A data correction method before copy number variation detection
By using the Holt linear method and the Loess locally weighted regression model to perform GC correction on sequencing data, the problem of uneven distribution of sequencing sequences in CNV detection was solved, improving the accuracy and consistency of the data. This method is applicable to other techniques for CNV detection.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SUZHOU BASECARE MEDICAL DEVICE CO LTD
- Filing Date
- 2022-05-12
- Publication Date
- 2026-05-15
AI Technical Summary
Existing CNV detection methods have high false positive and false negative rates in data analysis, and existing GC correction methods have failed to effectively reduce the fluctuations in the distribution of sequencing sequences on the genome, resulting in heterogeneous sequencing results.
Holt linear method was used for exponential smoothing combined with the Loess local weighted regression model to perform GC correction on sequencing data. Exponential smoothing reduced the distribution fluctuation of sequencing sequences on the genome, and correction weights were used for further adjustment.
It effectively reduces the uneven distribution of sequencing sequences on the genome, improves the accuracy and consistency of sequencing data, is easily extended to other CNV detection techniques, and improves the reliability of CNV detection.
Smart Images
Figure CN114944195B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of gene detection technology and relates to a data correction method before copy number variation detection. Background Technology
[0002] Copy number variations (CNVs) are a clinically important class of structural variations. Small CNVs are usually benign, but CNVs larger than 250 kb are closely associated with pathological consequences such as developmental disorders and cancer. Identifying intra- and inter-population CNVs is crucial for a better understanding of genomic plasticity and elucidating their potential contribution to disease or phenotypic traits. Although the link between SNPs and disease susceptibility has been well investigated, genome-wide association studies of CNVs are still scarce to date. This may be because CNV analysis is still slightly more complex than SNP analysis, leading to higher false positive and unknown false negative rates in bioinformatics work.
[0003] Next-generation sequencing (NGS) provides a rapid and inexpensive platform for identifying genomic copy number variations (CNVs), offering high resolution and sensitivity. However, the data generated still presents challenges in subsequent analysis. This is because imbalances in DNA library amplification are introduced during the DNA library construction process, such as primer selection, library fragment size choice, PCR amplification, and sequencing error probability. Studies have shown that PCR plays a dominant role in various stages before sequencing (Aird D, Ross MG, Chen WS, et al. Analyzing and minimizing PCR amplification bias in Illuminasequencing libraries[J]. Genome Biol, 2011, 12(2): R18.). Structural issues within the genome itself, such as GC-rich genomic regions often being over-amplified, and local structures of DNA or chromatin leading to uneven sequencing sequence coverage, result in an uneven distribution of the final sequencing results across the genome. Therefore, before CNV identification, it is essential to remove the distribution bias of the sequencing sequence across the genome to reduce technically induced noise and identify true CNVs.
[0004] GC content is one of the more clearly understood factors influencing the distribution of sequencing sequences. Most current methods correct for the GC effect, including using regression models such as Loess's (Boeva V, Zinovyev A, Bleakley K, et al. Control-free calling of copy number alterations in deep-sequencing data using GC-content normalization[J]. Bioinformatics, 2011, 27(2):268-9.) to correct the GC effect on the depth of coverage (DOC) of sequencing sequences within a specific window. In addition, some studies correct GC bias by comparing the number of sequences aligned to the genome at the nucleotide level rather than the genomic window level. However, the current correction methods still result in significant fluctuations in the distribution of sequences across the genome, failing to achieve satisfactory noise reduction.
[0005] In summary, providing a noise reduction method to reduce the fluctuations in the distribution of sequencing sequences across the genome, so that sequencing data can more accurately reflect the coverage of the sample genome, is one of the urgent problems to be solved in the field of CNV detection. Summary of the Invention
[0006] To address the shortcomings of existing technologies and practical needs, this invention provides a data correction method before copy number variation detection. This method can effectively reduce the fluctuations in the distribution of sequencing sequences on the genome, enabling the sequencing data to more accurately reflect the coverage of the sample genome.
[0007] To achieve the above objectives, the present invention adopts the following technical solution:
[0008] In a first aspect, the present invention provides a data correction method for copy number variation detection, the method comprising the following steps:
[0009] (1) Obtain the GC content and depth coverage value DOC under each window, and filter out the regions with GC content or DOC of 0;
[0010] (2) Exponential smoothing: The Holt linear method is used to perform exponential smoothing of the DOC value;
[0011] Alternatively, DOC can be exponentially smoothed using formula (1) to obtain DOC. ets ;
[0012] DOC ets =ets(DOC) Formula (1),
[0013] Where ets is a function in the forecast package of the R language;
[0014] (3) Using the locally weighted regression model Loess for DOC ets Perform GC correction, obtain the correction weight for each window, and multiply the inverse of the correction weight by the DOC. ets Obtain DOC loess ;
[0015] (4) According to formula (2), DOC loess Divide by the DOC in all windows loess The median value is used to obtain the final corrected DOC for each window, denoted as DOC. final ;
[0016] DOC final =DOC loess / median(DOC loess ) formula (2),
[0017] median(DOC loess ) indicates DOC under all windows loess The median.
[0018] In this invention, the Holt linear method with multiplicative error is used to perform exponential smoothing of the DOC value, and the specific formula is as follows:
[0019] By specifying the training error of the previous step as a relative error, we can obtain:
[0020] ε t =(y t -(l t-1 +b t-1 )) / (l t-1 +b t-1 (3);
[0021] Prediction equation: y t =(l t-1 +b t-1 (1+ε) t (4);
[0022] Horizontal equation: l t =(l t-1 +b t-1 (1+αε) t (5);
[0023] Trend equation: b t =b t-1 +β(l t-1 +b t-1 )ε t (6);
[0024] Where β=αβ* and ε t ~NID(0,σ2) (7).
[0025] Where y t This indicates that the predicted value of the time series at time t is determined by the level estimate, trend estimate, and relative error at time t-1; t This represents the estimated level of the time series at time t, where α is the smoothing parameter for the level (0 ≤ α ≤ 1); the level equation indicates that l t It is determined by the horizontal estimate, trend estimate, relative error, and horizontal smoothing parameter at time t-1; b t This represents an estimate of the trend (slope) of the time series at time t, where β* is the smoothing parameter for the trend 0 ≤ β* ≤ 1; the trend equation indicates that b t It is determined by the horizontal estimate, trend estimate, relative error, horizontal smoothing parameter, and trend smoothing parameter at time t-1.
[0026] Alternatively, the `ets` function in the `forecast` package of the R language can be used to directly perform exponential smoothing on the DOC value and tune the parameters α, β*, and initial l and b.
[0027] In this invention, the exponential smoothing method is applied to data denoising before copy number variation (CNV) detection in the context of high-throughput sequencing, providing a data correction method for CNV detection. This method can effectively reduce the uneven distribution of sequencing sequences on the genome caused by library preparation, sequencing, and other technologies, and is easily extended to other technologies for CNV detection, such as CNV detection based on whole exome sequencing (WES). It is also easy to integrate into CNV detection devices and has high application value.
[0028] Preferably, step (1) includes steps of aligning sequences and dividing windows.
[0029] Preferably, the alignment sequence includes acquiring sequencing data of the genome of the sample to be tested, aligning it with a reference genome, sorting it according to chromosomes, and removing duplicates from the alignment results.
[0030] Preferably, the windowing process includes dividing the reference genome into windows according to a given window size, generating the GC content corresponding to each window, and counting the number of sequences under each window, which is recorded as the depth coverage value (DOC) under each window.
[0031] As a preferred technical solution, the data correction method for copy number variation detection includes the following steps:
[0032] (1) Obtain sequencing data of the genome of the sample to be tested, compare it with the reference genome, sort it according to chromosomes, and remove duplicates from the comparison results;
[0033] (2) Divide the reference genome into windows according to the given window size, generate the GC content corresponding to each window, and count the number of sequences under each window, which is recorded as the depth coverage value (DOC) under each window;
[0034] (3) Filter out regions where the GC content or DOC is 0;
[0035] (4) Use the Holt linear method to perform exponential smoothing of DOC values;
[0036] Alternatively, DOC can be exponentially smoothed using formula (1) to obtain DOC. ets ;
[0037] (5) Using the locally weighted regression model Loess for DOC ets Perform GC correction, obtain the correction weight for each window, and multiply the inverse of the correction weight by the DOC. ets Obtain DOC loess ;
[0038] (6) According to formula (2), DOC loess Divide by the DOC in all windows loess The median value is used to obtain the final corrected DOC for each window, denoted as DOC. final .
[0039] Secondly, the present invention provides a method for detecting copy number variations for purposes other than disease diagnosis and / or treatment, the method comprising:
[0040] The sequencing data of the sample genome is corrected using the data correction method for copy number variation detection described in the first aspect, and copy number variation is determined.
[0041] This invention provides a method for detecting copy number variations for purposes other than disease diagnosis and / or treatment, which can be applied to scientific research related to copy number variations.
[0042] Preferably, the copy number variation detection method for purposes other than disease diagnosis and / or treatment includes the following steps:
[0043] (1') Obtain sequencing data of the genome of the sample to be tested, compare it with the reference genome, sort it according to chromosomes, and remove duplicates from the comparison results;
[0044] (2') Divide the reference genome into windows according to the given window size, generate the GC content corresponding to each window, and count the number of sequences under each window, which is recorded as the depth coverage value (DOC) under each window;
[0045] (3') Filter out regions where the GC content or DOC is 0;
[0046] (4') Use the Holt linear method to perform exponential smoothing of DOC values;
[0047] Alternatively, DOC can be exponentially smoothed using formula (1) to obtain DOC. ets ;
[0048] (5') Using the locally weighted regression model Loess for DOC ets Perform GC correction, obtain the correction weight for each window, and multiply the inverse of the correction weight by the DOC. ets Obtain DOC loess ;
[0049] (6') According to formula (2), DOC loess Divide by the DOC in all windows loess The median value is used to obtain the final corrected DOC for each window, denoted as DOC. final ;
[0050] (7') DOC final Perform data sharding, analyze the differences in the sharded data, and report copy number variation information.
[0051] Thirdly, the present invention provides a copy number variation detection device, the copy number variation detection device comprising:
[0052] Sequence alignment module: Used to acquire sequencing data of the genome of the sample to be tested and compare it with a reference genome;
[0053] Window partitioning module: connected to the sequence alignment module, used to partition the reference genome into windows according to a given window size, generate the GC content corresponding to each window, and count the number of sequences in each window;
[0054] Data correction module: connected to the window division module, used to execute the steps in the data correction method for copy number variation detection described in the first aspect;
[0055] Copy number mutation determination module: connected to the data correction module, used to determine copy number mutation.
[0056] Preferably, the sequence alignment module is used to perform the following steps:
[0057] (1”) Obtain sequencing data of the genome of the sample to be tested, compare it with the reference genome, and sort it according to chromosomes;
[0058] (2) Remove duplicates from the comparison results.
[0059] Preferably, the window partitioning module is used to perform the following steps:
[0060] The reference genome is divided into windows according to the given window size, the GC content corresponding to each window is generated, and the number of sequences under each window is counted and recorded as the depth coverage value (DOC) under each window.
[0061] Preferably, the copy number mutation determination module is used to perform the following steps:
[0062] The DOCs under each window obtained by the data correction module are fragmented, and the fragmented data are used for difference analysis to report copy number variation information.
[0063] Fourthly, the present invention provides a computer device including a memory and a processor, the memory storing a computer program that executes the steps of the data correction method for copy number variation detection described in the first aspect or the copy number variation detection method for non-disease diagnosis and / or treatment purposes described in the second aspect.
[0064] Fifthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, the computer program performing steps in the data correction method for copy number variation detection described in the first aspect or the copy number variation detection method for non-disease diagnosis and / or treatment purposes described in the second aspect.
[0065] Compared with the prior art, the present invention has the following beneficial effects:
[0066] This invention creatively applies the exponential smoothing method to data denoising before copy number variation (CNV) detection in the context of high-throughput sequencing, providing a data correction method for CNV detection. This method can effectively reduce the uneven distribution of sequencing sequences on the genome caused by library preparation, sequencing, and other technologies, and is easily extended to other technologies for CNV detection, such as CNV detection based on whole exome sequencing (WES). It is also easy to integrate into CNV detection devices and has high application value. Attached Figure Description
[0067] Figure 1 This is a flowchart of the experiment in Embodiment 1 of the present invention;
[0068] Figure 2 A graph showing the relationship between DOC and CG content in the original data;
[0069] Figure 3 A graph showing the relationship between DOC and CG content using only Loess-corrected data;
[0070] Figure 4 This is a graph showing the relationship between DOC and CG content in the corrected data of Example 1 of the present invention;
[0071] Figure 5 This is a distribution map of DOCs in the genome from the original data;
[0072] Figure 6 A map showing the distribution of DOCs across the genome using only Loess-corrected data;
[0073] Figure 7 This is a map showing the distribution of DOCs in the genome of the corrected data in Example 1 of this invention;
[0074] Figure 8 This is a distribution map of DOCs in the gene rearrangement region of chromosome 14 from the original data.
[0075] Figure 9 A map showing the distribution of DOCs in the gene rearrangement region of chromosome 14 using only Loess-corrected data;
[0076] Figure 10 This is a distribution map of the DOC of the corrected data in the gene rearrangement region of chromosome 14 in Example 1 of the present invention. Detailed Implementation
[0077] To further illustrate the technical means and effects of this invention, the following description, in conjunction with embodiments and accompanying drawings, provides a further explanation of the invention. It is understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it.
[0078] Where specific techniques or conditions are not specified in the examples, they shall be performed in accordance with the techniques or conditions described in the literature in this field, or in accordance with the product instructions. Reagents or instruments whose manufacturers are not specified are all conventional products that can be purchased through legitimate channels.
[0079] Example 1
[0080] This embodiment uses the method of the present invention to perform data correction before copy number variation detection, as shown in the flowchart below. Figure 1 As shown, the specific steps are as follows:
[0081] 1. DNA library construction and sequencing of samples.
[0082] In this embodiment, the test sample was cultured human B cells. The DNA library construction method was performed according to the product instructions (reagent catalog number: J000027) provided by Suzhou Beikang Medical Devices Co., Ltd. The sequencing was performed using the DA500 high-throughput sequencing platform of Suzhou Beikang Medical Devices Co., Ltd. The sequencing type was single end sequencing 150, and the number of sequencing sequences was about 20M.
[0083] 2. Segment the sequencing data
[0084] The sequencing data was randomly cut into 10M, 15M and 20M segments;
[0085] 3. Align the sequences to the reference genome, sort them, and remove duplicates.
[0086] The obtained offline data was compared with the human reference genome (UCSC hg19), and the generated BAM files were sorted and duplicates were removed.
[0087] 4. Window division of the reference genome
[0088] 4.1 The human genome was divided into windows of 5kb, 10kb, 15kb and 20kb and the GC content of each window was counted.
[0089] 4.2 Statistical analysis of the number of sequencing sequences that the BAM files obtained in step 3 fell into each window at different window sizes (5kb, 10kb, 15kb, and 20kb);
[0090] 4.3 Merge files 4.1 and 4.2 to obtain the GC content and DOC for each window;
[0091] 5. Data Correction
[0092] 5.1 For the GC content and DOC obtained in 4.3 for each window, first filter out the regions where the GC content or DOC is 0;
[0093] 5.2 The `ets` function from the R `forecast` package is used to perform exponential smoothing on the DOC results obtained in 5.1 to reduce the volatility of the data distribution across the reference genome; the model parameter is "MAN", and the formula is: DOC ets =ets(DOC,model="MAN");
[0094] 5.3 Using the locally weighted regression model Loess for DOC ets Perform GC correction to obtain the correction weight for each window; multiply the inverse of the correction weight by the DOC. ets Obtain DOC loess To correct sequencing sequence distribution bias caused by high GC;
[0095] 5.4 The DOC files under each window in 5.3 will be... loess Divide by all windows DOC loess The median yields the number of sequences in each window after final correction, using the formula: DOC final =DOC loess / median(DOC loess In this formula, DOC final This represents the final DOC value, median(DOC). loess ) represents the DOC of all windows loess The median value.
[0096] The statistical results of the distribution of corrected DOCs on the genome are shown in Tables 1 and 2. Table 1 shows the statistical data of the distribution of corrected DOCs on the genome with different sequencing data volumes (window bin = 20kb). The results show that the larger the sequencing volume, the smaller the fluctuation in the number of sequences covering the genome. Table 2 compares the statistical data of the distribution of DOCs on the genome with different window sizes after noise reduction under 20M sequencing reads. The results show that the larger the window, the smaller the fluctuation in the number of sequences covering the genome.
[0097] Table 1
[0098] 10M 15M 20M SD 0.38 0.33 0.31 Var 0.15 0.11 0.10 CV 38.72 33.17 31.32 Range 41.97 34.21 31.37
[0099] Table 2
[0100] 5kb 10kb 15kb 20kb SD 0.35 0.33 0.32 0.31 Var 0.12 0.11 0.10 0.10 CV 35.30 33.67 32.77 31.32 Range 41.39 36.69 33.22 31.37
[0101] Comparative Example 1
[0102] This comparative example compares the present invention with the original data, only using the Loess method of the locally weighted regression model to correct the differences. Specifically, steps 1-4 are the same as steps 1-4 in Example 1, while step 5 is slightly different, as the exponential smoothing process is not performed.
[0103] 1. Raw data:
[0104] The number of sequences in each window is obtained by dividing the number of unique alignment sequences in each window by the median of all window counts, as shown in the following formula: DOC rawfinal =DOC / median(DOC). In this formula, DOC rawfinal This represents the final DOC value, while median(DOC) represents the median DOC value across all windows.
[0105] 2. Loess correction:
[0106] (1) Use Loess to perform GC correction on the DOC under each window to obtain the correction weight under each window; multiply the inverse of the correction weight by the DOC under each window to obtain the DOC. loess .
[0107] (2) Divide the sequence count under each window obtained above by the median of all window counts to obtain the final corrected sequence count under each window. The formula is: DOC loessfinal =DOC loess / median(DOC loess In this formula, DOC loessfinal This represents the final DOC value, median(DOC). loess ) represents the DOC of all windows loess The median value.
[0108] Table 3 shows the statistical data on the distribution of DOCs on the genome after correction using only the Loess method with a sequencing volume of 20M and a window size of 20kb. Compared with the correction method of this invention, the distribution of DOCs after correction using only the Loess method is wider (range: 72.80>31.37) and the data fluctuation is greater (SD: 0.49>0.31; Var: 0.24>0.10; CV: 49.47>31.32).
[0109] Table 3
[0110]
[0111]
[0112] The relationship between DOC and GC content was analyzed for the original data, the data corrected using only the Loess method, and the data corrected using the method of this invention. The results are as follows: Figures 2-4 As shown, the DOC of the data corrected by the method of the present invention is less susceptible to changes in GC content, and the DOC is more concentrated in regions with different GC contents.
[0113] The distribution of DOCs on the genome was analyzed for the original data, data corrected only using the Loess method, and data corrected using the method of this invention. The results are as follows: Figures 5-7 As shown, the DOC of the data corrected by the method of the present invention is more concentrated on different chromosomes.
[0114] The distribution of DOCs in the gene rearrangement region of chromosome 14 was analyzed for the original data, the data corrected using only the Loess method, and the data corrected using the method of this invention. The results are as follows: Figures 8-10As shown, the DOC of the data corrected by the method of the present invention is more concentrated in the gene rearrangement region of chromosome 14, and the read deletions in the rearrangement region are more easily identified.
[0115] In summary, this invention creatively applies the exponential smoothing method to data denoising before copy number variation (CNV) detection in the context of high-throughput sequencing, providing a data correction method for CNV detection. This method can effectively reduce the uneven distribution of sequencing sequences on the genome caused by library preparation, sequencing, and other technologies. It is easily extended to other techniques for CNV detection, such as region-captured CNV detection, and can be easily integrated into CNV detection devices, thus possessing high application value.
[0116] The applicant declares that the detailed method of the present invention is illustrated by the above embodiments, but the present invention is not limited to the above detailed method, that is, it does not mean that the present invention must rely on the above detailed method to be implemented. Those skilled in the art should understand that any improvements to the present invention, equivalent substitutions of the raw materials of the product of the present invention, addition of auxiliary components, selection of specific methods, etc., all fall within the protection scope and disclosure scope of the present invention.
Claims
1. A data correction method for copy number variation detection, characterized in that, The method includes the following steps: (1) Obtain the GC content and depth coverage value DOC under each window, and filter out the regions with GC content or DOC of 0; (2) Exponential smoothing: The Holt linear method was used to perform exponential smoothing of the DOC values. Alternatively, DOC can be exponentially smoothed using formula (1) to obtain DOC. ets ; Where ets is a function in the forecast package of the R language; (3) Using the locally weighted regression model Loess for DOC ets Perform GC correction, obtain the correction weight for each window, and multiply the inverse of the correction weight by the DOC. ets Obtain DOC loess ; (4) According to formula (2), DOC loess Divide by the DOC in all windows loess The median value is used to obtain the final corrected DOC for each window, denoted as DOC. final ; median(DOC loess ) indicates DOC under all windows loess The median; The exponential smoothing of DOC values is performed using the Holt linear method with multiplicative error, as shown in the following formula: By specifying the training error of the previous step as a relative error, we can obtain: Where y t This indicates that the predicted value of the time series at time t is determined by the level estimate, trend estimate, and relative error at time t-1. This represents the estimated level of the time series at time t, where α is the smoothing parameter for the level (0 ≤ α ≤ 1); the level equation indicates... It is determined by the horizontal estimate, trend estimate, relative error, and horizontal smoothing parameter at time t-1; b t This represents an estimate of the trend of the time series at time t, where β* is the smoothing parameter for the trend 0 ≤ β* ≤ 1; the trend equation indicates that b t It is determined by the horizontal estimate, trend estimate, relative error, horizontal smoothing parameter, and trend smoothing parameter at time t-1.
2. The data correction method for copy number variation detection according to claim 1, characterized in that, Step (1) also includes the steps of aligning sequences and dividing windows.
3. The data correction method for copy number variation detection according to claim 2, characterized in that, The alignment sequence includes obtaining sequencing data of the genome of the sample to be tested, aligning it with a reference genome, sorting it according to chromosomes, and removing duplicates from the alignment results.
4. The data correction method for copy number variation detection according to claim 3, characterized in that, The partitioning window involves dividing the reference genome into windows according to a given window size, generating the GC content corresponding to each window, and counting the number of sequences under each window, which is recorded as the depth coverage value (DOC) under each window.
5. The data correction method for copy number variation detection according to any one of claims 1-4, characterized in that, The method includes the following steps: (1) Obtain sequencing data of the genome of the sample to be tested, compare it with the reference genome, sort it according to chromosomes, and remove duplicates from the comparison results; (2) Divide the reference genome into windows according to the given window size, generate the GC content corresponding to each window, and count the number of sequences under each window, which is recorded as the depth coverage value (DOC) under each window; (3) Filter out regions where the GC content or DOC is 0; (4) Use the Holt linear method to perform exponential smoothing of DOC values; Alternatively, DOC can be exponentially smoothed using formula (1) to obtain DOC. ets ; (5) Using the locally weighted regression model Loess for DOC ets Perform GC correction, obtain the correction weight for each window, and multiply the inverse of the correction weight by the DOC. ets Obtain DOC loess ; (6) According to formula (2), DOC loess Divide by the DOC in all windows loess The median value is used to obtain the final corrected DOC for each window, denoted as DOC. final .
6. A method for detecting copy number variations for purposes other than disease diagnosis and / or treatment, characterized in that, The method includes: The data correction method for copy number variation detection according to any one of claims 1-5 is used to correct the sequencing data of the sample genome and to determine the copy number variation.
7. The copy number variation detection method for non-disease diagnosis and / or treatment purposes according to claim 6, characterized in that, The method includes the following steps: (1') Obtain sequencing data of the genome of the sample to be tested, compare it with the reference genome, sort it according to chromosomes, and remove duplicates from the comparison results; (2') Divide the reference genome into windows according to the given window size, generate the GC content corresponding to each window, and count the number of sequences under each window, which is recorded as the depth coverage value (DOC) under each window; (3') Filter out regions where the GC content or DOC is 0; (4') Perform exponential smoothing of the DOC value using the Holt linear method; or, perform exponential smoothing of the DOC using the formula (1) to obtain the DOC value. ets ; (5') Using the locally weighted regression model loess for DOC ets Perform GC correction, obtain the correction weight for each window, and multiply the inverse of the correction weight by the DOC. ets Obtain DOC loess ; (6') DOC according to the formula (2) loess Divide by the DOC in all windows loess The median value is used to obtain the final corrected DOC for each window, denoted as DOC. final ; (7') DOC final Perform data sharding, analyze the differences in the sharded data, and report copy number variation information.
8. A copy number variation detection device, characterized in that, The copy number variation detection device includes: Sequence alignment module: Used to acquire sequencing data of the genome of the sample to be tested and compare it with a reference genome; Window partitioning module: connected to the sequence alignment module, used to partition the reference genome into windows according to a given window size, generate the GC content corresponding to each window, and count the number of sequences in each window; Data correction module: connected to the window division module, used to perform the steps in the data correction method for copy number variation detection as described in any one of claims 1-5; Copy number mutation determination module: connected to the data correction module, used to determine copy number mutation.
9. The copy number variation detection device according to claim 8, characterized in that, The sequence alignment module is used to perform the following steps: (1'') Obtain sequencing data of the genome of the sample to be tested, compare it with the reference genome, and sort it according to chromosomes; (2'') Remove duplicates from the comparison results.
10. The copy number variation detection device according to claim 9, characterized in that, The window partitioning module is used to perform the following steps: The reference genome is divided into windows according to the given window size, the GC content corresponding to each window is generated, and the number of sequences under each window is counted and recorded as the depth coverage value (DOC) under each window.
11. The copy number variation detection device according to claim 10, characterized in that, The copy number mutation determination module is used to perform the following steps: The DOCs under each window obtained by the data correction module are fragmented, and the fragmented data are used for difference analysis to report copy number variation information.
12. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, The computer program executes the steps of the data correction method for copy number variation detection as described in any one of claims 1-5 or the copy number variation detection method for non-disease diagnosis and / or treatment purposes as described in claim 6 or 7.
13. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer program executes the steps of the data correction method for copy number variation detection as described in any one of claims 1-5 or the copy number variation detection method for non-disease diagnosis and / or treatment purposes as described in claim 6 or 7.