CN118335203B - Coronavirus recombination detection method, system, equipment and medium for large-scale genome data - Google Patents

Abstract

The invention relates to the technical field of bioinformatics, and discloses a coronavirus recombination detection method, system, device and medium for large-scale genome data. The invention applies data mining related methods, combines bioinformatics tools, and establishes a coronavirus genome sequence fragment library and a query library by processing the coronavirus genome sequence; for the sequence of an unknown coronavirus, the genome source determination and recombination event determination of the coronavirus sequence are realized through mechanisms such as fragment matching and iterative voting; through query comparison, the parent sequences corresponding to different genome sources and their corresponding recombinant fragments are determined, while retaining the rich information of the large-scale massive genome data of the coronavirus, the library construction data is effectively used, the recombination detection and analysis tasks of the coronavirus sequence to be detected are completed efficiently and accurately, the recombination event detection for the large-scale coronavirus genome data is realized, and important data support is provided for subsequent research and application.

Description

Coronavirus recombination detection method, system, equipment and medium for large-scale genomic data

Technical Field

The present invention relates to the field of bioinformatics technology, and in particular to a coronavirus recombination detection method, system, equipment and medium for large-scale genome data.

Background Art

Coronavirus is a positive-sense single-stranded RNA virus that mainly infects the human respiratory tract, causing symptoms such as fever, cough, sore throat, and headache. SARS virus, MERS virus, and new coronavirus are all coronaviruses. Genetic recombination occurs during the genetic evolution of pathogens. Genetic recombination is a series of genomic fragments produced during viral replication in a host organism infected with multiple viruses at the same time. The genomic fragments of a certain virus are inserted or replaced into the genomic sequence of another virus, which will trigger the recombination between closely related genes of different viruses, and then cause a large change in the viral genome sequence. A major genetic evolution feature of coronaviruses is that there are frequent and complex recombination evolution events between the viruses within. For example, the dominant lineage of MERS virus produced by the genetic recombination of camel coronavirus is closely related to the MERS epidemic in South Korea in 2015. In addition, the outbreak of many coronaviruses, including SARS virus, is closely related to genetic recombination evolution. Based on the important impact of coronavirus genetic recombination, studying the evolutionary laws of coronavirus genetic recombination is of great significance to the pathogen monitoring and epidemic prevention and control of coronavirus.

The existing mainstream recombination detection technology generally uses models to perform recombination detection on the results of phylogenetic analysis of pathogenic microorganisms. This type of recombination detection technology has great limitations in the field of coronaviruses. On the one hand, traditional recombination analysis methods are only for small-scale genetic sequence data. In the context of the rapid development of next-generation sequencing technology, traditional methods are difficult to support the genetic analysis of rapidly growing large-scale genomic data, especially the rich data information in the massive coronavirus genome represented by the new coronavirus with tens of millions of sequence data. On the other hand, traditional recombination analysis methods only study the possible recombination events between genomic sequences within the input range, and cannot detect the recombination events of the input sequence relative to the entire coronavirus background. In addition, traditional recombination detection methods require the input sequence to complete sequence alignment, and most of the phylogenetic analysis is required during the recombination detection process, which is also difficult to achieve in large-scale coronavirus genome data.

Therefore, there is an urgent need for a coronavirus recombination detection method for large-scale genomic data to achieve coronavirus recombination detection efficiently and accurately.

Summary of the invention

The present invention provides a coronavirus recombination detection method, system, device and medium for large-scale genome data, so as to solve the defect that traditional recombination analysis methods are difficult to support recombination detection of massive coronavirus genome data.

The present invention provides a method for constructing a multi-dimensional reorganization detection corpus, comprising:

Obtaining coronavirus genome data, wherein the coronavirus genome data includes multiple genome sequence data;

performing quality screening on the multiple genome sequence data according to the illegal character rate of each genome sequence data in the multiple genome sequence data to obtain multiple high-quality genome sequence data;

Clustering multiple high-quality genome sequence data to obtain multiple genome sequence clusters, wherein each genome sequence cluster in the multiple genome sequence clusters is labeled with a classification label, and each high-quality genome sequence data in each genome sequence cluster is labeled with a biological category label, and the biological category label includes a main label and a grouping number;

According to the sequence length of each high-quality genome sequence data in the multiple high-quality genome sequence data, the multiple high-quality genome sequence data are screened, and according to the number of high-quality genome sequence data that meet the sequence length requirements in each genome sequence cluster, the multiple genome sequence clusters are screened to obtain multiple genome sequence clusters to be selected for library construction and the multiple high-quality genome sequence data contained therein;

According to a plurality of preset segmentation lengths, each high-quality genome sequence data in the plurality of high-quality genome sequence data to be selected for library construction is segmented into sequence fragments to obtain a plurality of base subsequence data corresponding to each preset segmentation length in the plurality of preset segmentation lengths, a coronavirus genome sequence fragment library is constructed for the plurality of base subsequence data corresponding to each preset segmentation length, and a source label is annotated for each base subsequence data in the plurality of base subsequence data according to the biological category label of each high-quality genome sequence data in the plurality of high-quality genome sequence data to be selected for library construction, wherein the plurality of preset segmentation lengths include a first preset segmentation length and a second preset segmentation length;

According to the primary tag of each high-quality genome sequence data in the multiple high-quality genome sequence data, the multiple high-quality genome sequence data are grouped to obtain multiple primary tag groups, and each primary tag group in the multiple primary tag groups is independently built to form a primary query library, wherein each primary tag group includes multiple high-quality genome sequence data with the same primary tag;

Building a library for multiple high-quality genome sequence data to be selected for library building to obtain a secondary query library;

A database is built for multiple high-quality genome sequence data after quality screening and clustering to obtain a three-level query library.

In one embodiment, the method of performing quality screening on the multiple genome sequence data according to the illegal character rate of each genome sequence data in the multiple genome sequence data to obtain multiple high-quality genome sequence data includes:

According to the number of nucleotide characters and the number of non-nucleotide characters of each genome sequence data, the illegal character rate of each genome sequence data is obtained;

The genome sequence data whose illegal character rate is less than or equal to a preset illegal character rate threshold is determined as high-quality genome sequence data, so as to screen out a plurality of high-quality genome sequence data from the plurality of genome sequence data.

In one embodiment, the method comprises screening multiple high-quality genome sequence data according to the sequence length of each high-quality genome sequence data in the multiple high-quality genome sequence data, and screening multiple genome sequence clusters according to the number of high-quality genome sequence data that meet the sequence length requirements in each genome sequence cluster, to obtain multiple genome sequence clusters to be selected for library construction and multiple high-quality genome sequence data contained therein, including:

According to the sequence length of each high-quality genome sequence data in the multiple high-quality genome sequence data, the high-quality genome sequence data whose sequence length is greater than or equal to a preset base number threshold is determined as the high-quality genome sequence data that meets the sequence length requirement;

According to the number of high-quality genome sequence data meeting the sequence length requirement in each genome sequence clustering cluster, the genome sequence clustering clusters containing the number of high-quality genome sequence data meeting the sequence length requirement greater than or equal to a preset number threshold are determined as genome sequence clustering clusters to be selected for library construction, and the high-quality genome sequence data contained in the genome sequence clustering clusters to be selected for library construction are determined as high-quality genome sequence data to be selected for library construction, so as to obtain multiple genome sequence clustering clusters to be selected for library construction and the multiple high-quality genome sequence data contained in them.

The present invention also provides a coronavirus recombination detection method for large-scale genome data, comprising:

Receive coronavirus genome data to be tested;

According to the coronavirus genome data to be detected, a priori knowledge query is performed using the secondary query library obtained by the construction method of the multi-dimensional recombinant detection corpus described in any of the above items;

According to the coronavirus genome data to be detected, a coronavirus genome sequence fragment library obtained by the construction method of the multi-dimensional recombinant detection corpus described in any of the above items is used to perform fragment matching, and according to the source label of the matched base subsequence data and the queried prior knowledge, the source label of the coronavirus genome data to be detected is obtained and the corresponding segment in the coronavirus genome data to be detected is stained, wherein the source label of the coronavirus genome data to be detected includes a main label and/or a special label, and the special label indicates that the source is unknown;

Judging whether there is a recombination event in the coronavirus genome data to be detected according to the number of main tags and special tags in the source tags of the coronavirus genome data to be detected;

When there is a recombination event in the coronavirus genome data to be detected, the base fragments of all the stained segments of each corresponding main tag in the coronavirus genome data to be detected are connected in series according to the chronological position in the coronavirus genome data to be detected to obtain a query subsequence, and the primary query library, the secondary query library, and the tertiary query library obtained by the construction method of the multi-dimensional recombination detection corpus described in any of the above items are used to perform a parent sequence query on the query subsequence to obtain the parent sequence data of the coronavirus genome data to be detected;

Subsequence each stained segment corresponding to the main label in the coronavirus genome data to be detected is extracted to obtain subsequence data, and the subsequence data is located in the parent sequence data using a sequence alignment method to obtain the recombinant fragment site data of the coronavirus genome data to be detected.

In one embodiment, the prior knowledge query is performed based on the coronavirus genome data to be detected using the secondary query library obtained by the construction method of the multi-dimensional recombinant detection corpus described in any one of the above items, including:

The similarity between the coronavirus genome data to be tested and the high-quality genome sequence data in the secondary query library is calculated to obtain the high-quality genome sequence data with the highest similarity to the coronavirus genome data to be tested and its biological category label as the prior knowledge of the coronavirus genome data to be tested.

In one embodiment, according to the coronavirus genome data to be detected, the coronavirus genome sequence fragment library obtained by the construction method of the multi-dimensional recombinant detection corpus described in any one of the above items is used to perform fragment matching, and according to the source label of the matched base subsequence data and the queried prior knowledge, the source label of the coronavirus genome data to be detected is obtained and the corresponding segment in the coronavirus genome data to be detected is stained, including:

According to a first preset segmentation length, the coronavirus genome data to be detected is segmented into sequence fragments to obtain a plurality of first sequence fragment data to be detected, each first sequence fragment data to be detected in the plurality of first sequence fragment data to be detected is matched with high-quality genome sequence data of a coronavirus genome sequence fragment library corresponding to the first preset segmentation length, and a biological category label of the high-quality genome sequence data obtained by matching each first sequence fragment data to be detected is used as a matching label corresponding to the first sequence fragment data to be detected, and voting for a plurality of matching labels according to the matching results of the plurality of first sequence fragment data to be detected;

When there is a unique matching tag with the most votes among multiple matching tags, the unique matching tag with the most votes is used as the optimal tag, and the optimal tag is used as the source tag to label and color the base sites corresponding to the first sequence fragment data to be detected that has voted for the optimal tag;

When there are multiple matching labels with the most votes among multiple matching labels, if the biological category label in the prior knowledge is one of the multiple matching labels with the most votes, the biological category label in the prior knowledge is the optimal label, otherwise, one of the multiple matching labels with the most votes is randomly selected as the optimal label, and the optimal label is used as the source label to label and color the base sites corresponding to the first sequence fragment data to be detected that has voted for the optimal label;

When the matching results of multiple first sequence fragment data to be detected show that none of the multiple first sequence fragment data to be detected have matching labels, segment the coronavirus genome data to be detected according to the second preset segmentation length to obtain multiple second sequence fragment data to be detected, match each second sequence fragment data to be detected in the multiple second sequence fragment data with the high-quality genome sequence data of the coronavirus genome sequence fragment library corresponding to the preset second segmentation length, use the biological category label of the high-quality genome sequence data obtained by matching each second sequence fragment data to be detected as the matching label corresponding to the second sequence fragment data to be detected, vote for the multiple matching labels according to the matching results of the multiple second sequence fragment data to be detected, and if the matching results of the multiple second sequence fragment data to be detected still show that none of the multiple second sequence fragment data to be detected have matching labels, mark the multiple second sequence fragment data to be detected with special labels;

For the stained segments in the coronavirus genome data to be detected, filtering out the stained segments whose base length is less than or equal to a preset stained length threshold;

For the unstained segments in the coronavirus genome data to be tested, if the base length of the unstained segment is less than or equal to the preset staining length threshold, the same staining is performed according to the context staining of the unstained segment. If the base length of the unstained segment is greater than the preset staining length threshold, a new round of fragment matching and matching label voting is started.

In one embodiment, judging whether there is a recombination event in the coronavirus genome data to be detected according to the number of main tags and special tags in the source tags of the coronavirus genome data to be detected includes:

In the case where there is no special tag in the source tag of the coronavirus genome data to be detected and only a main tag exists, when there is only one main tag in the source tag of the coronavirus genome data to be detected, it is determined that there is no recombination event in the coronavirus genome data to be detected; when the source tag of the coronavirus genome data to be detected includes multiple main tags, it is determined that there is a recombination event in the coronavirus genome data to be detected;

When the source tags of the coronavirus genome data to be detected contain both the special tag and the main tag, it is determined that the coronavirus genome data to be detected has a recombination event;

When only special tags exist in the source tags of the coronavirus genome data to be detected, it is determined that there is no recombination event in the coronavirus genome data to be detected.

In one embodiment, the base fragments of all the staining segments corresponding to the main tag in the coronavirus genome data to be detected are concatenated according to the chronological position in the coronavirus genome data to be detected to obtain a query subsequence, and the primary query library, the secondary query library, and the tertiary query library obtained by the construction method of the multidimensional recombinant detection corpus described in any one of the above items are used to perform a parent sequence query on the query subsequence to obtain the parent sequence data of the coronavirus genome data to be detected, including:

The query subsequence is queried for a parent sequence using the primary query library. When the parent sequence data of the coronavirus genome data to be detected is obtained, the step is terminated. Otherwise, the query subsequence is queried for a parent sequence again using the secondary query library. When the parent sequence data of the coronavirus genome data to be detected is obtained, the step is terminated. Otherwise, the query subsequence is queried for a parent sequence again using the tertiary query library. If the parent sequence data of the coronavirus genome data to be detected is still not obtained, the unknown parent data is used as the parent sequence data of the coronavirus genome data to be detected.

The present invention also provides a coronavirus recombination detection system for large-scale genome data, comprising:

A data receiving module, used to: receive coronavirus genome data to be detected;

A priori knowledge query module is used to perform prior knowledge query based on the coronavirus genome data to be detected using a secondary query library obtained by any of the above-mentioned methods for constructing a multi-dimensional recombinant detection corpus;

A sequence fragment matching module is used to: perform fragment matching based on the coronavirus genome data to be detected, using the coronavirus genome sequence fragment library obtained by the construction method of the multi-dimensional recombinant detection corpus described in any of the above items, and obtain the source label of the coronavirus genome data to be detected based on the source label of the matched base subsequence data and the queried prior knowledge, and color the corresponding segment in the coronavirus genome data to be detected, wherein the source label of the coronavirus genome data to be detected includes a main label and/or a special label, and the special label indicates that the source is unknown;

A recombination event determination module is used to determine whether there is a recombination event in the coronavirus genome data to be detected according to the number of main tags and special tags in the source tags of the coronavirus genome data to be detected;

A parent sequence query module, used for: when there is a recombination event in the coronavirus genome data to be detected, the base fragments of all the staining segments corresponding to the main tag in the coronavirus genome data to be detected are connected in series according to the chronological position in the coronavirus genome data to be detected, to obtain a query subsequence, and the primary query library, the secondary query library, and the tertiary query library obtained by the construction method of the multi-dimensional recombination detection corpus described in any of the above items are used to perform a parent sequence query on the query subsequence to obtain the parent sequence data of the coronavirus genome data to be detected;

The recombinant fragment site positioning module is used to: perform subsequence interception on each staining segment corresponding to the main tag in the coronavirus genome data to be detected to obtain subsequence data, and use the sequence alignment method to locate the site of the subsequence data in the parent sequence data to obtain the recombinant fragment site data of the coronavirus genome data to be detected.

The present invention also provides an electronic device, comprising a processor and a memory storing a computer program, wherein when the processor executes the computer program, the method for constructing a multi-dimensional recombination detection corpus and/or the coronavirus recombination detection method for large-scale genomic data as described above is implemented.

The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-mentioned methods for constructing a multidimensional recombination detection corpus and/or a coronavirus recombination detection method for large-scale genomic data.

The present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute any of the above-mentioned methods for constructing a multidimensional recombination detection corpus and/or the coronavirus recombination detection method for large-scale genomic data.

The present invention provides a coronavirus recombination detection method, system, device and medium for large-scale genome data, which has at least the following advantages:

1. The present invention applies data mining related methods, combines bioinformatics tools, and establishes a coronavirus genome sequence fragment library and a query library by processing the coronavirus genome sequence; for the sequence of unknown coronavirus, the genome source determination and recombination event determination of the coronavirus sequence are realized through mechanisms such as fragment matching and iterative voting; through query comparison, the parental sequences corresponding to different genome sources and their corresponding recombinant fragments are determined, which can realize recombination event detection for large-scale coronavirus genome data and provide important data support for subsequent research and application.

2. The present invention avoids the limitation of traditional recombination analysis methods that they are only applicable to small-scale genetic sequence data. Based on quality control, cluster annotation, pre-operation library construction and other means, while retaining the rich information of large-scale massive genome data of coronavirus, the present invention effectively uses the library construction data to efficiently and accurately complete the recombination detection and analysis tasks of the coronavirus sequences to be detected.

3. The present invention avoids the limitation of traditional recombination analysis methods that only study possible recombination events between genome sequences within a specific range and cannot detect recombination events of the coronavirus sequence to be detected relative to the entire coronavirus background. First, a multi-dimensional corpus is constructed using a rich and complete range of coronavirus genome data to effectively support the detection and evaluation of all possible recombination events and the sources of the recombinant sequences of the coronavirus sequence to be detected in the entire coronavirus background, and the data analysis results are more objective and reliable.

4. The present invention does not require the user to process the coronavirus sequence to be detected before detection. The user only needs to provide the coronavirus sequence to be detected, which makes it easier and more convenient for the user to use.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

Figure 1 is one of the flow charts of a coronavirus recombination detection method for large-scale genomic data provided by the present invention; Figure 2 is a second flow chart of a coronavirus recombination detection method for large-scale genomic data provided by the present invention; Figure 3 is a structural schematic diagram of a coronavirus recombination detection system for large-scale genomic data provided by the present invention; Figure 4 is a structural schematic diagram of an electronic device provided by the present invention.

DETAILED DESCRIPTION

In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings in the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments, and they should not be understood as limitations on the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. In the description of the present invention, it should be understood that the terms used are only for descriptive purposes and cannot be understood as indicating or implying relative importance.

The following describes the coronavirus recombination detection method, system, device and medium for large-scale genome data provided by the present invention in conjunction with Figures 1 to 4.

Figures 1 and 2 are schematic flow diagrams of a coronavirus recombination detection method for large-scale genomic data provided by the present invention. Referring to Figure 1, a coronavirus recombination detection method for large-scale genomic data provided by the present invention may include:

Step S110, constructing a multidimensional recombinant detection corpus, wherein the multidimensional recombinant detection corpus includes a coronavirus genome sequence fragment library (the coronavirus genome sequence fragment library may include two coronavirus genome sequence fragment libraries, one is a long-fragment coronavirus genome sequence fragment library (primary fragment library), and the other is a short-fragment coronavirus genome sequence fragment library (secondary fragment library)), a primary query library, a secondary query library, and a tertiary query library;

Step S120, receiving coronavirus genome data to be detected;

Step S130: Based on the coronavirus genome data to be detected, the secondary query library is used to perform a priori knowledge query. Without any prior knowledge, the primary query library cannot be used (labels are required), and the secondary query library contains all library construction sequences and excludes sequences in the tertiary library that do not meet the library construction quality requirements, so the secondary library is used;

Step S140: According to the coronavirus genome data to be detected, a coronavirus genome sequence fragment library is used to perform fragment matching, and according to the source label of the matched base subsequence data and the queried prior knowledge, the source label of the coronavirus genome data to be detected is obtained and the corresponding segment in the coronavirus genome data to be detected is stained, wherein the source label of the coronavirus genome data to be detected includes a main label and/or a special label, and the special label indicates that the source is unknown;

Step S150, judging whether there is a recombination event in the coronavirus genome data to be detected according to the number of main tags and special tags in the source tags of the coronavirus genome data to be detected;

Step S160: When there is a recombination event in the coronavirus genome data to be detected, the base fragments of all the dyeing segments corresponding to the main tag in the coronavirus genome data to be detected are connected in series according to the chronological positions in the coronavirus genome data to be detected to obtain a query subsequence, and the query subsequence is queried for a parent sequence using the primary query library, the secondary query library, and the tertiary query library to obtain the parent sequence data of the coronavirus genome data to be detected;

Step S170, subsequence each staining segment corresponding to the main tag in the coronavirus genome data to be detected to obtain subsequence data, and use the sequence alignment method to locate the site of the subsequence data in the parent sequence data to obtain the recombinant fragment site data of the coronavirus genome data to be detected.

It should be noted that the executor of the coronavirus recombination detection method for large-scale genomic data provided by the present invention can be any terminal-side device that meets the technical requirements, such as a coronavirus recombination detection device for large-scale genomic data.

It should be noted that step S110 only needs to pre-build a multi-dimensional recombination detection corpus before executing other coronavirus recombination detection steps (steps S120 to S170), and there is no need to rebuild it before each coronavirus recombination detection.

In one embodiment, step S110 may include:

Step S1101, obtaining coronavirus genome data, wherein the coronavirus genome data includes multiple genome sequence data. In this embodiment, step S1101 can obtain the genome sequence data of all coronaviruses from online public databases such as NCBI Genbank, GISAID, or local genome databases.

Step S1102: performing quality screening on the multiple genome sequence data according to the illegal character rate of each genome sequence data in the multiple genome sequence data to obtain multiple high-quality genome sequence data.

In this embodiment, the illegal character rate of each genome sequence data can be obtained based on the number of nucleotide characters and the number of non-nucleotide characters in each genome sequence data, and the genome sequence data with an illegal character rate less than or equal to a preset illegal character rate threshold is determined as high-quality genome sequence data, so as to screen out multiple high-quality genome sequence data from multiple genome sequence data.

Specifically, for a certain genome sequence data (nucleotide sequence), assuming that the total length of the sequence is L , and the total number of characters (illegal characters) other than the four nucleotides ACGT in the current sequence is e , this embodiment requires that the number of illegal characters in the current sequence is small, that is, the illegal character rate E≤0.10 , where the calculation formula of the illegal character rate E is as follows:

,

Sequences that meet the above illegal character rate requirements are considered high-quality genome sequences that meet the requirements and will be used in subsequent steps, and the remaining low-quality genome sequence data will be discarded.

Step S1103: cluster multiple high-quality genome sequence data to obtain multiple genome sequence clusters, wherein each genome sequence cluster in the multiple genome sequence clusters is marked with a classification label, and each high-quality genome sequence data in each genome sequence cluster is marked with a biological category label, and the biological category label includes a main label and a grouping number. In this embodiment, the CD-HIT clustering tool is used to cluster multiple high-quality genome sequence data.

Specifically, since the number of novel coronavirus sequences is much higher than that of other coronavirus sequences, and the internal similarity of novel coronavirus sequences is extremely high, the sequences can be resampled after CD-HIT clustering, while maintaining the diversity and integrity of the sequences as much as possible, while minimizing the data bias of the sequences, thereby constructing a non-redundant data set. The CD-HIT tool is used for nucleotide sequence clustering, which clusters the sequences according to their similarity and divides them into different classification labels (clustering labels). CD-HIT can be used to perform cluster calibration analysis on coronavirus sequences, and the label classification of sequences can be achieved by clustering and annotating the sequence data. After that, each retained high-quality genome sequence data obtains a label that represents its own biological category, such as SARS-rat_1, MERS_2, SARS-CoV-2_2, etc. Taking SARS-rat_1 as an example, the biological category label includes the main label and the grouping number, indicating that the biological category of the high-quality genome sequence is grouping No. 1 in the SARS-rat class.

Step S1104: screening multiple high-quality genome sequence data according to the sequence length of each high-quality genome sequence data in the multiple high-quality genome sequence data, and screening multiple genome sequence clusters according to the number of high-quality genome sequence data that meet the sequence length requirements in each genome sequence cluster, to obtain multiple genome sequence clusters to be selected for library construction and the multiple high-quality genome sequence data contained therein.

Since the recombinant strain itself is the result of the recombination of two or more sequences, its similarity with other sequences is often limited. By checking the label distribution of known recombinant sequences in the data set, it can be learned that the recombinant sequence data often form a CD-HIT label cluster in the form of "one or several full-length recombinant genome sequences (generally more than 20,000 bases) + several shorter recombinant fragments".

Therefore, in this embodiment, according to the sequence length of each high-quality genome sequence data in the multiple high-quality genome sequence data, the high-quality genome sequence data with a sequence length greater than or equal to a preset base number threshold (for example, 3500 bases) can be determined as the high-quality genome sequence data that meets the sequence length requirement, and then according to the number of high-quality genome sequence data that meet the sequence length requirement in each genome sequence clustering cluster, the genome sequence clustering cluster containing the number of high-quality genome sequence data that meet the sequence length requirement greater than or equal to the preset number threshold (for example, 10) can be determined as the genome sequence clustering cluster to be selected for library construction, and the high-quality genome sequence data contained in the genome sequence clustering cluster to be selected for library construction can be determined as the high-quality genome sequence data to be selected for library construction, so as to obtain multiple genome sequence clustering clusters to be selected for library construction and the multiple high-quality genome sequence data contained therein.

Step S1105: According to a plurality of preset segmentation lengths, each of the plurality of high-quality genome sequence data to be selected for library construction is segmented into sequence fragments to obtain a plurality of base subsequence data corresponding to each of the plurality of preset segmentation lengths, a coronavirus genome sequence fragment library is constructed for the plurality of base subsequence data corresponding to each preset segmentation length, and a source label is annotated for each of the plurality of base subsequence data according to the biological category label of each of the plurality of high-quality genome sequence data to be selected for library construction, wherein the plurality of preset segmentation lengths include a first preset segmentation length ( ) and the second preset segmentation length ( ), bp is the base length, the same below.

Specifically, for a label A Assume that its length is , starting from the first base, the sequence is cut into fixed lengths The base subsequence of each segmentation is 1, so a total of base subsequences (there may be duplications, but they do not affect subsequent results and need not be considered); in the sequence fragment library, save the segmented base subsequences (i.e., sequence fragments) and associate each sequence fragment with a tag It indicates that the sequence fragment is derived from the tag The segmentation of the genome sequence in the sequence fragment library; if a sequence fragment (and the fragment and a tag ), then a new association is added so that the sequence fragment is also associated with the tag With label Through the above process, up to 100 sequences in each tag are segmented (for tags with more than 100 sequences, only the first 100 are taken, and the extra sequences are considered redundant and not considered; for tags with less than 100 sequences, all are used), and a sequence fragment library can be constructed, which contains all sequences with a length of Each sequence fragment corresponds to the source of all its tag clusters, that is, the genomic source of the fragment, such as SARS-rat_1, MERS_2, SARS-CoV-2_2, etc. In this embodiment, two sequence fragment libraries are constructed at the same time. The primary sequence fragment library is taken , secondary sequence fragment library .

Step S1106: Group the multiple high-quality genome sequence data according to the primary tag of each high-quality genome sequence data in the multiple high-quality genome sequence data to obtain multiple primary tag groups, and build a separate library for each primary tag group in the multiple primary tag groups to form a primary query library, wherein each primary tag group includes multiple high-quality genome sequence data with the same standard primary tag.

Specifically, for each genome sequence, the corresponding biological category label is obtained based on CD-HIT cluster annotation, such as SARS-rat_1, MERS_2, SARS-CoV-2_2, etc., and the corresponding main labels are SARS-rat, MERS, and SARS-CoV-2. For each main label, all the genome sequences contained in it (regardless of whether they are involved in the construction of the sequence fragment library) can be BLAST-built using the makeblastdb command. A separate library is built for each main label.

Step S1107: construct a library for the multiple high-quality genome sequence data to be selected for library construction (regardless of whether they are involved in the construction of the sequence fragment library). The makeblastdb command can be used to perform BLAST library construction, and all sequences are constructed into one library to obtain a secondary query library.

Step S1108, construct a library for the multiple high-quality genome sequence data after quality screening and clustering. The makeblastdb command can be used to perform BLAST library construction, and all sequences are constructed into one library to obtain a three-level query library.

In one embodiment, step S120 may receive coronavirus genome data (X sequence) to be detected input by a user.

In one embodiment, step S130 can use BLAST (Basic Local Alignment Search Tool) to calculate the similarity between the coronavirus genome data to be detected and the high-quality genome sequence data in the secondary query library to obtain the high-quality genome sequence data with the highest similarity to the coronavirus genome data to be detected and its biological category label as the prior knowledge of the coronavirus genome data to be detected. If the query fails (No-hit), it is considered that there is no prior knowledge.

The X sequence is matched, and the matched fragments are voted according to the genome source in the sequence fragment library, and the genome source with the most votes is determined, and the corresponding genome fragments are determined to be of this source. Other fragments whose sources are not determined continue to be matched in multiple levels until the end. In one embodiment, step S140 may include:

According to the first preset segmentation length , segmenting the coronavirus genome data to be detected into sequence fragments to obtain a plurality of first sequence fragment data to be detected, matching each first sequence fragment data to be detected in the plurality of first sequence fragment data to be detected with high-quality genome sequence data of a coronavirus genome sequence fragment library corresponding to a first preset segmentation length, using a biological category label of the high-quality genome sequence data obtained by matching each first sequence fragment data to be detected as a matching label corresponding to the first sequence fragment data to be detected, and voting for a plurality of matching labels according to the matching results of the plurality of first sequence fragment data to be detected.

Specifically, firstly, the full-length fragments of X are matched and voted. , the input sequence X is also split into fixed length The base subsequence of the segmentation is 1, and the sequence fragments of sequence X are obtained. A voting pool is constructed. The sequence fragments obtained by segmentation are matched in the primary fragment library. The matching fragments have biological category labels in the fragment library, such as SARS-rat_1, MERS_2, SARS-CoV-2_2, etc.; each fragment segmented from the X sequence votes for its corresponding label in the voting pool. If a fragment is associated with multiple labels at the same time, the votes of these labels are increased by one.

After a round of voting is completed, the voting results are counted, among which there are several possibilities as follows.

When there is a unique matching tag with the most votes among multiple matching tags, the unique matching tag with the most votes is used as the optimal tag, and the optimal tag is used as the source tag to label and color the base sites corresponding to the first sequence fragment data to be detected that has voted for the optimal tag;

When there are multiple matching labels with the most votes among multiple matching labels, if the biological category label in the prior knowledge is one of the multiple matching labels with the most votes, the biological category label in the prior knowledge is the optimal label, otherwise, one of the multiple matching labels with the most votes is randomly selected as the optimal label, and the optimal label is used as the source label to label and color the base sites corresponding to the first sequence fragment data to be detected that has voted for the optimal label;

When the matching results of the plurality of first sequence fragment data to be detected show that none of the plurality of first sequence fragment data to be detected has a matching tag, the second preset segmentation length is used. , performing sequence segmentation on the coronavirus genome data to be detected to obtain a plurality of second sequence segment data to be detected, matching each second sequence segment data to be detected in the plurality of second sequence segment data to be detected with the high-quality genome sequence data of the coronavirus genome sequence segment library corresponding to the preset second segmentation length, using the biological category label of the high-quality genome sequence data obtained by matching each second sequence segment data to be detected as the matching label corresponding to the second sequence segment data to be detected, voting for the plurality of matching labels according to the matching results of the plurality of second sequence segment data to be detected, and if the matching results of the plurality of second sequence segment data to be detected still show that none of the plurality of second sequence segment data to be detected has a matching label, marking the plurality of second sequence segment data to be detected with a special label;

For the stained segments in the coronavirus genome data to be detected, filter out the stained segments whose base length is less than or equal to the preset stained length threshold （ Indicates that fragments shorter than this length are considered to be not recombinants but mutations, which is used to eliminate the interference of dense point mutations on sequence matching. ) of the stained segment;

For the unstained segments in the coronavirus genome data to be tested, if the base length of the unstained segment is less than or equal to the preset staining length threshold, the same staining is performed according to the context staining of the unstained segment. If the base length of the unstained segment is greater than the preset staining length threshold, a new round of fragment matching and matching label voting is started.

In addition, if there is no no prior knowledge label in the “prior knowledge search” stage, then after the full-length vote (the first vote), the label will be set as the prior knowledge label for use in subsequent iterative voting.

After step S140 is completed, the iteration returns and the results are summarized. Each base site is dyed. At this time, the length of each dyed segment is checked again from the beginning. If there is an individual dyed segment with a length less than or equal to , then delete the segment and perform the same coloring according to the coloring of the segment context.

Through the above-mentioned mechanisms of fragment matching, iterative voting, primary and secondary matching, etc., each base fragment of the unknown coronavirus genome sequence X input by the user can be dyed (marked) as a CD-HIT tag that characterizes its biological source category, so that the source of the viral genome can be determined.

In one embodiment, step S150 may include:

In the case where there is no special tag in the source tag of the coronavirus genome data to be detected and only a main tag exists, when there is only one main tag in the source tag of the coronavirus genome data to be detected, it is determined that there is no recombination event in the coronavirus genome data to be detected; when the source tag of the coronavirus genome data to be detected includes multiple main tags, it is determined that there is a recombination event in the coronavirus genome data to be detected;

When the source tags of the coronavirus genome data to be detected contain both the special tag and the main tag, it is determined that the coronavirus genome data to be detected has a recombination event;

When only special tags exist in the source tags of the coronavirus genome data to be detected, it is determined that there is no recombination event in the coronavirus genome data to be detected.

It should be noted that due to the high sequence similarity during CD-HIT clustering, the sequence similarity within the same tag, such as SARS-CoV-2_2, is very high, and there are even certain repetitions or redundant sequences; and different tags within the same main tag, such as SARS-CoV-2_1, SARS-CoV-2_2, and SARS-CoV-2_3 within SARS-CoV-2, have relatively high internal similarity due to the same or similar species. Therefore, the present invention focuses more on the recombination evolution events between different main tags (such as between MERS and SARS-CoV-2), and the inconsistency of the labels within the main tags is considered to be a normal genetic variation event. In the previous staining stage, the use of labeled tags is mainly for more accurate matching.

Specifically, at the algorithm level, the algorithm simplifies the labels corresponding to the staining marks of each staining segment for the X sequence that has been completely stained (labeled) in the previous stage, retains only the main label content (such as simplifying SARS-CoV-2_2 to SARS-CoV-2), and re-stains and labels according to the main label. Then the staining of the entire length of the X sequence is examined. If there is only one main label in the entire length of the sequence (such as SARS-CoV-2_1 and SARS-CoV-2_2 are both simplified to SARS-CoV-2), it is considered that there is no recombination; conversely, if there are multiple main labels (SARS-rat, SARS-CoV-2), it is considered that there is recombination. In particular, when the special label "Unknown" exists, there is almost certainly recombination, because it is different from any other main label; unless the entire length of the sequence is "Unknown", but this extreme case is almost impossible to occur in actual operation.

If the algorithm determines that there is no recombination in the full length of the X sequence, the main label determined by the full length of X can be directly output as the result of its genome source, and the current operation is terminated at the same time; if there is recombination, it is necessary to further query the parental sequence and locate the recombinant fragment based on the sequence staining segment division performed by the X sequence and the main label of each staining segment.

In one embodiment, step S160 can use the primary query library to perform a parent sequence query on the query subsequence. When the parent sequence data of the coronavirus genome data to be detected is obtained, the step is terminated. Otherwise, the query subsequence is queried again for the parent sequence using the secondary query library. When the parent sequence data of the coronavirus genome data to be detected is obtained, the step is terminated. Otherwise, the query subsequence is queried again for the parent sequence using the tertiary query library. If the parent sequence data of the coronavirus genome data to be detected is still not obtained, the unknown parent data is used as the parent sequence data of the coronavirus genome data to be detected.

Specifically, for a certain genome source tag (primary tag), all the corresponding chromatin segments are examined, and the base fragments of the corresponding chromatin segments are connected in series according to the order of their positions in the X genome sequence to obtain a "query subsequence". Use multi-level BLAST to query the parent sequence (the specific search method is BLASTn). Since the "query subsequence" corresponds to a primary tag, the BLAST search is first performed in the corresponding primary query library. If successful, the query is terminated directly to determine the parent sequence; if it fails (no-hit), the BLAST search is performed in the secondary query library. If successful, the parent sequence is determined; if the secondary query also fails, the BLAST search is performed in the tertiary query library. If successful, the parent sequence is determined; if the tertiary query also fails, the query fails and the parent sequence is returned as NA (unknown parent). The above operation is performed for each genome source tag (primary tag), and finally the parent sequence corresponding to all genome source primary tags of the X sequence is determined.

In one embodiment, step S170 can extract a subsequence from each staining segment of sequence X, and use a sequence alignment method to locate the corresponding site position of the subsequence fragment (recombinant fragment) in the parent sequence according to its corresponding parent sequence. In particular, for the case where the parent sequence is NA in the parent sequence query (this case is very rare), this part does not make an alignment and directly skips the fragment. The sequence alignment method in the field of bioinformatics can be used for sequence alignment. This method uses the pairwise2.align.localms() function of the pairwise2 package in the Biopython library to implement it. In actual operation, since each staining segment (recombinant fragment) needs to be sequence aligned and site located in the parent sequence, the time overhead is relatively large, so the multiprocessing library in the multi-process parallel library is used to accelerate the sequence alignment part in multi-process parallel.

Finally, for the X sequence determined to be recombined, the recombination determination result is output in csv format, such as "1-11285bp<-SARS-CoV-2|MT873842|235-11519bp" means that the sequence fragment 1-11285bp of this sequence comes from the 235-11519bp position fragment of the MT873842 sequence of SARS-CoV-2 (primary tag), and other recombinant fragments are deduced in the same way, and finally the complete result is output.

The following will provide Example 1 to describe the application of the coronavirus recombination detection method for large-scale genomic data provided by the present invention in mining recombination events in SARS-LIKE viruses (SARS-like viruses).

Example 1 collects and organizes 38 SARS-LIKE virus recombinant sequences through literature collection and RDP software detection, and combines them with 165 non-recombinant SARS-LIKE sequences to form a SARS-LIKE data set of 203 sequences. Example 1 aims to detect recombinant strains and their related recombination events (parental sequence sources, recombinant base sites) in the SARS-LIKE data set through this method.

1. Constructing a multi-dimensional reorganization detection corpus

1.1) Collection and preprocessing of genome sequences

The coronavirus genome sequences were downloaded from online databases such as NCBI Genbank and GISAID, totaling about 13 million, most of which were SARS-CoV-2 sequences, and about 40,000 were non-SARS-CoV-2 sequences. To reduce the bias, the sequences were screened and resampled to reduce the data bias of the sequences while maintaining sequence diversity and integrity. Then, the quality of each genome sequence was examined, and the illegal character rate was calculated. The high-quality sequences are retained. The CD-HIT clustering method is used to perform cluster calibration analysis on the coronavirus sequences, and the sequence label classification is achieved by clustering and annotating the sequence data.

1.2) Construction of viral genome sequence fragment library

Based on the above data preprocessed genome sequence, a viral genome sequence fragment library is constructed.

According to the quality control strategy described in this method, CD-HIT tag clusters and the sequences contained therein were screened, and only CD-HIT tag clusters containing at least ten sequences with a length greater than or equal to 3500 bp were retained. After quality control, 19 CD-HIT main tags were retained, containing 98 tag clusters.

Sequence segmentation and library construction are performed on the retained tag clusters to construct fragment lengths. The primary fragment library and Secondary fragment library.

1.3) Construction of viral genome sequence query library

Use the makeblastdb command to build the BLAST library. One primary query library is constructed for each major tag. A total of 19 primary query libraries are constructed for 19 major tags, plus one secondary query library and one tertiary query library, totaling 21 sequence query libraries.

2. Determination of the genomic origin of the input sequence

2.1) Determination of genome origin

According to the SARS-LIKE data set including 203 sequences used in the test, the sequences in the data set were read one by one. For each sequence, according to the method description, the prior knowledge was first acquired, and then the first-level fragment library was used for fragment matching and iterative voting. If necessary, the second-level fragment library was used for iterative voting according to the method description. After the iteration was returned, the length of each segment was checked and the segments with a length less than or equal to 1 were deleted. The dyed segments are directly dyed the same according to the context dyeing. Finally, for each input sequence, each base site of the sequence is dyed (the dyeing information represents the genome source), that is, the source of the viral genome is determined.

2.2) Determination of recombination events

According to the staining conditions and CD-HIT label information, each sequence in the data set is judged to be recombined or not. For each sequence, the labels corresponding to the staining marks of each staining segment are simplified first, and only the main label content is retained. If there is only one main label in the entire length of the sequence, it is considered that there is no recombination; on the contrary, if there are multiple main labels, it is considered that there is recombination. For sequences that are judged not to be recombined, the current result is directly output, and for recombined sequences, the subsequent parental sequence query and the location of the recombined fragment are continued.

3. Parental sequence query and positioning

3.1) Parent sequence query

For each sequence that has been determined to be recombinant in the SARS-LIKE data set, the following operations are performed: for each genomic source label (primary label) in the sequence, a corresponding query subsequence is constructed, and a multi-level BLAST search is used to query the parent sequence, and finally the parent sequence corresponding to all genomic sources of the sequence is determined.

3.2) Localization of recombinant fragments

For each sequence that has been determined to be recombined in the SARS-LIKE data set, the following operations are performed: for each segment of the sequence, sequence alignment is used to compare the sequence fragment of the segment (recombined fragment) with the corresponding position in the parent sequence to locate the site, and finally determine the source of the recombinant fragment and output the result.

Since the recombination of each sequence was known in advance through literature collection and RDP software detection when constructing the SARS-LIKE data set, the accuracy calculation can be performed based on the real recombination results and the recombination results determined by this method to evaluate the reliability of the results of this method. The accuracy results are shown in Table 1. The accuracy results of this method on the SARS-LIKE data set are all high, which shows the reliability of the detection results of the present invention.

In this application case, the coronavirus recombination detection method for large-scale genomic data provided by the present invention can effectively detect the recombination evolution events existing in SARS-LIKE coronaviruses, with high detection accuracy and relatively reliable results, and can provide important data support for subsequent research applications.

The coronavirus recombination detection method for large-scale genome data provided by the present invention has at least the following advantages: 1. The present invention applies data mining related methods, combines bioinformatics tools, and establishes a coronavirus genome sequence fragment library and a query library by processing the coronavirus genome sequence; for the sequence of unknown coronavirus, the genome source determination and recombination event determination of the coronavirus sequence are realized through mechanisms such as fragment matching and iterative voting; through query comparison, the parent sequences corresponding to different genome sources and their corresponding recombinant fragments are determined, which can realize recombination event detection for large-scale coronavirus genome data and provide important data support for subsequent research and application.

2. The present invention avoids the limitation of traditional recombination analysis methods that they are only applicable to small-scale genetic sequence data. Based on quality control, cluster annotation, pre-operation library construction and other means, while retaining the rich information of large-scale massive genome data of coronavirus, the present invention effectively uses the library construction data to efficiently and accurately complete the recombination detection and analysis tasks of the coronavirus sequences to be detected.

3. The present invention avoids the limitation of traditional recombination analysis methods that only study possible recombination events between genome sequences within a specific range and cannot detect recombination events of the coronavirus sequence to be detected relative to the entire coronavirus background. First, a multi-dimensional corpus is constructed using a rich and complete range of coronavirus genome data to effectively support the detection and evaluation of all possible recombination events and the sources of the recombinant sequences of the coronavirus sequence to be detected in the entire coronavirus background, and the data analysis results are more objective and reliable.

4. The present invention does not require the user to process the coronavirus sequence to be detected before detection. The user only needs to provide the coronavirus sequence to be detected, which makes it easier and more convenient for the user to use.

The construction of the multidimensional recombination detection corpus and the coronavirus recombination detection system for large-scale genomic data provided by the present invention are described below. The construction of the multidimensional recombination detection corpus and the coronavirus recombination detection system for large-scale genomic data described below and the construction method of the multidimensional recombination detection corpus and the coronavirus recombination detection method for large-scale genomic data described above can be referenced to each other.

The present invention provides a system for constructing a multi-dimensional reorganization detection corpus, which may include:

A data acquisition module is used to: acquire coronavirus genome data, wherein the coronavirus genome data includes a plurality of genome sequence data;

A quality screening module is used to: perform quality screening on the multiple genome sequence data according to the illegal character rate of each genome sequence data in the multiple genome sequence data to obtain multiple high-quality genome sequence data;

A clustering module is used to: cluster multiple high-quality genome sequence data to obtain multiple genome sequence clusters, wherein each genome sequence cluster in the multiple genome sequence clusters is marked with a classification label, and each high-quality genome sequence data in each genome sequence cluster is marked with a biological category label, and the biological category label includes a main label and a grouping number;

A length screening module is used to screen multiple high-quality genome sequence data according to the sequence length of each high-quality genome sequence data in the multiple high-quality genome sequence data, and screen multiple genome sequence clusters according to the number of high-quality genome sequence data that meet the sequence length requirements in each genome sequence cluster, to obtain multiple genome sequence clusters to be selected for library construction and multiple high-quality genome sequence data contained therein;

A segmentation module, used to: segment each high-quality genome sequence data in a plurality of high-quality genome sequence data to be selected for library construction into sequence fragments according to a plurality of preset segmentation lengths, obtain a plurality of base subsequence data corresponding to each preset segmentation length in a plurality of preset segmentation lengths, construct a coronavirus genome sequence fragment library for the plurality of base subsequence data corresponding to each preset segmentation length, and annotate a source label for each base subsequence data in the plurality of base subsequence data according to a biological category label of each high-quality genome sequence data in the plurality of high-quality genome sequence data to be selected for library construction, wherein the plurality of preset segmentation lengths include a first preset segmentation length and a second preset segmentation length;

The first library building module is used to: group the multiple high-quality genome sequence data according to the primary label of each high-quality genome sequence data in the multiple high-quality genome sequence data to obtain multiple primary label groups, and independently build a library for each primary label group in the multiple primary label groups to form a primary query library, wherein each primary label group includes multiple high-quality genome sequence data with the same primary label;

The second library building module is used to build a library for a plurality of high-quality genome sequence data to be selected for library building, and obtain a secondary query library;

The third library building module is used to build a library for multiple high-quality genome sequence data after quality screening and clustering to obtain a third-level query library.

Referring to FIG3 , a coronavirus recombination detection system for large-scale genome data provided by the present invention may include:

A data receiving module, used to: receive coronavirus genome data to be detected;

A priori knowledge query module is used to perform prior knowledge query based on the coronavirus genome data to be detected using a secondary query library obtained by any of the above-mentioned methods for constructing a multi-dimensional recombinant detection corpus;

A sequence fragment matching module is used to: perform fragment matching based on the coronavirus genome data to be detected, using the coronavirus genome sequence fragment library obtained by the construction method of the multi-dimensional recombinant detection corpus described in any of the above items, and obtain the source label of the coronavirus genome data to be detected based on the source label of the matched base subsequence data and the queried prior knowledge, and color the corresponding segment in the coronavirus genome data to be detected, wherein the source label of the coronavirus genome data to be detected includes a main label and/or a special label, and the special label indicates that the source is unknown;

A recombination event determination module is used to determine whether there is a recombination event in the coronavirus genome data to be detected according to the number of main tags and special tags in the source tags of the coronavirus genome data to be detected;

A parent sequence query module, used for: when there is a recombination event in the coronavirus genome data to be detected, the base fragments of all the staining segments corresponding to the main tag in the coronavirus genome data to be detected are connected in series according to the chronological position in the coronavirus genome data to be detected, to obtain a query subsequence, and the primary query library, the secondary query library, and the tertiary query library obtained by the construction method of the multi-dimensional recombination detection corpus described in any of the above items are used to perform a parent sequence query on the query subsequence to obtain the parent sequence data of the coronavirus genome data to be detected;

The recombinant fragment site positioning module is used to: perform subsequence interception on each staining segment corresponding to the main tag in the coronavirus genome data to be detected to obtain subsequence data, and use the sequence alignment method to locate the site of the subsequence data in the parent sequence data to obtain the recombinant fragment site data of the coronavirus genome data to be detected.

In one embodiment, the prior knowledge query module may include:

The similarity calculation submodule is used to calculate the similarity between the coronavirus genome data to be detected and the high-quality genome sequence data in the secondary query library, and obtain the high-quality genome sequence data with the highest similarity to the coronavirus genome data to be detected and its biological category label as the prior knowledge of the coronavirus genome data to be detected.

In one embodiment, the sequence fragment matching module may include:

The first molecular segmentation module is used to: segment the coronavirus genome data to be detected into sequence fragments according to a first preset segmentation length to obtain a plurality of first sequence fragment data to be detected, match each first sequence fragment data to be detected in the plurality of first sequence fragment data to be detected with the high-quality genome sequence data of the coronavirus genome sequence fragment library corresponding to the first preset segmentation length, use the biological category label of the high-quality genome sequence data obtained by matching each first sequence fragment data to be detected as the matching label corresponding to the first sequence fragment data to be detected, and vote for the plurality of matching labels according to the matching results of the plurality of first sequence fragment data to be detected;

The first voting submodule is used to: when there is a unique matching tag with the most votes among multiple matching tags, use the unique matching tag with the most votes as the optimal tag, and use the optimal tag as the source tag to label and color the base sites corresponding to the first sequence fragment data to be detected that has voted for the optimal tag;

The second voting submodule is used for: when there are multiple matching tags with the most votes among the multiple matching tags, if the biological category tag in the prior knowledge is one of the multiple matching tags with the most votes, the biological category tag in the prior knowledge is used as the optimal tag; otherwise, one of the multiple matching tags with the most votes is randomly selected as the optimal tag, and the optimal tag is used as the source tag to label and color the base sites corresponding to the first sequence fragment data to be detected that has voted for the optimal tag;

A third voting submodule is used for: when the matching results of multiple first sequence fragment data to be detected show that none of the multiple first sequence fragment data to be detected have matching labels, segmenting the coronavirus genome data to be detected according to a second preset segmentation length to obtain multiple second sequence fragment data to be detected, matching each second sequence fragment data to be detected in the multiple second sequence fragment data with the high-quality genome sequence data of the coronavirus genome sequence fragment library corresponding to the preset second segmentation length, using the biological category label of the high-quality genome sequence data obtained by matching each second sequence fragment data to be detected as the matching label corresponding to the second sequence fragment data to be detected, voting for the multiple matching labels according to the matching results of the multiple second sequence fragment data to be detected, and if the matching results of the multiple second sequence fragment data to be detected still show that none of the multiple second sequence fragment data to be detected have matching labels, marking the multiple second sequence fragment data to be detected with special labels;

The first optimization submodule is used to: for the stained segments in the coronavirus genome data to be detected, filter out the stained segments whose base length is less than or equal to a preset staining length threshold;

The second optimization submodule is used for: for the unstained segment in the coronavirus genome data to be detected, if the base length of the unstained segment is less than or equal to the preset staining length threshold, the same staining is performed according to the context staining of the unstained segment; if the base length of the unstained segment is greater than the preset staining length threshold, a new round of fragment matching and matching label voting is started.

In one embodiment, the recombination event determination module may include:

The first determination submodule is used to: when there is no special tag in the source tag of the coronavirus genome data to be detected and only the main tag exists, when there is only one main tag in the source tag of the coronavirus genome data to be detected, determine that there is no recombination event in the coronavirus genome data to be detected; when the source tag of the coronavirus genome data to be detected includes multiple main tags, determine that there is a recombination event in the coronavirus genome data to be detected;

The second determination submodule is used to: when the source tag of the coronavirus genome data to be detected contains both the special tag and the main tag, determine that the coronavirus genome data to be detected has a recombination event;

The third determination submodule is used to determine that there is no recombination event in the coronavirus genome data to be detected when only special tags exist in the source tags of the coronavirus genome data to be detected.

In one embodiment, the parent sequence query module may include:

The iterative query submodule is used to: use the primary query library to perform a parent sequence query on the query subsequence. When the parent sequence data of the coronavirus genome data to be detected is obtained, the step is terminated; otherwise, the parent sequence query is performed again on the query subsequence using the secondary query library. When the parent sequence data of the coronavirus genome data to be detected is obtained, the step is terminated; otherwise, the parent sequence query is performed again on the query subsequence using the tertiary query library. If the parent sequence data of the coronavirus genome data to be detected is still not obtained, the unknown parent data is used as the parent sequence data of the coronavirus genome data to be detected.

FIG4 illustrates a schematic diagram of the physical structure of an electronic device. As shown in FIG4 , the electronic device may include: a processor 810, a communication interface 820, a memory 830, and a communication bus 840, wherein the processor 810, the communication interface 820, and the memory 830 communicate with each other through the communication bus 840. The processor 810 may call the logic instructions in the memory 830 to execute the method for constructing a multi-dimensional recombination detection corpus provided by the above items and/or the coronavirus recombination detection method for large-scale genomic data.

In addition, the logic instructions in the above-mentioned memory 830 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the method for constructing a multidimensional recombination detection corpus provided by the above methods and/or the coronavirus recombination detection method for large-scale genomic data.

On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the method for constructing a multidimensional recombination detection corpus provided by the above methods and/or the coronavirus recombination detection method for large-scale genomic data.

The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.

Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM/RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims (10)

1. A method for constructing a multi-dimensional reorganization detection corpus, characterized by comprising: Obtaining coronavirus genome data, wherein the coronavirus genome data includes multiple genome sequence data; performing quality screening on the multiple genome sequence data according to the illegal character rate of each genome sequence data in the multiple genome sequence data to obtain multiple high-quality genome sequence data; Clustering multiple high-quality genome sequence data to obtain multiple genome sequence clusters, wherein each genome sequence cluster in the multiple genome sequence clusters is labeled with a classification label, and each high-quality genome sequence data in each genome sequence cluster is labeled with a biological category label, and the biological category label includes a main label and a grouping number; According to the sequence length of each high-quality genome sequence data in the multiple high-quality genome sequence data, the length of the clustered multiple high-quality genome sequence data is screened, and according to the number of high-quality genome sequence data that meet the sequence length requirements in each genome sequence cluster, the multiple genome sequence clusters are screened to obtain multiple genome sequence clusters to be selected for library construction and the multiple high-quality genome sequence data contained therein; According to a plurality of preset segmentation lengths, each high-quality genome sequence data in the plurality of high-quality genome sequence data to be selected for library construction is segmented into sequence fragments to obtain a plurality of base subsequence data corresponding to each preset segmentation length in the plurality of preset segmentation lengths, a coronavirus genome sequence fragment library is constructed for the plurality of base subsequence data corresponding to each preset segmentation length, and a source label is annotated for each base subsequence data in the plurality of base subsequence data according to the biological category label of each high-quality genome sequence data in the plurality of high-quality genome sequence data to be selected for library construction, wherein the plurality of preset segmentation lengths include a first preset segmentation length and a second preset segmentation length; According to the primary tag of each high-quality genome sequence data in the multiple high-quality genome sequence data, the multiple high-quality genome sequence data are grouped to obtain multiple primary tag groups, and each primary tag group in the multiple primary tag groups is independently built to form a primary query library, wherein each primary tag group includes multiple high-quality genome sequence data with the same primary tag; Building a library for multiple high-quality genome sequence data to be selected for library building to obtain a secondary query library; Building a database for multiple high-quality genome sequence data after quality screening and clustering to obtain a three-level query library; Among them, the secondary query library is used to perform prior knowledge query when receiving the coronavirus genome data to be detected, and the primary query library, the secondary query library, and the tertiary query library are used to perform parent sequence query on the query subsequence when there is a recombination event in the coronavirus genome data to be detected. 2. The method for constructing a multidimensional recombination detection corpus according to claim 1 is characterized in that the quality screening of the multiple genome sequence data is performed according to the illegal character rate of each genome sequence data in the multiple genome sequence data to obtain multiple high-quality genome sequence data, including: According to the number of nucleotide characters and the number of non-nucleotide characters of each genome sequence data, the illegal character rate of each genome sequence data is obtained; Determining genome sequence data whose illegal character rate is less than or equal to a preset illegal character rate threshold as high-quality genome sequence data, so as to screen out a plurality of high-quality genome sequence data from the plurality of genome sequence data; The method comprises screening the length of the clustered multiple high-quality genome sequence data according to the sequence length of each high-quality genome sequence data in the multiple high-quality genome sequence data, and screening the multiple genome sequence clusters according to the number of high-quality genome sequence data that meet the sequence length requirements in each genome sequence cluster, so as to obtain multiple genome sequence clusters to be selected for library construction and the multiple high-quality genome sequence data contained therein, including: According to the sequence length of each high-quality genome sequence data in the multiple high-quality genome sequence data, the high-quality genome sequence data whose sequence length is greater than or equal to a preset base number threshold is determined as the high-quality genome sequence data that meets the sequence length requirement; According to the number of high-quality genome sequence data meeting the sequence length requirement in each genome sequence clustering cluster, the genome sequence clustering clusters containing the number of high-quality genome sequence data meeting the sequence length requirement greater than or equal to a preset number threshold are determined as genome sequence clustering clusters to be selected for library construction, and the high-quality genome sequence data contained in the genome sequence clustering clusters to be selected for library construction are determined as high-quality genome sequence data to be selected for library construction, so as to obtain multiple genome sequence clustering clusters to be selected for library construction and the multiple high-quality genome sequence data contained in them. 3. A coronavirus recombination detection method for large-scale genomic data, characterized by comprising: Receive coronavirus genome data to be tested; According to the coronavirus genome data to be detected, a priori knowledge query is performed using the secondary query library obtained by the method for constructing the multi-dimensional recombinant detection corpus described in claim 1 or 2; According to the coronavirus genome data to be detected, the coronavirus genome sequence fragment library obtained by the construction method of the multidimensional recombinant detection corpus according to claim 1 or 2 is used to perform fragment matching, and according to the source label of the matched base subsequence data and the queried prior knowledge, the source label of the coronavirus genome data to be detected is obtained and the corresponding segment in the coronavirus genome data to be detected is stained, wherein the source label of the coronavirus genome data to be detected includes a main label and/or a special label, and the special label indicates that the source is unknown; Judging whether there is a recombination event in the coronavirus genome data to be detected according to the number of main tags and special tags in the source tags of the coronavirus genome data to be detected; When there is a recombination event in the coronavirus genome data to be detected, the base fragments of all the staining segments corresponding to the main tag in the coronavirus genome data to be detected are connected in series according to the chronological position in the coronavirus genome data to be detected to obtain a query subsequence, and the query subsequence is queried for a parent sequence using the primary query library, the secondary query library, and the tertiary query library obtained by the method for constructing a multidimensional recombination detection corpus according to claim 1 or 2 to obtain parent sequence data of the coronavirus genome data to be detected; Subsequence each stained segment corresponding to the main label in the coronavirus genome data to be detected is extracted to obtain subsequence data, and the subsequence data is located in the parent sequence data using a sequence alignment method to obtain the recombinant fragment site data of the coronavirus genome data to be detected. 4. The coronavirus recombination detection method for large-scale genome data according to claim 3 is characterized in that, according to the coronavirus genome data to be detected, the secondary query library obtained by the construction method of the multidimensional recombination detection corpus according to claim 1 or 2 is used to perform prior knowledge query, including: The similarity between the coronavirus genome data to be tested and the high-quality genome sequence data in the secondary query library is calculated to obtain the high-quality genome sequence data with the highest similarity to the coronavirus genome data to be tested and its biological category label as the prior knowledge of the coronavirus genome data to be tested. 5. The coronavirus recombination detection method for large-scale genome data according to claim 4 is characterized in that, according to the coronavirus genome data to be detected, the coronavirus genome sequence fragment library obtained by the construction method of the multidimensional recombination detection corpus according to claim 1 or 2 is used to perform fragment matching, and according to the source label of the matched base subsequence data and the queried prior knowledge, the source label of the coronavirus genome data to be detected is obtained and the corresponding segment in the coronavirus genome data to be detected is stained, comprising: According to a first preset segmentation length, the coronavirus genome data to be detected is segmented into sequence fragments to obtain a plurality of first sequence fragment data to be detected, each first sequence fragment data to be detected in the plurality of first sequence fragment data to be detected is matched with high-quality genome sequence data of a coronavirus genome sequence fragment library corresponding to the first preset segmentation length, and a biological category label of the high-quality genome sequence data obtained by matching each first sequence fragment data to be detected is used as a matching label corresponding to the first sequence fragment data to be detected, and voting for a plurality of matching labels according to the matching results of the plurality of first sequence fragment data to be detected; When there is a unique matching tag with the most votes among multiple matching tags, the unique matching tag with the most votes is used as the optimal tag, and the optimal tag is used as the source tag to label and color the base sites corresponding to the first sequence fragment data to be detected that has voted for the optimal tag; When there are multiple matching labels with the most votes among multiple matching labels, if the biological category label in the prior knowledge is one of the multiple matching labels with the most votes, the biological category label in the prior knowledge is the optimal label, otherwise, one of the multiple matching labels with the most votes is randomly selected as the optimal label, and the optimal label is used as the source label to label and color the base sites corresponding to the first sequence fragment data to be detected that has voted for the optimal label; When the matching results of multiple first sequence fragment data to be detected show that none of the multiple first sequence fragment data to be detected have matching labels, segment the coronavirus genome data to be detected according to the second preset segmentation length to obtain multiple second sequence fragment data to be detected, match each second sequence fragment data to be detected in the multiple second sequence fragment data with the high-quality genome sequence data of the coronavirus genome sequence fragment library corresponding to the preset second segmentation length, use the biological category label of the high-quality genome sequence data obtained by matching each second sequence fragment data to be detected as the matching label corresponding to the second sequence fragment data to be detected, vote for the multiple matching labels according to the matching results of the multiple second sequence fragment data to be detected, and if the matching results of the multiple second sequence fragment data to be detected still show that none of the multiple second sequence fragment data to be detected have matching labels, mark the multiple second sequence fragment data to be detected with special labels, wherein the first preset segmentation length is greater than the preset second segmentation length; For the stained segments in the coronavirus genome data to be detected, filtering out the stained segments whose base length is less than or equal to a preset stained length threshold; For the unstained segments in the coronavirus genome data to be tested, if the base length of the unstained segment is less than or equal to the preset staining length threshold, the same staining is performed according to the context staining of the unstained segment. If the base length of the unstained segment is greater than the preset staining length threshold, a new round of fragment matching and matching label voting is started. 6. The coronavirus recombination detection method for large-scale genome data according to claim 5 is characterized in that the method comprises: judging whether there is a recombination event in the coronavirus genome data to be detected according to the number of main tags and special tags in the source tags of the coronavirus genome data to be detected, comprising: In the case where there is no special tag in the source tag of the coronavirus genome data to be detected and only a main tag exists, when there is only one main tag in the source tag of the coronavirus genome data to be detected, it is determined that there is no recombination event in the coronavirus genome data to be detected; when the source tag of the coronavirus genome data to be detected includes multiple main tags, it is determined that there is a recombination event in the coronavirus genome data to be detected; When the source tags of the coronavirus genome data to be detected contain both the special tag and the main tag, it is determined that the coronavirus genome data to be detected has a recombination event; When only special tags exist in the source tags of the coronavirus genome data to be detected, it is determined that there is no recombination event in the coronavirus genome data to be detected. 7. The coronavirus recombination detection method for large-scale genome data according to any one of claims 3 to 6, characterized in that the base fragments of all the stained segments corresponding to the main tags in the coronavirus genome data to be detected are connected in series according to the chronological positions in the coronavirus genome data to be detected to obtain a query subsequence, and the primary query library, the secondary query library, and the tertiary query library obtained by the construction method of the multidimensional recombination detection corpus according to claim 1 or 2 are used to perform a parent sequence query on the query subsequence to obtain the parent sequence data of the coronavirus genome data to be detected, including: The query subsequence is queried for a parent sequence using the primary query library. When the parent sequence data of the coronavirus genome data to be detected is obtained, the step is terminated. Otherwise, the query subsequence is queried for a parent sequence again using the secondary query library. When the parent sequence data of the coronavirus genome data to be detected is obtained, the step is terminated. Otherwise, the query subsequence is queried for a parent sequence again using the tertiary query library. If the parent sequence data of the coronavirus genome data to be detected is still not obtained, the unknown parent data is used as the parent sequence data of the coronavirus genome data to be detected. 8. A coronavirus recombination detection system for large-scale genomic data, comprising: A data receiving module, used to: receive coronavirus genome data to be detected; A priori knowledge query module, used to: perform prior knowledge query based on the coronavirus genome data to be detected, using the secondary query library obtained by the method for constructing the multi-dimensional recombinant detection corpus described in claim 1 or 2; A sequence fragment matching module is used to: perform fragment matching based on the coronavirus genome data to be detected, using the coronavirus genome sequence fragment library obtained by the construction method of the multi-dimensional recombinant detection corpus according to claim 1 or 2, and obtain the source label of the coronavirus genome data to be detected and color the corresponding segment in the coronavirus genome data to be detected based on the source label of the matched base subsequence data and the queried prior knowledge, wherein the source label of the coronavirus genome data to be detected includes a main label and/or a special label, and the special label indicates that the source is unknown; A recombination event determination module is used to determine whether there is a recombination event in the coronavirus genome data to be detected according to the number of main tags and special tags in the source tags of the coronavirus genome data to be detected; A parent sequence query module, used for: when there is a recombination event in the coronavirus genome data to be detected, concatenating the base fragments of all the staining segments corresponding to the main tag in the coronavirus genome data to be detected according to the chronological position in the coronavirus genome data to be detected, to obtain a query subsequence, and performing a parent sequence query on the query subsequence using the primary query library, the secondary query library, and the tertiary query library obtained by the method for constructing the multidimensional recombination detection corpus according to claim 1 or 2, to obtain parent sequence data of the coronavirus genome data to be detected; The recombinant fragment site positioning module is used to: perform subsequence interception on each staining segment corresponding to the main tag in the coronavirus genome data to be detected to obtain subsequence data, and use the sequence alignment method to locate the site of the subsequence data in the parent sequence data to obtain the recombinant fragment site data of the coronavirus genome data to be detected. 9. An electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for constructing a multidimensional recombination detection corpus as described in claim 1 or 2 and/or the coronavirus recombination detection method for large-scale genomic data as described in any one of claims 3 to 7 is implemented. 10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, it implements the method for constructing a multidimensional recombination detection corpus as described in claim 1 or 2 and/or the coronavirus recombination detection method for large-scale genomic data as described in any one of claims 3 to 7.