Structural variation merging method based on third-generation sequencing data and considering sequence similarity and related device

Abstract

The invention discloses a structure variation merging method based on third-generation sequencing data and considering sequence similarity and a related device. The method comprises the following steps: acquiring single-sample variation data and multi-sample variation data; dividing the single-sample variation data or the multi-sample variation data into different windows by adopting coarse clustering based on a sliding window; fine clustering based on sequence similarity is carried out in each window; combining the fine clustering results; and integrating the combination results to obtain a consensus sequence. According to the method, on the basis of variation length and position information, variation sequence similarity is considered, more similar structure variations can be found, and higher merging precision is achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of genome structural variation detection, and specifically relates to a structural variation merging method and related devices based on third-generation sequencing data and considering sequence similarity. Background Art

[0002] Structural variation refers to genome rearrangements longer than 50 bp, usually including deletions, insertions, inversions, duplications and translocations. Structural variation is closely related to each of us, mainly reflected in human diseases, chromosome evolution and gene regulation. With the development of sequencing technology, especially the third-generation sequencing technology, which has the characteristics of ultra-long read length, high throughput and high accuracy, it has shown great advantages in many research fields, and human research on structural variation has gradually deepened. There are many structural variation detection tools based on third-generation sequencing data, such as Sniffles published in 2018, SVIM published in 2019, and cuteSV published in 2020. However, among these structural variation detection tools, even the tools with the best comprehensive performance have a large number of false positive detection results in some high-complexity regions of the genome. In addition, these tools detect single samples and cannot directly obtain structural variation results on a population scale. To this end, structural variation merging is needed to process the results after detection in order to obtain a more accurate variation set and meet the needs of population structural variation research.

[0003] At present, there are some tools for merging structural variations. These tools mainly follow two different ideas when merging structural variations. The first idea is to directly use the position information of structural variations on the reference genome for merging, that is, by analyzing the alignment position and length of two structural variations on the genome, calculating the degree of overlap between them, and using this as the basis for merging. SURVIVOR is a representative tool that adopts this merging method. The second idea is to convert structural variations into graphs, and then use graph clustering methods to achieve merging, such as the merging tool Jasmine published in 2023. Both of these ideas essentially only use the position information of the variation, which cannot fully describe the characteristics of structural variations and limit the performance of merging. For example, SURVIVOR is prone to mistakenly merge different insertion and deletion variations at the same position, and the merging results are relatively rough. Although Jasmine uses a more detailed merging method, it often produces misjudgments in high-repetitive regions of the genome, affecting the accuracy of downstream analysis. Therefore, it is not enough to merge based only on the position of structural variations, and more information needs to be considered as the basis for merging structural variations. Summary of the invention

[0004] In order to overcome the technical problem of inaccurate results of genomic structural variation detection in the prior art, the purpose of the present invention is to provide a structural variation merging method and related devices based on third-generation sequencing data and considering sequence similarity. The present invention can not only effectively reduce the false positive rate of the detection results in a single-sample single-detection set scenario and improve the overall quality of the detection set, but also achieve better merging effects in a multi-sample scenario. At the same time, in a single-sample scenario, the present invention also has the ability to identify some complex structural variations.

[0005] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0006] The first aspect of the present invention provides a method for merging structural variations based on third-generation sequencing data and considering sequence similarity, comprising the following steps:

[0007] Obtain single sample variation data and multi-sample variation data;

[0008] Divide single sample variation data or multi-sample variation data into different windows using coarse clustering based on sliding windows;

[0009] Fine clustering based on sequence similarity was performed within each window;

[0010] Merge the fine clustering results;

[0011] The merged results were integrated to obtain a consensus sequence.

[0012] Furthermore, for single-sample variation data, fine clustering based on sequence similarity is performed in each window, including the following steps: for the sequences in each window, the variation sequences are first split into k-mer sets, and then the sequence feature matrix is ​​constructed in a high-dimensional space by counting the k-mer frequencies. Finally, the structural variation sequence similarity in the same window is calculated based on the cosine distance between the feature matrices.

[0013] Furthermore, for multi-sample variation data, fine clustering based on sequence similarity is performed in each window, including the following steps: encoding the k-mer in quaternary, then using the code and frequency as key-value pairs, using a hash table to store the k-mer frequency and performing KLD distance calculation to obtain a distance matrix, and based on the distance matrix, using a multidimensional scaling transformation to map the variation to the Euclidean space while maintaining the relative distance.

[0014] Furthermore, for single sample variation data, the fine clustering results are merged, including the following steps:

[0015] According to the results of fine clustering, based on the distance matrix between structural variations within the window, a bottom-up hierarchical clustering method is used to iteratively cluster and update the matrix until the set distance threshold is reached.

[0016] Furthermore, for multi-sample variation data, the fine clustering results are merged, including the following steps:

[0017] According to the fine clustering results, the KD tree optimized density clustering is used for merging.

[0018] Further, for single sample variant data, the merging results are integrated to obtain a consensus sequence, including the following steps: based on the merging results, a hierarchical clustering tree structure is used to guide the path of double sequence alignment to perform multiple sequence alignment and generate a consensus sequence.

[0019] Furthermore, for multi-sample variant data, the merged results are integrated to obtain a consensus sequence, including the following steps:

[0020] Based on the merge results, a high-dimensional matrix is ​​used to record the scores of the multiple sequence alignments, a partial alignment graph is constructed, and the maximum likelihood path is solved to generate a consensus sequence.

[0021] The second aspect of the present invention provides a structural variation merging system based on third-generation sequencing data and considering sequence similarity, comprising:

[0022] A data acquisition module, used to acquire single-sample variation data and multi-sample variation data;

[0023] A coarse clustering module is used to divide single-sample variation data or multi-sample variation data into different windows using coarse clustering based on sliding windows;

[0024] A fine clustering module, used to perform fine clustering based on sequence similarity within each window;

[0025] A merging module is used to merge the fine clustering results;

[0026] The integration module is used to integrate the merging results to obtain the consensus sequence.

[0027] The third aspect of the present invention provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the structural variation merging method based on third-generation sequencing data and considering sequence similarity when executing the computer program.

[0028] The fourth aspect of the present invention provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the structural variation merging method based on third-generation sequencing data and considering sequence similarity.

[0029] Compared with the prior art, the present invention has the following beneficial effects:

[0030] (1) The present invention considers the sequence similarity of variants in addition to the length and position information of the variants, and can discover more similar structural variants with higher merging accuracy;

[0031] (2) In the single-sample single-test scenario, potential complex structural variations can be discovered by merging different types of variations, providing a certain reference for subsequent related research;

[0032] (3) The merging ability is stable, and the improvement of the performance of a single detection set is effective under different supporting reads, and it still has good merging ability in complex regions of the genome;

[0033] (4) Even with the addition of sequence similarity calculation, it still has high efficiency;

[0034] (5) In multi-sample scenarios, it has stronger merging capabilities than mainstream merging tools.

[0035] The present invention can provide researchers with a more effective way to merge structural variations in single-sample and multi-sample scenarios, thereby obtaining higher-quality variation sets at individual and group scales. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 The overall framework of the structural variation merging algorithm proposed in the present invention;

[0037] Figure 2 Schematic diagram of the sliding window coarse clustering method;

[0038] Figure 3 Schematic diagram of the similarity calculation method based on k-mer frequency;

[0039] Figure 4 Schematic diagram of the optimization method for sequence feature representation of quaternary coded hash;

[0040] Figure 5 It is a schematic diagram of the mapping between KLD distance and MDS Euclidean space;

[0041] Figure 6 Schematic diagram of the AGNES hierarchical clustering merging method; (a) is the distance matrix, and (b) is the merging process;

[0042] Figure 7 Schematic diagram of the density clustering merging method optimized for KD tree;

[0043] Figure 8 A schematic diagram of the length splicing algorithm based on support reads;

[0044] Fig. 9 Schematic diagram of the consensus sequence generation algorithm based on tree alignment;

[0045] Fig.10 Schematic diagram of consensus sequence generation algorithm based on partial order alignment;

[0046] Fig.11 The figures are comparisons between the detection results of the method of the present invention and the results of existing detection tools; wherein (a) is SVIM, (b) is cuteSV, and (c) is Sniffles2;

[0047] Fig.12 The figure is a comparison diagram between the results of merging using the method of the present invention and the results of existing merging tools.

[0048] Fig.13 It is a schematic diagram of the structural variation merging system based on third-generation sequencing data and considering sequence similarity of the present invention. DETAILED DESCRIPTION

[0049] In order to facilitate the understanding of the present invention, the present invention will be described more fully below with reference to the relevant drawings. The preferred embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thoroughly and comprehensively understood.

[0050] See also Figure 1 , a structural variation merging method SeqSV based on third-generation sequencing data and considering sequence similarity, is divided into two merging modes: single sample and multi-sample, namely single sample structural variation merging algorithm SeqSV-S and multi-sample structural variation merging algorithm SeqSV-M. The two have similar processes, both of which include four main steps: data preprocessing, coarse clustering, fine clustering, and merging information integration. However, different implementation methods are used in some steps according to the characteristics of single sample and multi-sample scenarios, mainly including the following steps:

[0051] In the single-sample structural variation merging algorithm SeqSV-S, coarse clustering based on sliding windows is first used to divide the variation into different windows. Each window is independent of each other in the subsequent merging process, thereby reducing the granularity of the merging task and improving the concurrency of the program. In addition, the process also allows different types of variation to exist in the same window, thereby supporting the discovery of potential complex structural variations. Subsequently, within each window, the algorithm performs fine clustering based on sequence similarity, converts the sequence distance into a high-dimensional sparse matrix of k-mer frequency as a feature representation, uses cosine distance to obtain the distance matrix between variations, and uses AGNES hierarchical clustering based on the distance matrix to merge variations. In the information integration stage, SeqSV-S uses different basic information integration methods for variation types. For the integration of sequence information, SeqSV-S uses the tree structure of hierarchical clustering to guide the path of double sequence alignment to perform multiple sequence alignments, and generates a consensus sequence as the merged sequence. Experiments on real data sets have shown that for current mainstream structural variation detection tools, SeqSV-S can not only effectively reduce the false positive rate of a single detection set and significantly improve the overall performance of the detection results, but also obtain higher quality detection results by merging multiple detection sets from different tools.

[0052] The multi-sample structural variation merging algorithm SeqSV-M uses a variety of optimization methods to address the larger data scale in population scale scenarios, including: using quaternary encoded hash tables to replace the high-dimensional sparse matrix in the single-sample structural variation merging algorithm, optimizing the k-mer extraction and storage process, and improving the calculation efficiency of the cosine distance; proposing the KLD distance to replace the variation distance measurement method that only considers sequence similarity in a single sample, to meet the variation merging needs when the sequence is default in multi-sample scenarios, and mapping the variation to the Euclidean space through multi-dimensional scaling transformation; using the DBSCAN density clustering optimized by the KD tree instead of the hierarchical clustering, to improve the merging efficiency in multi-sample scenarios while adapting to the irregular distribution of similar variations in the Euclidean space.

[0053] Example 1

[0054] The structural variation merging method SeqSV based on third-generation sequencing data and considering sequence similarity of the present invention comprises the following steps:

[0055] (1) Sliding window-based coarse clustering method: In both single-sample and multi-sample modes, sliding windows are used to divide the chromosomes and positions where structural variations are located with a fixed step size (default 1kbp), and each window is used as the basic unit for merging. In the single-sample single-test set scenario, the window does not distinguish between variant types to support the discovery of potential complex structural variations, such as Figure 2 As shown;

[0056] (2) Sequence similarity calculation based on k-mer frequency: In the single sample mode, for each sequence in the window, the variant sequence is first split into a k-mer set, and then the sequence feature matrix is ​​constructed in a high-dimensional space by counting the k-mer frequencies (see Figure 3 ), and finally, the structural variation sequence similarity under the same window is calculated based on the cosine distance between the feature matrices (the calculation process is a well-known technology in the art), such as Figure 3 As shown: Taking K=3 as an example, the dimension of the matrix is ​​also 3. Each base in the 3-mer corresponds to its coordinate in the corresponding dimension. Then each 3-mer will be mapped to the specified position of the feature matrix, and the final value of the matrix is ​​obtained by statistical frequency.

[0057] (3) Optimization of sequence feature representation of quaternary coded hash: In view of the larger data scale in multi-sample scenarios, the k-mer is first encoded according to the quaternary system. Then, the code and frequency are used as key-value pairs, and the k-mer frequency is stored in a hash table and the distance is calculated to optimize memory usage and computing efficiency. The process is as follows: Figure 4 As shown;

[0058] (4) KLD distance and MDS Euclidean space mapping: In the multi-sample model, KLD distance is proposed as a measure of the distance between variants, taking into account the length of the variant and the sequence similarity, and multidimensional scaling (MDS) is used to map the variant to the Euclidean space while maintaining the relative distance. Figure 5 As shown in the figure, sv1 to sv4 are used as examples for explanation, where Figure (a) is the distance matrix obtained by calculating the KLD distance of four SV sequences. Based on this matrix, MDS can be used to map Sv to the Euclidean space (Figure (b)), and try to make the Euclidean distance between each SV consistent with that in the distance matrix;

[0059] (5) AGNES hierarchical clustering-based merging method: In the single-sample mode, based on the distance matrix between structural variations within the window, a bottom-up hierarchical clustering method is used to iteratively cluster and update the matrix until the specified distance threshold is reached. The process is as follows: Figure 6 As shown: Based on the distance matrix of SV1 to SV4, each SV is a cluster at the beginning, and the two SVs with the shortest distance are merged, and the distance between the merged cluster and other clusters is updated using the full chain method until the shortest distance is greater than the threshold;

[0060] (6) KD tree optimized DBSCAN density clustering merging method: see Figure 7 , use KD tree optimized density clustering to merge in multi-sample mode, while improving the efficiency of merging, adapting to the situation where similar variants are mapped in Euclidean space with irregular distribution shapes;

[0061] (7) Basic information integration: In single sample merging, variants of different types before merging are marked as potential complex structural variants, and the length of insertion and deletion variants is calculated using a length concatenation algorithm based on supporting reads. In other scenarios, the average method is used for calculation. Figure 8 As shown in the figure, the process of using the length splicing algorithm based on support reads to calculate the length of insertion and deletion variants is as follows: maintain a hash table for each cluster, record all the supporting reads of the structural variants before merging and the splicing length corresponding to each read. When traversing the structural variants in the cluster, the unique identifier of all supporting reads of the variant and the length of the variant are put into the hash table as the key and value in sequence. If the current read identifier already exists, the corresponding value is modified to the sum of the original value and the current variant length, that is, length splicing. Finally, sort the values ​​corresponding to all reads, take the upper quartile as the length after merging, and the position after merging also takes the upper quartile of all positions.

[0062] (8) Consensus sequence generation based on tree alignment: see Fig. 9 , using the tree alignment algorithm Tree Align, using the clustering tree in the single sample merging process to guide the path of pairwise sequence alignment, gradually generating multiple sequence alignment results, and generating a consensus sequence as the merged sequence;

[0063] (9) Consensus sequence generation based on partial alignment: see Fig.10 ,In the multi-sample mode, a high-dimensional matrix is ​​used to record the scores of the multiple ,sequence alignment, a partial alignment graph is constructed and the maximum likelihood ,path is solved to generate the consensus sequence, as shown in Table 1.

[0064] Table 1 Results before and after the merger

[0065]

[0066] Experiments on the HG002 family data show that in a single-sample, single-dataset scenario, SeqSV-S can improve the F1 value of the results of current mainstream detection tools by merging (as shown in Table 2 and Fig.11 In the single-sample multi-dataset scenario, the high-quality results obtained are better than any single test set in F1 (as shown in Table 2). In the multi-sample scenario, compared with the two most mainstream structural variation merging tools, the number of de-nova mutations after merging by the multi-sample structural variation merging algorithm SeqSv-M is 4699, which is more accurate than the current mainstream merging tools Jasmine (5746) and SURVIVOR (6377) (see Fig.12 ).

[0067] Table 2 Comparison of the results of the present invention and existing detection tools

[0068]

[0069] The method of the present invention adds sequence similarity calculation based on k-mer frequency on the basis of structural variation position and length information, which can not only effectively reduce the false positive rate of the detection results in the single sample single detection set scenario and improve the overall quality of the detection set, but also achieve better merging effect in the multi-sample scenario. At the same time, in the single sample scenario, the method also has the ability to identify some complex structural variations.

[0070] See also Fig.13 , another embodiment of the present invention provides a structural variation merging system based on third-generation sequencing data and considering sequence similarity, comprising: a data acquisition module, a coarse clustering module, a fine clustering module, a merging module and an integration module;

[0071] A data acquisition module, used to acquire single-sample variation data and multi-sample variation data;

[0072] A coarse clustering module is used to divide single-sample variation data or multi-sample variation data into different windows using coarse clustering based on sliding windows;

[0073] A fine clustering module, used to perform fine clustering based on sequence similarity within each window;

[0074] A merging module is used to merge the fine clustering results;

[0075] The integration module is used to integrate the merging results to obtain the consensus sequence.

[0076] Another embodiment of the present invention provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the structural variation merging method based on third-generation sequencing data and considering sequence similarity when executing the computer program.

[0077] Another embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for merging structural variations based on third-generation sequencing data and considering sequence similarity is implemented.

[0078] The above description is only for the best embodiment of the present invention, but it should not be understood as limiting the claims. The present invention is not limited to the above embodiments, and its specific structure is allowed to be changed. However, all changes made within the protection scope of the independent claims of the present invention are within the protection scope of the present invention.

[0079] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more related listed items.

Claims

1. A method for merging structural variations based on third-generation sequencing data and considering sequence similarity, characterized in that: The following steps are involved: Obtain single sample variation data and multi-sample variation data; Divide single sample variation data or multi-sample variation data into different windows using coarse clustering based on sliding windows; Fine clustering based on sequence similarity was performed within each window; Merge the fine clustering results; The merged results were integrated to obtain a consensus sequence.

2. The method for merging structural variations based on third-generation sequencing data and considering sequence similarity according to claim 1, characterized in that: For single-sample variation data, fine clustering based on sequence similarity is performed in each window, including the following steps: for the sequences in each window, the variation sequences are first split into k-mer sets, and then the sequence feature matrix is ​​constructed in a high-dimensional space by counting the k-mer frequencies. Finally, the structural variation sequence similarity in the same window is calculated based on the cosine distance between the feature matrices.

3. The method for merging structural variations based on third-generation sequencing data and considering sequence similarity according to claim 1, characterized in that: For multi-sample variation data, fine clustering based on sequence similarity is performed in each window, including the following steps: encoding k-mers in quaternary form, then using the code and frequency as key-value pairs, using a hash table to store the k-mer frequencies and performing KLD distance calculations to obtain a distance matrix. Based on the distance matrix, multidimensional scaling is used to map the variation to Euclidean space while maintaining relative distances.

4. The method for merging structural variations based on third-generation sequencing data and considering sequence similarity according to claim 1, characterized in that: For single-sample variation data, the fine clustering results are merged, including the following steps: According to the results of fine clustering, based on the distance matrix between structural variations within the window, a bottom-up hierarchical clustering method is used to iteratively cluster and update the matrix until the set distance threshold is reached.

5. The method for merging structural variations based on third-generation sequencing data and considering sequence similarity according to claim 1, characterized in that: For multi-sample variation data, the fine clustering results are merged, including the following steps: According to the fine clustering results, the KD tree optimized density clustering is used for merging.

6. The method for merging structural variations based on third-generation sequencing data and considering sequence similarity according to claim 1, characterized in that: For single-sample variant data, the merging results are integrated to obtain a consensus sequence, including the following steps: Based on the merging results, a hierarchical clustering tree structure is used to guide the path of double sequence alignment to perform multiple sequence alignment and generate a consensus sequence.

7. The method for merging structural variations based on third-generation sequencing data and considering sequence similarity according to claim 1, characterized in that: For multi-sample variant data, the merge results are integrated to obtain a consensus sequence, including the following steps: Based on the merge results, a high-dimensional matrix is ​​used to record the scores of the multiple sequence alignments, a partial alignment graph is constructed, and the maximum likelihood path is solved to generate a consensus sequence.

8. A structural variation merging system based on third-generation sequencing data and considering sequence similarity, characterized in that: include: A data acquisition module, used to acquire single-sample variation data and multi-sample variation data; A coarse clustering module is used to divide single-sample variation data or multi-sample variation data into different windows using coarse clustering based on sliding windows; A fine clustering module, used to perform fine clustering based on sequence similarity within each window; A merging module is used to merge the fine clustering results; The integration module is used to integrate the merging results to obtain the consensus sequence.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method for merging structural variations based on third-generation sequencing data and considering sequence similarity as described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for merging structural variations based on third-generation sequencing data and considering sequence similarity as described in any one of claims 1 to 7 is implemented.

Images