Fetal concentration calculation method and electronic device

Abstract

The invention provides a fetal concentration calculation method and an electronic device. The calculation method comprises the following steps: S1) sequencing a to-be-detected sample to obtain a set of genotypes, covered by SNP loci in a database, of gene loci in the to-be-detected sample, and then carrying out SNP interpolation on the set to obtain a first genotype set; s2) obtaining the genotype of each gene in the first genotype set on the SNP site covered by each window in the SNP site window set, and calculating the proportion of the heterozygous site to the total site in the genotype corresponding to each window in the first genotype set; gathering the proportions of the heterozygous sites corresponding to all the windows in the total sites as a first vector; and S3) inputting the first vector into the fetal concentration prediction model to obtain the fetal concentration of the to-be-detected sample. The method can solve the problem of limitation of fetal concentration detection in the prior art, and is suitable for the field of non-invasive prenatal detection of fetuses.

Description

Technical Field

[0001] This invention relates to the field of non-invasive prenatal testing of fetuses, and more specifically, to a method for calculating fetal concentration and an electronic device. Background Technology

[0002] Non-invasive prenatal testing (NIPT), a revolutionary technology, improves the efficiency and safety of prenatal disease diagnosis by analyzing fetal cell-free DNA in maternal blood. Compared to traditional fetal trisomy screening methods, NIPT significantly reduces false positive rates, avoids the risks associated with invasive procedures, and provides more accurate results. However, a core challenge in the practical application of NIPT is accurately calculating the proportion of fetal DNA in maternal plasma, i.e., fetal concentration. The accuracy of fetal concentration directly affects the reliability of test results and the diagnostic rate of diseases.

[0003] Several algorithms have been proposed and put into practice for calculating fetal concentration, but each has its own obvious limitations, mainly in the following aspects:

[0004] Y chromosome-dependent methods: These algorithms are based on the specificity of the Y chromosome and can provide relatively accurate fetal concentration calculations, but their applicability is limited to the detection of male fetuses, thus limiting their universality and application value in all pregnancy cases.

[0005] Sex chromosome-independent methods: These algorithms further explore computational strategies independent of sex chromosomes, including schemes based on plasma DNA fragment length, methylation status, and fetal-specific SNP sites. While these methods aim to broaden applicability, in practice, they all require the collection of additional experimental data, such as paired-end sequencing (PE) data or dedicated methylation sequencing data. This not only increases costs but also prolongs the testing cycle, reducing the accessibility and practicality of NIPT in a broad population.

[0006] Low-depth whole-genome sequencing methods: Given the shortcomings of the methods mentioned above, there are three existing methods for calculating low-depth single-end sequencing (SE) data from whole-genome sequencing that only require routine NIPT. Among them, seqFF utilizes the genome-wide distribution characteristics of cell-free DNA fragments for calculation, but it struggles to provide accurate results when fetal concentrations are low; another method based on nucleosome fragment distribution simplifies data requirements, but its accuracy is somewhat lacking; hetFF attempts to estimate fetal concentration through heterozygote distribution, but its accuracy drops significantly when the data depth is below 0.5×, posing an obstacle for situations with limited resources or small sample sizes.

[0007] In summary, existing methods for calculating fetal concentration have drawbacks such as being limited by gender, low detection accuracy, or high detection costs. These problems not only affect the widespread adoption of NIPT but also limit its application potential in cases of low fetal concentration and insufficient data depth. Summary of the Invention

[0008] The main objective of this invention is to provide a method and electronic device for calculating fetal concentration, in order to solve the limitations of existing fetal concentration detection technologies.

[0009] To achieve the above objectives, according to a first aspect of the present invention, a method for calculating fetal concentration is provided, the method comprising:

[0010] S1) Sequencing the sample to be tested yields a set of genotypes for each gene locus covered by SNP loci in the database. SNP imputation is then performed on this set to obtain the first genotype set. S2) The genotypes of each gene in the first genotype set are obtained at each SNP locus covered by a window in the SNP locus window set. The proportion of heterozygous loci to the total number of loci in each window of the first genotype set is calculated. The set of proportions of heterozygous loci to the total number of loci across all windows is used as the first vector. The SNP locus window set is defined as follows: the reference genome is divided into multiple uniform, continuous, and non-overlapping regions based on chromosome physical length intervals. Each window covers multiple SNP loci. The windows and the SNP loci covered within them constitute the SNP locus window set. S3) The first vector is input into the fetal concentration prediction model to obtain the fetal concentration of the sample to be tested.

[0011] Further, the method for obtaining the SNP site window set includes: i) dividing the autosomes of the reference genome into a first window set by dividing the chromosomes into windows of 2-5M each, based on the physical length of the chromosomes, where adjacent windows in the first window set do not overlap; filtering out windows in the first window set where the N base ratio is greater than 10% or the number of SNP sites is less than 100, to obtain the second window set; ii) traversing the sequencing data of all samples in the screening set to obtain the genotype of each sample at each SNP site in the database, thus obtaining the genotype set of the screening set SNP sites; statistically analyzing the proportion of heterozygous sites in each window of the second window set for each sample in the screening set SNP site genotype set, thus obtaining the proportion of heterozygous sites for each sample; iii) dividing the genotype set by fetal concentration region... The samples in the screening set are classified into three concentration groups from low to high: a first concentration group, a second concentration group, and a third concentration group. The distribution of the heterozygous locus proportion for each sample in each window of the second window set is statistically analyzed for the samples in the first, second, and third concentration groups. The distribution values ​​corresponding to the first, second, and third concentration groups are compared pairwise. If the p-value between the pairwise comparisons is less than 0.05, the window corresponding to the distribution value of the heterozygous locus proportion is retained, thus obtaining the SNP locus window set. Preferably, the screening set is a sample set used to construct the SNP locus window set. Preferably, the database is the 1000 Genomes Database. Preferably, autosomes do not include chromosomes 13, 18, and 21.

[0012] Furthermore, in the screening set, samples with fetal concentrations of 0.1% to 10% were designated as the first concentration group; samples with fetal concentrations of 10% to 30% were designated as the second concentration group; and samples with fetal concentrations greater than 30% were designated as the third concentration group.

[0013] Furthermore, the method for constructing the fetal concentration prediction model includes: a) sequencing the training set samples to obtain a set of genotypes for each gene locus in the training set samples, and performing SNP imputation on the set to obtain a second genotype set; b) calculating the proportion of heterozygous loci to total loci in the SNP locus window set for each sample in the second genotype set, and obtaining a vector set of the training set samples; c) using the fetal concentration of each sample in the training set samples as the expected result, inputting the vector set of the training set samples into the model for training, and obtaining the fetal concentration prediction model; preferably, the model includes one or more of the following: multivariate logistic regression model, deep neural network, support vector machine, or Bayesian model.

[0014] Furthermore, the samples to be tested included peripheral blood from pregnant women.

[0015] Furthermore, the software for SNP interpolation includes one or more of PLINK software, stitch software, beagle 5.4 software, or BIMBAM software.

[0016] To achieve the above objectives, according to a second aspect of the present invention, an electronic device for calculating fetal concentration is provided, the electronic device comprising a genotype set acquisition unit, a first vector calculation unit, and a fetal concentration calculation unit;

[0017] The system includes the following components: a genotype set acquisition unit, which sequences the sample to be tested, obtains a set of genotypes for each gene locus in the sample that are covered by SNP loci in the database, and then performs SNP imputation on the set to obtain a first genotype set; a first vector calculation unit, which obtains the genotypes of each gene in the first genotype set at the SNP loci covered by each window in the SNP locus window set, calculates the proportion of heterozygous loci to the total loci in the genotypes corresponding to each window in the first genotype set, and uses the set of proportions of heterozygous loci to the total loci for all windows as the first vector; the SNP locus window set is defined as follows: the reference genome is divided into multiple uniform, continuous, and non-overlapping regions according to the physical length interval of chromosomes, and each window covers multiple SNP loci; the windows and the SNP loci covered by the windows constitute the SNP locus window set; and a fetal concentration calculation unit, which inputs the first vector into the fetal concentration prediction model to obtain the fetal concentration of the sample to be tested.

[0018] Furthermore, the first vector calculation unit also includes an SNP locus window set acquisition unit, which includes a window partitioning unit, a window filtering unit, a screening set sample SNP locus genotype set acquisition unit, a heterozygous locus proportion statistics unit, and a distribution value calculation unit. Specifically, the window partitioning unit is used to divide the autosomes of the reference genome according to the chromosome physical length, with a window size of 2-5M, to obtain a first window set, where adjacent windows in the first window set do not overlap. The window filtering unit is used to filter windows in the first window set where the N base ratio is greater than 10% or the number of SNP loci is less than 100, to obtain a second window set. The screening set sample SNP locus genotype set acquisition unit uses... The sequencing data of all samples in the screening set are traversed to obtain the genotype of each SNP locus in the database for each sample, thus obtaining the genotype set of SNP loci in the screening set; the heterozygous locus proportion statistics unit calculates the proportion of heterozygous loci to the total loci in each window of the second window set of the screening set SNP locus genotype set, thus obtaining the heterozygous locus proportion of each sample; the distribution value calculation unit includes a concentration interval division unit, a distribution value statistics unit, a distribution value comparison unit, and a judgment unit; preferably, the screening set samples are a sample set used to construct the SNP locus window set; preferably, the database is the 1000 Genomes Database; preferably, autosomes do not include chromosomes 13, 18, and 21.

[0019] Furthermore, the concentration range division unit is used to classify the samples in the screening set into the first concentration group, the second concentration group, and the third concentration group according to the fetal concentration range, from low to high.

[0020] Among them, samples with fetal concentration of 0.1% to 10% were designated as the first concentration group; samples with fetal concentration of 10% to 30% were designated as the second concentration group; and samples with fetal concentration greater than 30% were designated as the third concentration group.

[0021] The distribution value statistical unit is used to statistically analyze the distribution value of the proportion of heterozygous sites for each sample in each window of the second window set, in the samples of the first concentration group, the second concentration group, and the third concentration group respectively.

[0022] The distribution value comparison unit compares the distribution values ​​corresponding to the first concentration group, the second concentration group, and the third concentration group pairwise.

[0023] The judgment unit is used to judge the results of the distribution value comparison unit. If the statistical p-value between the pairwise comparisons of distribution values ​​is less than 0.05, the window corresponding to the distribution value of the heterozygous locus proportion is retained to obtain the SNP locus window set.

[0024] Furthermore, the fetal concentration calculation unit also includes a fetal concentration prediction model construction unit, which includes a second genotype set acquisition unit for training set samples, a training set sample vector set acquisition unit, and a model training unit. The second genotype set acquisition unit is used to sequence the training set samples to obtain a set of genotypes for each gene locus in the training set samples, and then perform SNP interpolation on the set to obtain the second genotype set. The training set sample vector set acquisition unit is used to calculate the proportion of heterozygous loci to total loci for each sample in the second genotype set within the SNP locus window, thus obtaining the vector set of the training set samples. The model training unit uses the fetal concentration of each sample in the training set samples as the expected result, inputs the vector set of the training set samples into the model for training, and obtains the fetal concentration prediction model. Preferably, the model is one or more of the following: multivariate logistic regression model, deep neural network, support vector machine, or Bayesian model.

[0025] Furthermore, the sample to be tested includes peripheral blood from pregnant women; preferably, the software for SNP interpolation includes one or more of PLINK software, stitch software, beagle 5.4 software, or BIMBAM software.

[0026] To achieve the above objectives, according to a third aspect of the present invention, an electronic device is provided, comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to perform the above-described method for calculating fetal concentration by executing the executable instructions.

[0027] To achieve the above objectives, according to a fourth aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, the device on which the computer-readable storage medium is located executes the method for calculating fetal concentration as described above.

[0028] By applying the technical solution of this invention and utilizing the fetal concentration calculation method of this application, SNP imputation is performed on the genotype set to ensure that the genotype set contains as many genotypes as possible that are not covered by polymorphic sites. Then, based on the calculation of the proportion of polymorphic sites (SNPs), a highly accurate fetal concentration can be calculated. The fetal concentration calculation method of this application based on the proportion of polymorphic sites can maintain accuracy even with a small amount of data, is not limited by the sex of the fetus, and has a lower detection cost. Compared with the fetal concentration detection methods of the prior art, it is more suitable for promotion. Attached Figure Description

[0029] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0030] Figure 1 A schematic diagram of the calculation method for fetal concentration as described in this application specification is shown.

[0031] Figure 2 A schematic diagram of the electronic device for calculating fetal concentration, as described in this application specification, is shown.

[0032] Figure 3 A hardware block diagram of the method for calculating fetal concentration as described in this application specification is shown.

[0033] Figure 4 This diagram illustrates a comparison between the calculated concentrations of 300 test samples from Example 3 of this application and the known concentrations of fetuses.

[0034] Figure 5 The following graph shows the results of fetal concentration calculation using the method of this application and the seqFF method in Example 4 of this application specification; wherein, Figure 5 Figure A shows the results of fetal concentration calculation using the method described in this application. Figure 5 Figure B shows the results of fetal concentration calculation using the seqFF method.

[0035] Figure 6 The following graph shows the results of fetal concentration calculation using the method of this application and the seqFF method in Example 5 of this application specification; wherein, Figure 6 Figure A shows the results of fetal concentration calculation using the method described in this application. Figure 6 Figure B shows the results of fetal concentration calculation using the seqFF method. Detailed Implementation

[0036] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The present invention will now be described in detail with reference to the embodiments.

[0037] Terminology Explanation:

[0038] SNP (Single Nucleotide Polymorphism) sites: These are sites where a single nucleotide has two or more stable, heritable variations.

[0039] Linkage disequilibrium (LD) is a core concept in population genetics, describing the non-random association of alleles at two or more gene loci in a population.

[0040] As mentioned in the background section, the existing methods for calculating fetal concentration have limitations, including being limited by the sex of the fetus, relying on special detection conditions, high detection costs, or low detection accuracy. Based on this, the inventors in this application attempt to develop a new method for calculating fetal concentration, and thus propose a series of protection schemes in this application.

[0041] In a first typical embodiment of this application, a method for calculating fetal concentration is provided. A schematic flowchart of the fetal concentration calculation method of this application is shown below. Figure 1 As shown, the calculation method includes:

[0042] S1) Sequencing the sample to be tested yields a set of genotypes for each gene locus covered by SNP loci in the database. SNP imputation is then performed on this set to obtain the first genotype set. S2) The genotypes of each gene in the first genotype set are obtained at each SNP locus covered by a window in the SNP locus window set. The proportion of heterozygous loci among the genotypes corresponding to the SNP loci covered by the SNP locus window set in the first genotype set is calculated as the first vector. The SNP locus window set is formed by dividing the reference genome into multiple uniform, continuous, and non-overlapping regions based on chromosome physical length intervals. Each window covers multiple SNP loci. The windows and the SNP loci covered within them constitute the SNP locus window set. S3) The first vector is input into the fetal concentration prediction model to obtain the fetal concentration of the sample to be tested. The SNP locus window set is a set of multiple windows containing SNP loci, continuously divided according to chromosome physical length intervals.

[0043] The aforementioned SNP site window set refers to a series of continuous and non-overlapping regional windows uniformly divided according to the physical length of the chromosome. Each window covers multiple variable SNP sites, forming a comprehensive set containing information on all partitioned windows and their SNP sites. For example, given the complexity of the genome and the distribution characteristics of SNP sites, a specific chromosome (such as chromosome 1) is continuously divided along its physical length according to a preset window size (e.g., every 5kb). If a 2Mb region of chromosome 1 is divided into multiple 5kb windows, then each window will independently carry the SNP information within its own range, and the set of these windows constitutes the aforementioned "SNP site window set".

[0044] This application develops a novel method for calculating fetal concentration based on linear linkage disequilibrium. The method for calculating fetal concentration in this application relies on determining the linkage disequilibrium of SNP loci genotypes. Therefore, the method does not depend on gene fragment distribution characteristics and does not require fragment size calculation. Thus, the method is applicable regardless of whether the sequencing data comes from single-end (SE) sequencing or paired-end (PE) sequencing.

[0045] SNP imputation is a technique that uses genotype data from known single nucleotide polymorphism (SNP) sites to statistically infer or "fill in" genotypes for missing or undetected SNP sites. Currently, SNP imputation is mainly used in: ① Genome-wide association studies (GWAS) to improve the detection efficiency of disease loci; ② Precision medicine in clinical settings to improve pharmacogenomics and rare disease diagnosis; ③ Population genetics research to reconstruct key haplotypes in population evolutionary history; and ④ Genome breeding in plants and animals to optimize the accuracy of genome selection. However, SNP imputation carries the risk of imputation errors due to population matching of reference data and rare variants (MAF < 0.5%).

[0046] Currently, there is no existing method for applying SNP imputation technology to the calculation of fetal concentration. In this application, SNP imputation technology is integrated into the fetal concentration calculation process to enhance data integrity and accuracy. This application considers that in low-depth sequencing environments (e.g., 0.1×), many SNP loci may not have their accurate genotype status directly obtained due to insufficient sequencing reads, which would undoubtedly have an adverse impact on SNP-based fetal concentration calculations. Therefore, in the process of calculating fetal concentration in this application, known high-density SNP locus data (e.g., from large genome databases such as the 1000 Genomes Project) are used to estimate and supplement missing or poor-quality SNP loci in the sequencing samples. This process effectively improves the coverage and accuracy of sample genotypes, providing solid data support for subsequent fetal concentration calculations based on linkage disequilibrium.

[0047] The core steps of SNP imputation can be summarized into two main stages: data preprocessing and imputation calculation. Taking PLINK software as an example: First, strict quality control is performed (plink--bfiledata--geno0.02--hwe 1e-6--maf 0.01--make-bed) to convert the genotyping data into IMPUTE2 format; then, haplotypes are constructed using phase tools (such as Eagle2), and then region imputation is completed by calling a reference panel (such as HRC); finally, the results are merged and low-quality loci are filtered (plink--data outputimpute--score Rsq 0.3) to generate high-density genotype data for downstream analysis. Taking STITCH software as an example, its interpolation process is as follows: First, vcftools is used to generate a VCF file from the sample's BAM; second, based on databases such as the 1000 Genomes Project (1KGP) or the UCSC database as reference data, STITCH is called using R language, setting key parameters (k=4, nGen=100, nCores=8) to complete the interpolation; finally, bcftools is used to evaluate the accuracy of the interpolation, retaining high-confidence interpolation results for downstream analysis. Those skilled in the art can choose any SNP interpolation method to perform SNP interpolation on the set of genotypes whose gene loci in the test samples of this application are covered by SNP loci in the database, all of which can achieve the corresponding results. Those skilled in the art can flexibly select and adjust according to actual needs.

[0048] By incorporating SNP imputation into the fetal concentration calculation in this application, it is possible to obtain near-genome-wide SNP genotypic information even with low data depth, thereby calculating the proportion of heterozygous SNP loci within each window. The richness and accuracy of this proportional data are crucial for constructing a stable and reliable fetal concentration prediction model. More importantly, the introduction of SNP imputation technology ensures that even with low fetal concentrations (e.g., <4%), the method can still accurately reflect the proportion of fetal DNA in maternal plasma, achieving high-precision fetal concentration calculation. This application's introduction of SNP imputation technology into the fetal concentration calculation process of non-invasive prenatal testing overcomes the information loss problem caused by low-depth sequencing data, greatly enhancing the robustness and predictive ability of the calculation model, and providing a more accurate and practical solution for prenatal genetic disease screening.

[0049] The fetal concentration calculation method of this application is not limited by fetal sex and does not require special testing conditions. Fetal concentration can be calculated simply by calculating the proportion of SNP loci. Furthermore, even in extreme cases where the data depth is only 0.1× and the fetal concentration is as low as 4%, the results obtained using the fetal concentration calculation method of this application maintain high accuracy. The fetal concentration calculation method of this application has a simple algorithm and low testing cost, making it more conducive to promotion compared to existing technologies and improving the popularization and application efficiency of NIPT technology.

[0050] In the method for calculating fetal concentration in this application, the sequencing method of the sample to be tested includes, but is not limited to, single-end sequencing or paired-end sequencing; the sequencing depth is not less than 0.1×; the sequencing technology is preferably next-generation sequencing (NGS); the data processing tools for obtaining the set of genotypes of each gene locus in the sample to be tested include, but are not limited to, the mplieup module of samtools, which can be flexibly adjusted and selected by those skilled in the art according to actual conditions.

[0051] Using the samtools mplieup module, the sequencing information for each site in the sample to be tested was obtained (the number of reads covering each site and the number of reads supporting each base). The gene sites covered by SNP sites recorded in high-density SNP site data (e.g., from large genome databases such as the 1000 Genomes Project) in the sample file were then identified. For example, if chr1 2 (chromosome coordinate 2) in the sample file was not recorded as an SNP site in the database, it was not included in the analysis set.

[0052] In a preferred embodiment, the method for obtaining the SNP site window set includes:

[0053] i) Divide the autosomes of the reference genome into a first window set according to the physical length of the chromosomes, with each window being 5M in size. Adjacent windows in the first window set do not overlap. Filter out windows in the first window set that have an N base ratio greater than 10% or fewer than 100 SNP sites to obtain a second window set.

[0054] ii) Traverse the sequencing data of all samples in the screening set to obtain the genotype of each sample at each SNP site in the database, and obtain the genotype set of SNP sites in the screening set; count the proportion of heterozygous sites in each window of the second window set of the genotype set of SNP sites in the screening set, and obtain the proportion of heterozygous sites in each sample.

[0055] iii) Based on the fetal concentration range, the samples in the screening set are classified into a first concentration group, a second concentration group, and a third concentration group from low to high; the distribution values ​​of the heterozygous locus proportion of each sample in each window of the second window set are statistically analyzed for the samples in the first concentration group, the second concentration group, and the third concentration group; the distribution values ​​corresponding to the first concentration group, the second concentration group, and the third concentration group are compared pairwise; if the p-value between the pairwise comparison distribution values ​​is less than 0.05, the window corresponding to the distribution value of the heterozygous locus proportion is retained to obtain the SNP locus window set; preferably, the screening set samples are the sample set used to construct the SNP locus window set; preferably, the database is the 1000 Genomes Database; preferably, autosomes do not include chromosomes 13, 18, and 21.

[0056] Those skilled in the art can flexibly select and set the sample size of the above-mentioned screening set according to the actual situation. The above-mentioned screening set consists of male fetal samples from the SNP locus window set, and the number of male fetal samples in the screening set can be flexibly set and selected according to the actual situation.

[0057] In a preferred embodiment, in the samples of the screening set, samples with a fetal concentration of 0.1% to 10% are designated as the first concentration group; samples with a fetal concentration of 10% to 30% are designated as the second concentration group (excluding the 10% endpoint); and samples with a fetal concentration greater than 30% are designated as the third concentration group.

[0058] In a particularly preferred embodiment, the present invention provides a strategy for selecting a window set of SNP sites for calculating fetal concentrations, a process specifically comprising the following steps:

[0059] i) Window division and initial screening:

[0060] First, based on the physical length of chromosomes, within the range of 2M to 5M, those skilled in the art can select the standard window size and divide the windows according to the actual situation. Based on the selected intervals, the autosomes in the reference genome are systematically segmented to construct the initial first window set. Each window is spatially independent, meaning there is no overlap between adjacent windows, improving the independence and completeness of data analysis.

[0061] Subsequently, the resulting first window set underwent quality control, removing windows with an N base ratio exceeding 10% and regions containing fewer than 100 SNP sites within each window. This was done to eliminate computational biases that might arise from poor data quality or insufficient variable site density, thereby obtaining a cleaner and more effective second window set for subsequent statistical analysis.

[0062] ii) Statistics and classification of heterozygous site proportions:

[0063] The sequencing data of all samples in the selection set were traversed, with the focus on determining the specific genotype of each SNP locus in the 1000-person East Asian population database for each sample, thereby constructing the genotype set of SNP loci in the selection set. Based on this, each window in the second window set was analyzed to calculate the proportion of SNPs belonging to heterozygous loci in the total number of loci in each sample's genotype, thus obtaining the set of heterozygous locus proportions for each sample.

[0064] Furthermore, considering the significant impact of fetal concentration on calculation accuracy, the samples in the screening set were divided into three concentration groups based on their fetal concentration. The first concentration group covered fetal concentrations ranging from 0% to 10%, the second concentration group was in the 10%-30% range, and the third concentration group included samples with concentrations higher than 30%. The distribution of heterozygous sites in each window within the second window set for each group was recorded for subsequent comparative analysis.

[0065] iii) Final determination of the window set:

[0066] The distribution of heterozygous sites in each window of the samples from the three concentration groups was compared pairwise to identify windows that showed significant statistical differences between the different concentration groups, because the SNP sites in these windows were prominent in linkage disequilibrium and could contribute more effective information to the calculation of fetal concentration.

[0067] From the perspective of haploid genomes, humans have 22 autosomes and 2 sex chromosomes. Since the number of sex chromosomes differs between males and females, including all of them in the calculation would be difficult to perform. Furthermore, variations on autosomes 13, 18, and 21 are more frequent and may interfere with the accuracy of fetal chromosome concentration calculations. Therefore, the calculation method in this application only considers the remaining 19 chromosomes. The aforementioned 1000 Genomes database (from http: / / www.internationalgenome.org) refers to the gene database of the 1000 Genomes Project. A subset can be selected based on the mother's population; for example, if the mother is East Asian, the East Asian (EAS) subset from the 1000 Genomes Project database can be selected.

[0068] In a preferred embodiment, the method for constructing the fetal concentration prediction model includes: a) sequencing the training set samples to obtain a set of genotypes for each gene locus in the training set samples, and performing SNP interpolation on the set to obtain a second genotype set; b) calculating the proportion of heterozygous loci to total loci in the SNP locus window set for each sample in the second genotype set, and obtaining a vector set of the training set samples; c) using the fetal concentration of each sample in the training set samples as the expected result, inputting the vector set of the training set samples into the model for training, and obtaining the fetal concentration prediction model; preferably, one or more of the following are used: multivariate logistic regression model, deep neural network, support vector machine, or Bayesian model.

[0069] In this application, the model used to train the aforementioned training set of sample vectors can be any model known to those skilled in the art, and can be flexibly selected according to actual conditions. Those skilled in the art can flexibly select and set the number of samples in the aforementioned training set according to actual circumstances.

[0070] In a preferred embodiment, the sample to be tested includes peripheral blood from a pregnant woman.

[0071] In a preferred embodiment, the SNP interpolation software includes one or more of PLINK software, stitch software, beagle5.4 software, or BIMBAM software.

[0072] In a second typical embodiment of this application, an electronic device for calculating fetal concentration is provided, and a schematic diagram of each unit in the electronic device is shown below. Figure 2 As shown, the electronic device includes a genotype set acquisition unit 01, a first vector calculation unit 02, and a fetal concentration calculation unit 03;

[0073] The unit includes a genotype set acquisition unit, which obtains a set of genotypes for each gene locus in the test sample that are covered by SNP loci in the database after sequencing the sample, and then performs SNP imputation on the set to obtain a first genotype set. A first vector calculation unit is used to obtain the genotypes of each gene in the first genotype set at the SNP loci covered by each window in the SNP locus window set, calculate the proportion of heterozygous loci to the total loci in each window of the first genotype set, and use the set of proportions of heterozygous loci to the total loci for all windows as the first vector. The SNP locus window set is defined as follows: the reference genome is divided into multiple uniform, continuous, and non-overlapping regions according to the chromosome physical length interval, with each window covering multiple SNP loci; the windows and the SNP loci covered by the windows constitute the SNP locus window set.

[0074] The fetal concentration calculation unit is used to input the first vector into the fetal concentration prediction model to obtain the fetal concentration of the sample to be tested; the SNP site window set is a set of multiple windows containing SNP sites that are continuously divided into the reference genome according to the physical length interval of the chromosome.

[0075] In a preferred embodiment, the first vector calculation unit further includes an SNP site window set acquisition unit, which includes a window partitioning unit, a window filtering unit, a screening set sample SNP site genotype set acquisition unit, a heterozygous site proportion statistics unit, and a distribution value calculation unit.

[0076] The window division unit is used to divide the autosomes of the reference genome according to the physical length of the chromosome, with a window size of 2-5M, to obtain the first window set. Adjacent windows in the first window set do not overlap. The window filtering unit is used to filter the windows in the first window set with an N base ratio greater than 10% or fewer than 100 SNP sites to obtain the second window set.

[0077] The selection set sample SNP locus genotype set acquisition unit is used to traverse the sequencing data of all samples in the selection set, obtain the genotype of each SNP locus in the database for each sample, and obtain the selection set SNP locus genotype set; the heterozygous locus proportion statistics unit calculates the proportion of heterozygous loci in each window of the selection set SNP locus genotype set in the second window set, and obtains the heterozygous locus proportion of each sample; the distribution value calculation unit includes a concentration interval division unit, a distribution value statistics unit, a distribution value comparison unit, and a judgment unit; preferably, the database is the 1000-person genome database; preferably, autosomes do not include chromosomes 13, 18, and 21; preferably, the selection set sample is a sample set used to construct the SNP locus window set.

[0078] In a preferred embodiment, the concentration range division unit is used to classify the samples in the screening set into a first concentration group, a second concentration group, and a third concentration group according to the fetal concentration range, from low to high; wherein, samples with a fetal concentration of 0.1% to 10% are denoted as the first concentration group; samples with a fetal concentration of 10% to 30% are denoted as the second concentration group; and samples with a fetal concentration greater than 30% are denoted as the third concentration group.

[0079] The distribution value statistics unit is used to separately count the distribution values ​​of the heterozygous locus proportion of each sample in the first concentration group, the second concentration group, and the third concentration group, in each window of the second window set; the distribution value comparison unit compares the distribution values ​​corresponding to the first, second, and third concentration groups pairwise; the judgment unit is used to judge the results of the distribution value comparison unit. If the p-value between the pairwise comparison distribution values ​​is less than 0.05, the window corresponding to the distribution value of the heterozygous locus proportion is retained to obtain the SNP locus window set.

[0080] In a preferred embodiment, the fetal concentration calculation unit further includes a fetal concentration prediction model construction unit, which includes a second genotype set acquisition unit for training set samples, a training set sample vector set acquisition unit, and a model training unit.

[0081] The second genotype set acquisition unit is used to sequence the training set samples to obtain a set of genotypes for each gene locus in the training set samples, and to perform SNP interpolation on the set to obtain the second genotype set; the training set sample vector set acquisition unit is used to compare the second genotype set with the SNP locus window set, calculate the proportion of heterozygous loci to the total loci, and obtain the vector set of the training set samples; the model training unit is used to use the known fetal concentration of each sample in the training set samples as the expected result, input the vector set of the training set samples into the model for training, and obtain the fetal concentration prediction model; preferably, the model is one or more of the following: multivariate logistic regression model, deep neural network, support vector machine or Bayesian model.

[0082] In a preferred embodiment, the sample to be tested includes peripheral blood from a pregnant woman; preferably, the SNP interpolation software includes one or more of PLINK software, stitch software, beagle 5.4 software, or BIMBAM software.

[0083] In a third typical embodiment of this application, an electronic device is provided, comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to perform the above-described method for calculating fetal concentration by executing the executable instructions.

[0084] In a fourth typical embodiment of this application, a computer-readable storage medium is provided, which includes a stored computer program, wherein the computer program controls the device where the computer-readable storage medium is located to execute the above-described method for calculating fetal concentration when it is running.

[0085] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, because according to this invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions involved are not necessarily essential to this application.

[0086] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus hardware devices such as detection devices. Based on this understanding, the data processing part of the technical solution of this application can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of the embodiments of this application.

[0087] This application can be used in a wide range of general-purpose or special-purpose computing system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics devices, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices.

[0088] The method provided in this application can be executed on a terminal, computer terminal, or similar computing device. Taking running on a terminal as an example, Figure 3 This is a hardware structure block diagram of the calculation model of the fetal concentration calculation method according to an embodiment of the present invention. For example... Figure 3 As shown, a terminal may include one or more ( Figure 3 Only one is shown in the diagram. A processor A1 (processor A1 may include, but is not limited to, a microprocessor (MCU) or a programmable logic device (FPGA), etc.) and a memory B1 for storing data are also shown. Optionally, the terminal may further include a transmission device C1 for communication functions and an input / output device D1. Those skilled in the art will understand that... Figure 3 The structure shown is for illustrative purposes only and does not limit the structure of the terminal described above. For example, the terminal may also include components that are more... Figure 3 The more or fewer components shown, or having the same Figure 3 The different configurations shown.

[0089] Memory B1 can be used to store computer programs, such as application software programs and modules, like the computer programs corresponding to the methods of segment concatenation, clustering, and consistency processing in this embodiment of the invention. Processor A1 executes various functional applications and data processing by running the computer programs stored in memory B1, thereby implementing the methods described above. Memory B1 may include high-speed random access memory and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, memory B1 may further include memory remotely located relative to processor A1, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0090] Transmission device C1 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the terminal's communication provider. In one example, transmission device C1 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, transmission device C1 may be a Radio Frequency (RF) module used for wireless communication with the Internet.

[0091] Obviously, those skilled in the art should understand that some modules or steps of this application described above can be implemented in general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby storing them in a storage device for execution by a computing device, or fabricating them separately as individual integrated circuit modules, or fabricating multiple modules or steps into a single integrated circuit module. Thus, this application is not limited to any particular hardware and software combination.

[0092] The beneficial effects of this application will be explained in more detail below with reference to specific embodiments.

[0093] Example 1

[0094] 1. Window division: Except for the two sex chromosomes and chromosomes 13, 18, and 21, the remaining 19 chromosomes in the reference genome are divided into windows of 5M each (adjacent windows do not overlap) to obtain the first window set;

[0095] 2. Calculate the proportion of N bases in the reference genome sequence corresponding to each window in the first window set, and the SNP sites covered by the East Asian population in each window of the 1000 Genomes Study. Filter out windows with more than 10% N bases or fewer than 100 SNP sites to obtain the second window set.

[0096] 3. One hundred male fetal samples were selected as the screening set. Fetal concentration was calculated based on the Y chromosome as the gold standard. The fetal concentration of the 100 male fetal samples was calculated and divided into three groups according to the fetal concentration: 0.1%-10% was the first concentration group, 10-30% (excluding the 10% endpoint value) was the second concentration group, and greater than 30% was the third concentration group.

[0097] One hundred male fetal samples were subjected to single-end sequencing (SE50) at a sequencing depth of 5×. The whole genome variant analysis was performed using the software samtoolsmplieup to obtain the genotype of each polymorphic site in the East Asian population database of the 1000-person group, i.e., the selection set SNP site genotype set.

[0098] In the statistical screening set of SNP loci genotypes, in each window of the second window set, the genotype of each sample is the proportion of heterozygous loci to the total loci, thus obtaining the proportion of heterozygous loci for each sample;

[0099] 4. Statistically analyze the distribution of heterozygous site proportions for each sample in each window of the second window set, for samples in the first concentration group, the second concentration group, and the third concentration group respectively.

[0100] The distribution values ​​corresponding to the first, second, and third concentration groups are compared pairwise.

[0101] If there is a statistical difference, the window corresponding to the distribution value of the heterozygous locus proportion is retained to obtain the SNP locus window set, resulting in a total of 507 windows.

[0102] Example 2

[0103] I. Training Reference Set

[0104] 1. Using 1000 male singleton fetuses as the training set, single-end SE sequencing was performed on each sample at a sequencing depth of 0.4×.

[0105] 2. Calculate the fetal concentration of 1000 male fetuses based on the Y chromosome as the gold standard (calculate the proportions a and b of the number of reads aligned to the X and Y chromosomes to the total number of reads in the sample, respectively, Y concentration = 2b / (a+b)), in preparation for the next step of training;

[0106] 3. Use samtools mplieup to calculate the genotypes of the loci covered by 1000 samples, and use STITCH to perform SNP imputation. Correct the genotypes obtained by samtools to obtain the final genotype set, i.e., the second genotype set.

[0107] 4. Statistically analyze the training set samples. Each sample is in a set of 507 windows (SNP locus windows). The proportion of heterozygous loci in each window is the total number of loci. Each sample yields a total of 507 values, which are combined into a vector to obtain the vector set of the training set samples.

[0108] 5. Using the multivariate logistic regression algorithm, the fetal concentration of the training set samples is used as the expected result, and 507×1000 (the vector set of the training set samples) is used as the input of the observation value for training to obtain the fetal concentration prediction model.

[0109] II. Calculation of fetal concentration

[0110] 1. The male fetal samples to be tested (with a known fetal concentration of 14.2%, samples not in the training set) were subjected to single-end SE sequencing at a sequencing depth of 0.4×.

[0111] 2. Obtain the sequencing data of the sample to be tested, use samtools mplieup to obtain the genotypes at each sequencing point, and then use STITCH to perform SNP interpolation to obtain the genotypes of the sample to be tested at each point, i.e., the first genotype set;

[0112] 3. Calculate the proportion of heterozygous sites to the total number of sites in each window (SNP site window set) of the sample to be tested. Calculate the proportion of 507 windows to obtain a vector containing 507 elements, which is the first vector.

[0113] 4. Inputting the first vector into the fetal concentration prediction model, the fetal concentration of the sample was found to be 14.6%, which is very close to the value obtained by the gold standard detection.

[0114] Example 3

[0115] 1. Using 300 male fetal samples (excluding training and selection sets), single-end SE sequencing was performed at a sequencing depth of 0.4×.

[0116] 2. After obtaining the sequencing data of the above samples, use samtools mplieup to obtain the genotypes of each sequencing locus, and then use STITCH to perform SNP interpolation to obtain the genotypes of the sample to be tested at each locus.

[0117] 3. Calculate the proportion of heterozygous sites to the total number of sites in each window of the sample to be tested. Calculate the proportion of the 507 windows (SNP site window set). Each sample will be given a vector containing 507 elements (the first vector).

[0118] 4. Input the first vector of the above samples into the fetal concentration prediction model to obtain the fetal concentration results of 300 samples. Compare these results with the known fetal concentrations of these 300 samples and calculate the correlation coefficient, which reaches 0.967. The results are as follows. Figure 4 As shown, Figure 4 The "cor" in this context is an abbreviation for "correlation," which translates to correlation coefficient (the meanings of "cor" or "r" in other accompanying figures are the same). The correlation coefficient, originally designed by statistician Carl Pearson, is a statistical indicator used to study the degree of linear correlation between variables. It is generally represented by the letters r, ρx,y, or Corr(X,Y). The correlation coefficient is a dimensionless quantity used to describe the linear relationship between X and Y, unaffected by units. A higher correlation coefficient indicates a stronger linear correlation between the two values, leading to more accurate fetal concentration calculations.

[0119] Example 4

[0120] One hundred male fetal samples were collected, library constructed, and sequenced to obtain FASTQ files. The FASTQ files were preprocessed, aligned to a reference genome, deduplicated, and the uniquely aligned sequences were selected to obtain BAM files. Fetal concentrations in these 100 samples were calculated based on the Y chromosome, serving as the gold standard for comparative analysis. Downsampling was performed based on the sample data volume, reducing the sequencing depth of all samples to 0.1×. Fetal concentrations were then calculated using both the method described in Example 3 of this application and seqFF, and the accuracy of the two methods was compared. The results are as follows: Figure 5 As shown.

[0121] Using the method of this application embodiment, the calculated correlation coefficient can reach 0.88, while the seqFF method can only reach 0.453, which proves the superiority of the calculation method of this application.

[0122] Example 5

[0123] One hundred male fetal samples with low fetal concentrations (all known to be <4%) were collected, library constructed, and sequenced to obtain FASTQ files. The FASTQ files were preprocessed, aligned to a reference genome, deduplicated, and the uniquely aligned sequences were selected to obtain BAM files. Fetal concentrations of these 100 samples were calculated based on the Y chromosome as the gold standard for comparative analysis. Downsampling was performed based on the sample size, reducing all samples to 0.1×. Fetal concentrations were then calculated using both the method described in Example 3 and seqFF, and the accuracy of the two methods was compared. Results are as follows: Figure 6 As shown.

[0124] Using the calculation method of this application, the correlation coefficient can reach 0.949, while the seqFF method can only reach 0.718, which proves the superiority of the calculation method of this application.

[0125] As can be seen from the above description, the embodiments of the present invention achieve the following technical effects: The method for calculating fetal concentration in this application avoids reliance on additional experimental data and overcomes the limitations of fetal sex determination. Furthermore, even with low fetal concentrations, the method of this application can still obtain relatively accurate results. The method for calculating fetal concentration in this application has a simple algorithm. Applying it to non-invasive prenatal testing (NIPT) can reduce the overall cost and complexity of the operation, while maintaining high accuracy. It provides a more reliable, practical, and economical testing tool for clinical use, and is expected to improve the accuracy and efficiency of early diagnosis of genetic diseases, thus promoting the development of non-invasive prenatal testing technology.

[0126] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method of calculating a concentration of a fetus, characterized by, The calculation method comprises: S1) sequencing the sample to be tested to obtain a set of genotypes of each gene locus covered by the SNP locus in the database in the sample to be tested, and then performing SNP interpolation on the set to obtain a first genotype set; S2) obtaining the genotypes of each window in the SNP locus window set covered by each window of each gene in the first genotype set, and calculating the proportion of heterozygous sites in the total sites in the corresponding genotype of each window of the first genotype set; The set of proportions of the total sites of the heterozygous sites corresponding to all the windows is taken as a first vector; The SNP locus window set is: according to the chromosome physical length interval, the reference genome is divided into a plurality of uniform, continuous and non-overlapping region windows, and a plurality of SNP loci are covered in each window; the window and the SNP loci covered in the window constitute the SNP locus window set; S3) inputting the first vector into a fetal concentration prediction model to obtain the fetal concentration of the sample to be tested.

2. The computational method of claim 1, wherein, The method for obtaining the SNP locus window set comprises: i) dividing the autosomes of the reference genome according to the chromosome physical length, taking 2-5M as the size of one window, to obtain a first window set, and the adjacent windows in the first window set are non-overlapping; Filtering the windows with an N base proportion greater than 10% or less than 100 SNP loci in the first window set to obtain a second window set; ii) traversing the sequencing data of all samples in the screening set to obtain the genotype of each sample at each SNP locus in the database, and obtaining a SNP locus genotype set of the screening set; Statistically calculating the proportion of heterozygous sites of each sample in the total sites in each window of the second window set in the SNP locus genotype set of the screening set to obtain the proportion of heterozygous sites of each sample; iii) classifying the samples in the screening set from low to high according to the fetal concentration interval into a first concentration group, a second concentration group and a third concentration group; Statistically calculating the distribution value of the proportion of heterozygous sites of each sample in each window of the second window set in the first concentration group, the second concentration group and the third concentration group, respectively; Comparing the distribution values corresponding to the first concentration group, the second concentration group and the third concentration group with each other; If the P value in the statistical comparison between the distribution values is less than 0.05, the window corresponding to the distribution value of the proportion of heterozygous sites is retained to obtain the SNP locus window set; Preferably, the sample set for constructing the SNP locus window set is the screening set sample; Preferably, the database is a thousand genome database; Preferably, the autosomes do not include chromosome 13, chromosome 18 and chromosome 21.

3. The computational method of claim 2, wherein, In the screening set sample, The sample with a fetal concentration of 0.1-10% is recorded as the first concentration group; The sample with a fetal concentration of 10-30% is recorded as the second concentration group; The sample with a fetal concentration greater than 30% is recorded as the third concentration group.

4. The computational method of claim 1, wherein, The construction method of the fetal concentration prediction model comprises: a) sequencing the training set samples to obtain a set of genotypes of each genetic locus in the training set samples, and performing SNP imputation on the set to obtain a second genotype set; b) calculating the proportion of heterozygous loci in the set of genotypes of each sample in the second genotype set at the SNP locus window set to obtain a vector set of the training set samples; c) inputting the vector set of the training set samples into a model to train the model with the fetal concentration of each sample in the training set samples as the expected result to obtain the fetal concentration prediction model; Preferably, the model comprises one or more of a multivariate logistic regression model, a deep neural network, a support vector machine or a Bayesian model.

5. The computational method of claim 1, wherein, The sample to be tested comprises peripheral blood of a pregnant woman.

6. The computational method of claim 1, wherein, The SNP imputation software comprises one or more of PLINK software, stitch software, beagle5.4 software or BIMBAM software.

7. An electronic device for calculating fetal concentration, characterized in that, The electronic device comprises a genotype set acquisition unit, a first vector calculation unit and a fetal concentration calculation unit; The genotype set acquisition unit is configured to sequence the sample to be tested to obtain a set of genotypes of each genetic locus covered by the SNP loci in the database in the sample to be tested, and perform SNP imputation on the set to obtain a first genotype set; The first vector calculation unit is configured to obtain the genotypes of each gene in the first genotype set at the SNP loci covered by each window in the SNP locus window set, calculate the proportion of heterozygous loci in the total loci in the corresponding genotype of each window, and obtain a first vector by collecting the proportions of the heterozygous loci in the total loci corresponding to all the windows. The SNP locus window set is obtained by dividing the reference genome into a plurality of regions with the same size, continuity and non-overlapping according to the physical length interval of the chromosome, and each window covers a plurality of SNP loci. The fetal concentration calculation unit is configured to input the first vector into the fetal concentration prediction model to obtain the fetal concentration of the sample to be tested. 8.The electronic device of claim 7, wherein, The first vector calculation unit further comprises an SNP locus window set acquisition unit, which comprises a window division unit, a window filtering unit, a screening set sample SNP locus genotype set acquisition unit, a heterozygous locus proportion statistical unit and a distribution value calculation unit. The window division unit is configured to divide the autosomes of the reference genome into a first window set by taking 2-5M as the size of one window according to the physical length of the chromosome, and the adjacent windows in the first window set are non-overlapping. The window filtering unit is configured to filter the windows with an N base proportion greater than 10% or less than 100 SNP loci in the first window set to obtain a second window set. The screening set sample SNP locus genotype set acquisition unit is configured to traverse the sequencing data of all samples in the screening set to obtain the genotype of each sample at each SNP locus in the database and obtain a screening set SNP locus genotype set. The fetal concentration calculation unit is configured to input the first vector into the fetal concentration prediction model to obtain the fetal concentration of the sample to be tested. The hybrid site proportion statistical unit is configured to calculate the proportion of hybrid sites in the genotype set of each sample in each window in the second window set, and obtain the proportion of hybrid sites of each sample; The distribution value calculation unit comprises a concentration interval division unit, a distribution value statistical unit, a distribution value comparison unit, and a judgment unit; Preferably, the screening set sample is a sample set used for constructing the SNP site window set. Preferably, the database is a thousand human genomes database. Preferably, the autosomes do not include chromosome 13, chromosome 18, and chromosome 21. 9.The electronic device of claim 8, wherein, The concentration interval division unit is configured to classify the samples in the screening set into a first concentration group, a second concentration group, and a third concentration group from low to high according to the fetal concentration interval. The samples with a fetal concentration of 0.1-10% are recorded as the first concentration group; the samples with a fetal concentration of 10-30% are recorded as the second concentration group; and the samples with a fetal concentration greater than 30% are recorded as the third concentration group. The distribution value statistical unit is configured to respectively calculate the distribution value of the proportion of hybrid sites of the samples in the first concentration group, the samples in the second concentration group, and the samples in the third concentration group in each window in the second window set. The distribution value comparison unit is configured to compare the distribution values corresponding to the first concentration group, the second concentration group, and the third concentration group in pairs. The judgment unit is configured to judge the result of the distribution value comparison unit. If the P value between the distribution values in the pair comparison is less than 0.05, the window corresponding to the distribution value of the proportion of hybrid sites is retained, and the SNP site window set is obtained. 10.The electronic device of claim 7, wherein, The fetal concentration calculation unit further comprises a fetal concentration prediction model construction unit, and the fetal concentration prediction model construction unit comprises a second genotype set acquisition unit of training set samples, a training set sample vector set acquisition unit, and a model training unit. The second genotype set acquisition unit is configured to sequence the training set samples to obtain a set of genotypes of each genetic site in the training set samples, and perform SNP interpolation on the set to obtain a second genotype set. The training set sample vector set acquisition unit is configured to calculate the proportion of hybrid sites in the SNP site window set in the genotype of each sample in the second genotype set, and obtain a vector set of the training set samples. The model training unit is configured to take the fetal concentration of each sample in the training set samples as an expected result, input the vector set of the training set samples into a model for training, and obtain the fetal concentration prediction model. Preferably, the model is one or more of a multivariate logistic regression model, a deep neural network, a support vector machine, or a Bayesian model. 11.The electronic device of claim 7, wherein, The sample to be tested comprises peripheral blood of a pregnant woman; and preferably, the software for SNP interpolation comprises one or more of PLINK software, stitch software, beagle5.4 software, or BIMBAM software.

12. An electronic device, comprising: The electronic device comprises: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to perform the method of calculating the fetal concentration of any one of claims 1 to 6 via execution of the executable instructions.

13. A computer-readable storage medium, characterized in that, The computer readable storage medium comprises a stored computer program, wherein the computer readable storage medium controls the device where the computer readable storage medium is located to perform the method of calculating the fetal concentration of any one of claims 1 to 6 when the computer program is running.

Images