combined into a single score, which establishes a connection between disease and trait. iPS cells have developed rapidly in the field of regenerative medicine and can differentiate into a variety of human cell subtypes. However, their harmfulness is also of great concern, especially their carcinogenicity. An international polygenic scoring database, PGS Catalog, has been established, which covers almost many types of diseases including cancer. As for the current algorithm development through polygenic scoring, a variety of models such as P+T, PRS-CS, and PRSmix have been applied to the field of predicting pathogenicity. However, after these methods obtain genotype data to establish models, the misjudgment rate and calculation time are high, and most importantly, the linkage disequilibrium effect is not considered. Summary of the invention

[0003] In view of this, the purpose of the present invention is to provide a prediction method, device and storage medium for predicting cell carcinogenicity based on genotype data, so as to overcome the above-mentioned technical problems existing in the prior art.

[0004] The Bayesian model is a classic probability model that collects data and combines prior knowledge to update the posterior probability and make predictions. The core of the Bayesian model design method is to use the Bayesian formula to calculate the posterior probability of an object through the prior probability of the object, that is, the probability that the object belongs to a certain class. The class with the largest posterior probability is selected as the class to which the object belongs. The Bayesian model can effectively solve the classification problem.

[0005] The present invention uses a Bayesian model to consider all SNP sites as a whole, and rationally selects those SNP sites related to the prediction of cell carcinogenicity, which has good interpretability and also has extremely high analysis speed and accuracy. The present invention adopts the following technical solutions to achieve the above invention objectives:

[0006] In a first aspect, the present invention provides a method for predicting cell carcinogenicity based on genotype data, the method comprising:

[0007] Obtaining genotype data of the SNP site of the cells to be tested;

[0008] Inputting the genotype data of the SNP site into a constructed prediction model, wherein the prediction model predicts the carcinogenicity of the cells to be tested based on the genotype data of the SNP site;

[0009] Predicting and outputting the carcinogenicity prediction result of the cell to be tested based on the output result of the prediction model;

[0010] The SNP sites include: rs13303065, rs13303010, rs112905931, rs13303368, rs34712273, rs9442392, rs2298214, rs3121567, rs13303172, rs4615788, rs4970394, rs2710872, rs9442393, rs4072537, rs9442380, rs7538773, rs12135382, rs78021605, rs75972122, rs3766186, rs12093154, rs78446752, rs115270700, rs1886773, rs307348, rs182532, rs71628956, rs36066545, rs17160669, rs12408050, rs2765035, rs2765033, rs1240709, rs3766176, rs7366884, rs3765908, rs61776804, rs2294487, rs1014988, rs494264, rs1062090, rs2072927, rs4751, rs1107910, rs17162854, rs4648629, rs12044597, rs9660180, rs12034740, rs7546498, rs4648727, rs9660106, rs4648729, rs6603803, rs2803298, rs2748983, rs2803316, rs2803328, rs2459985, rs2678949, rs3107137, rs10910029, rs3128339, rs3107148, rs2503712, rs2503715, rs2480714, rs2460000, rs2480712, rs1692580, rs377599, rs262680, rs263527, rs260508, rs903908, rs4648819, rs10910045, rs12045693, rs12731820, rs12743493, rs2279702, rs2843152, rs61762079, rs2643893, rs10910060, rs2843128, rs1123571, rs903911, rs6698680, rs6685829, rs6693447, rs3795269, rs6671730, rs2494638, rs4648356, rs2377041, rs897628rs7537581, rs76244841, rs6656067. ,

[0011] Furthermore, the prediction model is a Bayesian model, and the method for constructing the prediction model includes:

[0012] Obtain linkage disequilibrium data, genotype data and GWAS summary statistics for human population cells;

[0013] The human population cells are divided into a training set and a test set, based on the linkage disequilibrium data, genotype data and GWAS summary statistics corresponding to the training set, and based on N groups of different hyperparameter values ​​h 2 and p, using the function snp_ldpred2_auto to fit to obtain N different Bayesian models, and at the same time obtaining the SNP loci as described above, selecting the Bayesian model that meets the predetermined requirements as the prediction model, and verifying it through the test set to evaluate the effectiveness of the prediction model, and obtaining a constructed prediction model;

[0014] Wherein, N is a natural number between 1000 and 1500;

[0015] Optionally, the cells to be tested are cells to be tested of human origin.

[0016] Furthermore, the Bayesian model is constructed based on the LDpred2 software package;

[0017] Optionally, the selecting the Bayesian model that meets the predetermined requirements includes: among the N different Bayesian models constructed, selecting a Bayesian model that meets the following requirements as the Bayesian model that meets the predetermined requirements: a Bayesian model with a minimum misjudgment rate.

[0018] Furthermore, the parameter value p of the Bayesian model is between 0.0002211056 and 0.0003621337, and h 2 Between 0.2063467 and 0.2116096.

[0019] Further, the SNP site is a single base mutation site;

[0020] Optionally, the linkage disequilibrium data, genotype data and GWAS summary statistics of the human population cells are obtained by the following method:

[0021] Linkage disequilibrium data were obtained through the database hapmap-3 (https: / / www.broadinstitute.org / medical-and-population-genetics / hapmap-3), and genotype data and GWAS summary statistics were obtained through the database GWAS Catalog (https: / / www.ebi.ac.uk / gwas / );

[0022] Optionally, the preprocessing includes the following process:

[0023] The downloaded GWAS summary statistics data are quality controlled to obtain quality-controlled GWAS summary statistics data; the quality-controlled GWAS summary statistics data and linkage disequilibrium data are used to calculate the correlation matrix through snp_cor; the downloaded linkage disequilibrium data are used to calculate the heritability h through a regression model 2 .

[0024] In a second aspect, the present invention provides a prediction system for predicting cell carcinogenicity based on genotype data, the system comprising:

[0025] A data acquisition unit is used to acquire genotype data of the SNP site of the cell to be tested;

[0026] An analysis and prediction unit, used for inputting the genotype data of the SNP site into a constructed prediction model, wherein the prediction model predicts the carcinogenicity of the cells to be tested based on the genotype data of the SNP site;

[0027] An output result unit, used for predicting and outputting the carcinogenicity prediction result of the cell to be tested based on the output result of the prediction model;

[0028] The SNP site is the SNP site described in the first aspect of the present invention;

[0029] The prediction result is obtained based on the prediction model, and the method for constructing the prediction model includes:

[0030] Obtain linkage disequilibrium data, genotype data and GWAS summary statistics for human population cells;

[0031] The human population cells are divided into a training set and a test set, based on the linkage disequilibrium data, genotype data and GWAS summary statistics corresponding to the training set, and based on N groups of different hyperparameter values ​​h 2and p, using the function snp_ldpred2_auto to fit to obtain N different Bayesian models, and at the same time obtain the SNP loci described in the first aspect of the present invention, selecting the Bayesian model that meets the predetermined requirements as the prediction model, and verifying it through the test set to evaluate the effectiveness of the prediction model, and obtaining a constructed prediction model;

[0032] Wherein, N is a natural number between 1000 and 1500.

[0033] Furthermore, the Bayesian model is constructed based on the LDpred2 software package;

[0034] Optionally, the selecting the Bayesian model that meets the predetermined requirements includes: selecting a Bayesian model that meets the following requirements from among the N different Bayesian models constructed as the Bayesian model that meets the predetermined requirements: a Bayesian model with the smallest misjudgment rate;

[0035] Optionally, the parameter value p of the Bayesian model is between 0.0002211056 and 0.0003621337, and h 2 Between 0.2063467 and 0.2116096.

[0036] In a third aspect, the present invention provides an electronic device comprising a memory and a processor; the memory is used to store program instructions; the processor is used to call program instructions, and when the program instructions are executed, they are used to execute the steps of the prediction method for predicting cell carcinogenicity based on genotype data as described in the first aspect of the present invention.

[0037] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for predicting cell carcinogenicity based on genotype data as described in the first aspect of the present invention.

[0038] In a fifth aspect, the present invention provides the prediction method for predicting cell carcinogenicity based on genotype data described in the first aspect of the present invention, the prediction system for predicting cell carcinogenicity based on genotype data described in the second aspect of the present invention, the electronic device described in the third aspect of the present invention, or the computer-readable storage medium described in the fourth aspect of the present invention, in predicting the carcinogenicity of human cells.

[0039] The present invention has the following beneficial effects:

[0040] The present invention provides a new, highly accurate and efficient method for predicting cell carcinogenicity based on genotype data in the field of human cell carcinogenicity prediction. The present invention uses a Bayesian model to find specific SNP sites related to cell carcinogenicity among all SNP sites, and realizes cell carcinogenicity prediction by detecting the genotype of the specific SNP site of the test cell. This is of great significance for the prediction of human cell carcinogenicity and has broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0042] Figure 1 A schematic flow chart of a prediction method for predicting cell carcinogenicity based on genotype data provided in an embodiment of the present invention;

[0043] Figure 2 A schematic diagram of a prediction system for predicting cell carcinogenicity based on genotype data provided by an embodiment of the present invention;

[0044] Figure 3 A schematic diagram of an electronic device for predicting cell carcinogenicity based on genotype data provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0045] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0046] In some of the processes described in the specification and claims of the present invention and the above-mentioned drawings, multiple operations appearing in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this document or may be executed in parallel. The serial numbers of the operations, such as S101, S102, etc., are only used to distinguish different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel.

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0048] Figure 1 1 is a flow chart of a prediction method for predicting cell carcinogenicity based on genotype data provided by an embodiment of the present invention. Specifically, the method comprises the following steps:

[0049] S101: Obtaining genotype data of the SNP site of the cell to be tested;

[0050] The SNP sites include: rs13303065, rs13303010, rs112905931, rs13303368, rs34712273, rs9442392, rs2298214, rs3121567, rs13303172, rs4615788, rs4970394, rs2710872, rs9442393, rs4072537, rs9442380, rs7538773, rs12135382, rs78021605, rs75972122, rs3766186, rs12093154, rs78446752, rs115270700, rs1886773, rs307348, rs182532, rs71628956, rs36066545, rs17160669, rs12408050, rs2765035, rs2765033, rs1240709, rs3766176, rs7366884, rs3765908, rs61776804, rs2294487, rs1014988, rs494264, rs1062090, rs2072927, rs4751, rs1107910, rs17162854, rs4648629, rs12044597, rs9660180, rs12034740, rs7546498, rs4648727, rs9660106, rs4648729, rs6603803, rs2803298, rs2748983, rs2803316, rs2803328, rs2459985, rs2678949, rs3107137, rs10910029, rs3128339, rs3107148, rs2503712, rs2503715, rs2480714, rs2460000, rs2480712, rs1692580, rs377599, rs262680, rs263527, rs260508, rs903908, rs4648819, rs10910045, rs12045693, rs12731820, rs12743493, rs2279702, rs2843152, rs61762079, rs2643893, rs10910060, rs2843128, rs1123571, rs903911, rs6698680, rs6685829, rs6693447, rs3795269, rs6671730, rs2494638, rs4648356, rs2377041, rs897628rs7537581, rs76244841, rs6656067. ,

[0051] In one embodiment, the cells to be tested are cells to be tested that are derived from mammals, including humans and non-human mammals, and the non-human mammals include but are not limited to: experimental animals or other animals, such as rabbits, rodents (such as mice, rats, hamsters, gerbils and guinea pigs), cattle, sheep, pigs, goats, horses, dogs, cats, primates (such as chimpanzees, rhesus monkeys, and cynomolgus monkeys). In a preferred embodiment, the cells to be tested are cells to be tested that are derived from humans.

[0052] In one embodiment, the single nucleotide polymorphism (SNP) refers to a DNA sequence polymorphism caused by a variation of a single nucleotide at the genomic level. It is the most common type of heritable variation in humans, accounting for more than 90% of all known polymorphisms. SNP can be a variation within a gene sequence or a variation on a non-coding sequence outside a gene. This variation can be caused by a conversion or transversion of a single base, or by an insertion or deletion of a base.

[0053] In one embodiment, the genotype data of the SNP site can be obtained by detecting the cells to be tested by a variety of detection methods. Exemplarily, the detection methods include but are not limited to: Sanger sequencing, TaqMan probe method, SNaPshot method, ARMS-PCR, molecular beacon, high-resolution melting curve method, CAPS, KASP, gene chip, mass spectrometry.

[0054] In one embodiment, the Sanger sequencing is a method based on the dideoxynucleotide termination reaction, which is considered the "gold standard" for SNP detection by generating fragments of different lengths for sequencing. It can determine the type and location of mutations and discover unknown SNP sites, but the cost is relatively high and it is suitable for situations with few sites and few samples.

[0055] In one embodiment, the TaqMan probe method utilizes real-time fluorescent PCR technology to amplify by designing specific probes and primers. The 5' end and 3' end of the probe are respectively labeled with a reporter fluorescent group and a quencher fluorescent group. When the PCR product exists in the solution, the probe anneals with the template to generate a fluorescent signal, thereby realizing the detection of the SNP site.

[0056] In one embodiment, the SNaPshot method is a typing technology based on the principle of fluorescent labeled single base extension, also known as small sequencing, which is mainly aimed at medium-throughput SNP typing projects. In a reaction system containing sequencing enzyme, four fluorescent labeled ddNTPs, extension primers of different lengths and PCR product templates, the primer is extended by one base and then terminated. After detection, the SNP site is determined according to the position and color of the peak.

[0057] In one embodiment, the ARMS-PCR method is an allele-specific PCR method, which is based on the principle that Taq DNA polymerase cannot repair a single base mismatch at the 3' end of the primer, thereby hindering amplification. By designing specific primers and probes, the wild-type and mutant alleles can be detected.

[0058] In one embodiment, the molecular beacon method uses the signal changes generated when the molecular beacon probe binds to the target DNA for detection. The molecular beacon is a special oligonucleotide probe that can emit fluorescent signals under specific conditions, and the SNP site is determined by detecting the changes in these signals.

[0059] In one embodiment, the high-resolution melting curve method analyzes SNP sites by monitoring the melting temperature of PCR products. Different SNP sites have different melting temperatures, and the type of SNP can be determined by high-resolution melting curve analysis.

[0060] In one embodiment, the CAPS method is a method based on restriction endonucleases, and the SNP site is determined by analyzing the products after enzyme cleavage.

[0061] In one embodiment, the KASP method uses specific primers and probes to perform PCR amplification and determines the SNP site by detecting changes in fluorescent signals.

[0062] In one embodiment, the gene chip method hybridizes the sample DNA with the probes on the chip and determines the SNP site by detecting the hybridization signal.

[0063] In one embodiment, the mass spectrometry method analyzes the mass of DNA molecules by mass spectrometry technology to determine the SNP site. This method has the characteristics of high accuracy and high throughput, but the cost is relatively high.

[0064] S102: inputting the genotype data of the SNP site into a constructed prediction model, wherein the prediction model predicts the carcinogenicity of the cells to be tested based on the genotype data of the SNP site;

[0065] In one embodiment, the prediction model is a Bayesian model, and the method for constructing the prediction model includes:

[0066] Obtain linkage disequilibrium data, genotype data and GWAS summary statistics for human population cells;

[0067] The human population cells are divided into a training set and a test set, based on the linkage disequilibrium data, genotype data and GWAS summary statistics corresponding to the training set, and based on N groups of different hyperparameter values ​​h 2 and p, using the function snp_ldpred2_auto to fit to obtain N different Bayesian models, and at the same time obtaining the SNP loci as described above, selecting the Bayesian model that meets the predetermined requirements as the prediction model, and verifying it through the test set to evaluate the effectiveness of the prediction model, and obtaining a constructed prediction model;

[0068] Wherein, N is a natural number between 1000 and 1500.

[0069] In one embodiment, the Bayesian model performs feature selection based on the genotype data, and the constructed prediction model predicts the carcinogenicity of cells by detecting the genotypes of the multiple SNP sites of the individual cells to be tested.

[0070] In one embodiment, the Bayesian model is constructed based on the LDpred2 software package. The present invention uses LDpred2, which uses genotype data for feature selection, to construct a Bayesian model to obtain a model with higher accuracy, which can achieve efficient and accurate prediction of cell carcinogenicity. LDpred2 is a Bayesian method for polygenic risk score analysis, which is an improved version of the LDpred algorithm. This algorithm is mainly used to predict individual polygenic risk scores based on summary statistics of genome-wide association studies (GWAS), and then estimate individual genetic traits.

[0071] In one embodiment, the selecting of the Bayesian model that meets the predetermined requirements includes: among the N different Bayesian models constructed, selecting a Bayesian model that meets the following requirements as the Bayesian model that meets the predetermined requirements: a Bayesian model with the smallest misjudgment rate.

[0072] In one embodiment, the parameter value p of the Bayesian model is between 0.0002211056 and 0.0003621337, and h 2 Between 0.2063467 and 0.2116096.

[0073] In one embodiment, the Bayesian model for predicting cell carcinogenicity obtained by constructing and screening has excellent performance, and the excellent performance is: the AUC index is above 70.

[0074] In one embodiment, the p obtained by the present invention is 0.0002398453 and h 2 The Bayesian model with a prediction AUC index of 0.2091662 is above 70, which can be used to accurately predict cell carcinogenicity.

[0075] In one embodiment, the SNP site is a single base mutation site; the linkage disequilibrium data, genotype data and GWAS summary statistics of the human population cells are obtained by the following method:

[0076] Linkage disequilibrium data were obtained through the database hapmap-3 (https: / / www.broadinstitute.org / medical-and-population-genetics / hapmap-3), and genotype data and GWAS summary statistics were obtained through the database GWAS Catalog (https: / / www.ebi.ac.uk / gwas / ).

[0077] In one embodiment, by accessing https: / / figshare.com / ndownloader / files / 37802721, the linkage disequilibrium data of the human population is obtained after data downloading, and by accessing https: / / www.ebi.ac.uk / gwas / , a target cancer research project is selected, and after data downloading, gwas summarystatistic data, i.e., the genotype data and GWAS summary statistics of the human population, are obtained.

[0078] In one embodiment, the preprocessing includes the following steps: performing quality control on the downloaded GWAS summary statistics data to obtain quality-controlled GWAS summary statistics data; calculating the correlation matrix between the quality-controlled GWAS summary statistics data and linkage disequilibrium data through snp_cor; calculating the heritability h of the downloaded linkage disequilibrium data through a regression model. 2 .

[0079] S103: predicting and outputting the carcinogenicity prediction result of the cell to be tested based on the output result of the prediction model;

[0080] In one embodiment, the carcinogenicity of the cells to be tested is predicted based on the genotype data of the SNP sites of the cells to be tested and the constructed prediction model.

[0081] Figure 2 Schematic diagram of a prediction system for predicting cell carcinogenicity based on genotype data provided by an embodiment of the present invention. Specifically, the system includes:

[0082] A data acquisition unit is used to acquire genotype data of the SNP site of the cell to be tested;

[0083] An analysis and prediction unit, used for inputting the genotype data of the SNP site into a constructed prediction model, wherein the prediction model predicts the carcinogenicity of the cells to be tested based on the genotype data of the SNP site;

[0084] An output result unit, used for predicting and outputting the carcinogenicity prediction result of the cell to be tested based on the output result of the prediction model;

[0085] The SNP site is the SNP site described above;

[0086] The prediction result is obtained based on the prediction model, and the method for constructing the prediction model includes:

[0087] Obtain linkage disequilibrium data, genotype data and GWAS summary statistics for human population cells;

[0088] The human population cells are divided into a training set and a test set, based on the linkage disequilibrium data, genotype data and GWAS summary statistics corresponding to the training set, and based on N groups of different hyperparameter values ​​h 2 and p, using the function snp_ldpred2_auto to fit to obtain N different Bayesian models, and at the same time obtaining the SNP loci as described above, selecting the Bayesian model that meets the predetermined requirements as the prediction model, and verifying it through the test set to evaluate the effectiveness of the prediction model, and obtaining a constructed prediction model;

[0089] Wherein, N is a natural number between 1000 and 1500.

[0090] In one embodiment, the Bayesian model performs feature selection based on the genotype data, and the constructed prediction model predicts the carcinogenicity of cells by detecting the genotypes of the multiple SNP sites of the individual cells to be tested.

[0091] In one embodiment, the Bayesian model is constructed based on the LDpred2 software package. The present invention uses LDpred2, which uses genotype data for feature selection, to construct a Bayesian model to obtain a model with high accuracy, which can achieve efficient and accurate prediction of cell carcinogenicity.

[0092] In one embodiment, the selecting of the Bayesian model that meets the predetermined requirements includes: among the N different Bayesian models constructed, selecting a Bayesian model that meets the following requirements as the Bayesian model that meets the predetermined requirements: a Bayesian model with the smallest misjudgment rate.

[0093] In one embodiment, the parameter value p of the Bayesian model is between 0.0002211056 and 0.0003621337, and h 2 Between 0.2063467 and 0.2116096.

[0094] In one embodiment, the Bayesian model for predicting cell carcinogenicity obtained by constructing and screening has excellent performance, and the excellent performance is: the AUC index is above 70.

[0095] In one embodiment, the p obtained by the present invention is 0.0002398453 and h 2 The Bayesian model with a prediction AUC index of 0.2091662 is above 70, which can be used to accurately predict cell carcinogenicity.

[0096] Figure 3 A schematic diagram of an electronic device for predicting cell carcinogenicity based on genotype data provided by an embodiment of the present invention, specifically, the electronic device includes:

[0097] A memory and a processor; the memory is used to store program instructions; the processor is used to call the program instructions, and when the program instructions are executed, it is used to execute the steps of the prediction method for predicting cell carcinogenicity based on genotype data as described above.

[0098] An embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the prediction method for predicting cell carcinogenicity based on genotype data as described above are implemented.

[0099] In one embodiment, the storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc. The computer-readable storage medium may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function, etc.; the data storage area may store the created related data, etc.

[0100] The verification results of this verification example show that assigning inherent weights to indications can moderately improve the performance of the method relative to the default settings. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here.

[0101] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0102] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0103] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0104] A person of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium may include: a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, etc.

[0105] A person skilled in the art can understand that all or part of the steps in the above-mentioned embodiment method can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. The above-mentioned storage medium can be a read-only memory, a disk or an optical disk, etc.

[0106] The above is a detailed introduction to a computer device provided by the present invention. For a person skilled in the art, according to the concept of the embodiments of the present invention, there may be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

[0107] Example 1

[0108] This embodiment provides a method for predicting the carcinogenicity of cells and a method for constructing a related prediction model, which specifically includes the following steps:

[0109] 1. Data download

[0110] Genotype data and GWAS summary statistics were obtained by accessing specific cancer research projects at https: / / www.ebi.ac.uk / gwas / and downloading the corresponding GWAS summary statistics data. Linkage disequilibrium data of the human population were obtained by accessing https: / / figshare.com / ndownloader / files / 37802721 and downloading the data.

[0111] 2. Modeling

[0112] 2.1 Preprocess the downloaded GWAS summary statistic data

[0113] Low-quality SNP sites were filtered out, and correction calculations were performed based on the statistical values ​​of the experimental and control groups of the samples.

[0114] 2.1.1 Calculate the correlation matrix

[0115] The snp_cor function is used to calculate the downloaded GWAS summary statistic and linkage disequilibrium data.

[0116] 2.2 Bayesian fitting modeling

[0117] First, randomly select 90% of the individuals as the training set; randomly select about 10% of the data as the test set. The test set does not participate in the modeling process and is only used to predict the effect of the evaluation model. Call the snp_ldpred2_auto function in the LDpred2 package to model, and tune the model through 9-fold cross validation to find the hyperparameter values ​​p and h of the optimal performance model 2 .

[0118] 2.2.1 Model establishment

[0119] (1) The snp_ldpred2_auto function uses the train data to fit the model multiple times without the need for test data verification. The parameter seq_log = (0.0001, 0.2) allows the algorithm to automatically select 1500 different p and h 2 Value, fitting 1500 models with different coefficients;

[0120] These include:

[0121] i) Mean AUC: The average AUC is an important indicator because it provides a summary of the overall performance of the model on different datasets. If a model has a high AUC value on multiple different datasets, then it can be considered that the model has good generalization ability. The mean AUC is 70.2, and the number of nonzero SNPs (non-zero sites) is 1,038,338. Compared with the total number of human SNP sites, the total number is reduced to about 1 / 3.

[0122] 3. Relevant formulas and mathematical concepts used for calculations

[0123] Prior distribution: LDpred2 uses a prior distribution to express the initial belief about the effect size of the SNPs. This usually assumes that most SNPs have an effect size close to zero.

[0124] Posterior distribution: The posterior distribution combines prior knowledge and GWAS data to update the estimate of the SNP effect size. The posterior distribution is the conditional probability of the parameter given the data.

[0125] Likelihood function: The likelihood function measures the probability of observing GWAS data under certain parameters. In LDpred2, the likelihood function is usually based on a multivariate normal distribution.

[0126] Bayesian Update: Use Bayes’ theorem to combine the prior distribution and the likelihood function to calculate the posterior distribution:

[0127] P(θ|D)∝P(D|θ)·P(θ)

[0128] Among them, P(θ|D) is the posterior distribution, P(D|θ) is the likelihood function, and P(θ) is the prior distribution.

[0129] Polygenic score (PRS): The polygenic score is calculated by weighting the effect size and genotype of each SNP:

[0130]

[0131] where gi is the genotype of the ith SNP (usually coded as 0, 1, or 2) and βi is the corresponding effect size.

[0132] Linkage disequilibrium (LD): LDpred2 takes into account linkage disequilibrium between SNPs, which is represented by an LD matrix that contains the correlations between pairs of SNPs.

[0133] Gibbs Sampler: LDpred2 generates polygenic risk scores using the Gibbs sampler, a random sampling method for generating samples from complex multivariate posterior distributions.

[0134] Calibrated probabilities: LDpred2 provides a calibration of the causal probability for each SNP, which helps identify variants that are likely to have a causal role.

[0135] 4. Feature selection SNP loci list of the optimal model

[0136] rs13303065、rs13303010、rs112905931、rs13303368、rs34712273、rs9442392、rs2298214、rs3121567、rs13303172、rs4615788、rs4970394、rs2710872、rs9442393、rs4072537、rs9442380、rs7538773、rs12135382、rs78021605、rs75972122、rs3766186、rs12093154、rs78446752、rs115270700、rs1886773、rs307348、rs182532、rs71628956、rs36066545、rs17160669、rs12408050、rs2765035、rs2765033、rs1240709、rs3766176、rs7366884、rs3765908、rs61776804、rs2294487、rs1014988、rs494264、rs1062090、rs2072927、rs4751、rs1107910、rs17162854、rs4648629、rs12044597、rs9660180、rs12034740、rs7546498、rs4648727、rs9660106、rs4648729、rs6603803、rs2803298、rs2748983、rs2803316、rs2803328、rs2459985、rs2678949、rs3107137、rs10910029、rs3128339、rs3107148、rs2503712、rs2503715、rs2480714、rs2460000、rs2480712、rs1692580、rs377599、rs262680、rs263527、rs260508、rs903908、rs4648819、rs10910045、rs12045693、rs12731820、rs12743493、rs2279702、rs2843152、rs61762079、rs2643893、rs10910060、rs2843128、rs1123571、rs903911、rs6698680、rs6685829、rs6693447、rs3795269、rs6671730、rs2494638、rs4648356、rs2377041、rs897628、rs7537581、rs76244841、rs6656067。、

Claims

1. A prediction method for predicting cell carcinogenicity based on genotype data, It is characterized in that The method comprises: Obtaining genotype data of the SNP site of the cells to be tested; Inputting the genotype data of the SNP site into a constructed prediction model, wherein the prediction model predicts the carcinogenicity of the cells to be tested based on the genotype data of the SNP site; Predicting and outputting the carcinogenicity prediction result of the cell to be tested based on the output result of the prediction model; The SNP sites include: rs13303065, rs13303010, rs112905931, rs13303368, rs34712273, rs9442392, rs2298214, rs3121567, rs13303172, rs4615788, rs4970394, rs2710872, rs9442393, rs4072537, rs9442380, rs7538773, rs12135382, rs78021605, rs75972122, rs3766186, rs12093154, rs78446752, rs115270700, rs1886773, rs307348, rs182532, rs71628956, rs36066545, rs17160669, rs12408050, rs2765035, rs2765033, rs1240709, rs3766176, rs7366884, rs3765908, rs61776804, rs2294487, rs1014988, rs494264, rs1062090, rs2072927, rs4751, rs1107910, rs17162854, rs4648629, rs12044597, rs9660180, rs12034740, rs7546498, rs4648727, rs9660106, rs4648729, rs6603803, rs2803298, rs2748983, rs2803316, rs2803328, rs2459985, rs2678949, rs3107137, rs10910029, rs3128339, rs3107148, rs2503712, rs2503715, rs2480714, rs2460000, rs2480712, rs1692580, rs377599, rs262680, rs263527, rs260508, rs903908, rs4648819, rs10910045, rs12045693, rs12731820, rs12743493, rs2279702, rs2843152, rs61762079, rs2643893, rs10910060, rs2843128, rs1123571, rs903911, rs6698680, rs6685829, rs6693447, rs3795269, rs6671730, rs2494638, rs4648356, rs2377041, rs897628rs7537581、rs76244841、rs6656067。、 2. The method according to claim 1, It is characterized in that The prediction model is a Bayesian model, and the method for constructing the prediction model includes: Obtain linkage disequilibrium data, genotype data and GWAS summary statistics for human population cells; The human population cells are divided into a training set and a test set, based on the linkage disequilibrium data, genotype data and GWAS summary statistics corresponding to the training set, and based on N groups of different hyperparameter values ​​h 2 and p, using the function snp_ldpred2_auto to fit to obtain N different Bayesian models, and at the same time obtain the SNP site described in claim 1, selecting the Bayesian model that meets the predetermined requirements as the prediction model, and verifying it through the test set to evaluate the effectiveness of the prediction model, and obtaining a constructed prediction model; Wherein, N is a natural number between 1000 and 1500; Optionally, the cells to be tested are cells to be tested of human origin.

3. The method according to claim 2, It is characterized in that The Bayesian model is constructed based on the LDpred2 software package; Optionally, the selecting the Bayesian model that meets the predetermined requirements includes: among the N different Bayesian models constructed, selecting a Bayesian model that meets the following requirements as the Bayesian model that meets the predetermined requirements: a Bayesian model with a minimum misjudgment rate.

4. The method according to claim 3, It is characterized in that The parameter value p of the Bayesian model is between 0.0002211056 and 0.0003621337, and h 2 Between 0.2063467 and 0.2116096.

5. The method according to claim 2, It is characterized in that The SNP site is a single base mutation site; Optionally, the linkage disequilibrium data, genotype data and GWAS summary statistics of the human population cells are obtained by the following method: Linkage disequilibrium data were obtained through the database hapmap-3 (https: / / www.broadinstitute.org / medical-and-population-genetics / hapmap-3), and genotype data and GWAS summary statistics were obtained through the database GWAS Catalog (https: / / www.ebi.ac.uk / gwas / ); Optionally, the preprocessing includes the following process: The downloaded GWAS summary statistics data are quality controlled to obtain quality-controlled GWAS summary statistics data; the quality-controlled GWAS summary statistics data and linkage disequilibrium data are used to calculate the correlation matrix through snp_cor; the downloaded linkage disequilibrium data are used to calculate the heritability h through a regression model 2 .

6. A prediction system for predicting cell carcinogenicity based on genotype data, It is characterized in that The system comprises: A data acquisition unit is used to acquire genotype data of the SNP site of the cell to be tested; An analysis and prediction unit, used for inputting the genotype data of the SNP site into a constructed prediction model, wherein the prediction model predicts the carcinogenicity of the cells to be tested based on the genotype data of the SNP site; An output result unit, used for predicting and outputting the carcinogenicity prediction result of the cell to be tested based on the output result of the prediction model; The SNP site is the SNP site described in claim 1; The prediction result is obtained based on the prediction model, and the method for constructing the prediction model includes: Obtain linkage disequilibrium data, genotype data and GWAS summary statistics for human population cells; The human population cells are divided into a training set and a test set, based on the linkage disequilibrium data, genotype data and GWAS summary statistics corresponding to the training set, and based on N groups of different hyperparameter values ​​h 2 and p, using the function snp_ldpred2_auto to fit to obtain N different Bayesian models, and at the same time obtain the SNP site described in claim 1, selecting the Bayesian model that meets the predetermined requirements as the prediction model, and verifying it through the test set to evaluate the effectiveness of the prediction model, and obtaining a constructed prediction model; Wherein, N is a natural number between 1000 and 1500.

7. The prediction system for predicting cell carcinogenicity based on genotype data according to claim 6, It is characterized in that The Bayesian model is constructed based on the LDpred2 software package; Optionally, the selecting the Bayesian model that meets the predetermined requirements includes: selecting a Bayesian model that meets the following requirements from among the N different Bayesian models constructed as the Bayesian model that meets the predetermined requirements: a Bayesian model with the smallest misjudgment rate; Optionally, the parameter value p of the Bayesian model is between 0.0002211056 and 0.0003621337, and h 2 Between 0.2063467 and 0.2116096.

8. An electronic device, It is characterized in that The electronic device includes a memory and a processor; the memory is used to store program instructions; the processor is used to call the program instructions, and when the program instructions are executed, it is used to execute the steps of the prediction method for predicting cell carcinogenicity based on genotype data as described in any one of claims 1-5.

9. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for predicting cell carcinogenicity based on genotype data described in any one of claims 1 to 5 are implemented.

10. Use of the prediction method for predicting cell carcinogenicity based on genotype data according to any one of claims 1 to 5, the prediction system for predicting cell carcinogenicity based on genotype data according to claim 6 or 7, the electronic device according to claim 8 or the computer-readable storage medium according to claim 9 in predicting human cell carcinogenicity.