Computer apparatus, methods, and computer program products for identifying similar sequences between sequences or repetitive sequences within a sequence

Abstract

The present application discloses computer device, method and computer program product for recognizing similar sequences between sequences or repetitive sequences within sequences. The computer device or computer program product of the present application establishes a permutation matrix of the sequence to be analyzed, fills a score matrix H of the sequence to be analyzed according to the permutation matrix, the value of each element H ij in the score matrix H is the maximum value between the value of the element H i‑1，j‑1 in the upper left corner of the element and the value of the sum of the score of the permutation matrix at the position of the element and 0; then backtracking is performed by selecting the maximum value of the element in the score matrix H, and similar sequences between genes or amino acid sequences or repetitive sequences within sequences are detected and located. The present application can quickly and batch recognize and locate infinite partial repetitive sequence fragments within sequences, can be applied to repetitive analysis of massive biological sequences, and can be applied to detection of fusion genes.

Description

Technical Field

[0001] This invention belongs to the field of bioinformatics, and specifically relates to computer devices, methods and computer program products for identifying similar sequences between sequences or repetitive sequences within a sequence. Background Technology

[0002] With the emergence of a wide variety of biological sequences, more and more repetitive DNA sequences have been discovered. Repetitive sequences have been detected in almost all phyla, and statistically, repetitions are more common in eukaryotes than in prokaryotes. In the past few decades of research, the biological functions of repetitions have been explored, and can be summarized as follows: (1) Many repetitions have been shown to be associated with human genetic diseases. For example, human malformations and syndromes are often associated with repetitions in specific chromosomal regions. (2) Some repetitions have been shown to affect bacterial virulence by acting as the molecular basis of mechanisms for successful colonization and infection of different human individuals. (3) Some specific repetitions are essential genomic elements in lower eukaryotes, and their sites may provide useful markers for the identification and tracking of bacterial strains. (4) Many studies have demonstrated the significant role of repetitions in bacterial adaptation to the environment, and even as an important prerequisite for bacterial stage changes and adaptation. For example, some repetitions play an important role in controlling the surface exposure of active protein domains and antigenic variation. (5) Repetitions also have functional effects on DNA structure. For example, hairpin folds of RNA transcribed from the trinucleotide CTG repeat sequence in the 3'UTR of the myotonic dystrophy protein kinase gene bind to and activate dsRNA-activated protein kinases. (6) Repeats can have a decisive effect on transcriptional activity. For example, the first intron of the gene encoding human epidermal growth factor contains an AC repeat that affects transcriptional activity in vivo and in vitro. In addition, gene expression is turned on or off by repeat variations in promoter regions. (7) Studies have found that pathogens can evade vaccine-induced immune responses by simply altering the repeat composition in the αC protein gene. This is of great significance for developing vaccines targeting the αC protein portion. (8) In eukaryotes, certain types of repetitive DNA extensions are located in the 5' or 3' flanking regions of genes. These repeats may be involved in nucleosome organization, recombination, or regulation of gene expression or gene product activity. (9) Changes in repeat sequences within the coding sequence can lead to changes in proteins. For example, changes in the surface opacity (Opa) protein of bacterial isolates occur through rearrangements in the coding repeat sequence. The composition of the Opa protein is closely related to invasion of human epithelial cells. (10) The importance of repetitions in understanding biological functions also lies in their ability to endow proteins with multiple binding and structural roles. The most common function of repetition sets is protein binding. For example, leucine-rich repetitive sequences (LRRs) are associated with a variety of functions, including signal transduction, transmembrane receptors, DNA repair, cell adhesion, and extracellular matrix proteins. A common function among LRRs is that they form complexes with other proteins. However, the functions of most repetitions are still unknown and await deciphering by biologists.

[0003] Repeats in sequences can be categorized into two types based on direction: forward repeats and reverse repeats. Repeats are often uniformly distributed throughout a sequence. Short repeats appear to primarily participate in the regulation of gene expression; for example, when short motif repeats are located within a gene and are not 3 or 6 nucleotides long, they can significantly influence the coding potential of a given transcript. Slightly longer repeats appear to have other functions. For example, they can be involved in changes in the size of cell wall or membrane-associated proteins, potentially leading to increased or decreased exposure of bacterial surfactant protein domains. Moreover, these longer repeats are candidate regions for determining phylogenetic relationships between species or strains. Furthermore, the longer the repeat unit, the greater the chance of introducing point mutations. Experimental manipulation of repeat length has demonstrated a linear correlation between repeat number and cell adhesion. Therefore, variations in repeat length appear to allow for gradual and fully reversible functional changes, enabling rapid evolutionary adaptation to specific environments. The repeating structure of DNA holds many undiscovered secrets. Systematic studies of repetitive DNA require extensive algorithmic support. The identification and functional analysis of repetitive DNA have become timely topics in experimental research. Identifying perfectly identical repeats is relatively straightforward. However, detecting repeats with some degree of sequence similarity presents a significant challenge. The more difficult part is defining the repetition boundary.

[0004] The repeat-match algorithm in the MUMmer toolkit (related literature: Kurtz S, Phillippy A, Delcher AL, Smoot M, Shumway M, Antonescu C, Salzberg SL. Versatile and open software for comparing large genomes. Genome Biol. 2004;5(2):R12. doi:10.1186 / gb-2004-5-2-r12. Epub 2004Jan30.) is a representative algorithm for identifying all repetitions within a sequence. Its principle is based on a suffix tree data structure, finding all the largest perfect matches within a sequence, i.e., repetitions, by matching character substrings. Unfortunately, this method identifies only perfect repeat sequences (i.e., the repeating sequences are completely identical, with 100% sequence identity), ignoring degenerate repeat sequences (i.e., repeat sequences with sequence or nucleotide differences; see: Kurtz S, Choudhuri JV, Ohlebusch E, Schleiermacher C, Stoye J, Giegerich R. REPuter: the manifold applications of repeat analysis on a genomic scale. Nucleic Acids Res. 2001 Nov 15; 29(22):4633-42. doi:10.1093 / nar / 29.22.4633.). Furthermore, this tool lacks the ability to identify inverted repeat sequences. Another method for displaying all repetitions within a sequence is a dot-matrix sequence alignment map, such as the dotter tool (related literature: Sonnhammer EL, Durbin RA dot-matrix program with dynamic threshold control suited for genomic DNA and protein sequence analysis. Gene. 1995 Dec 29; 167(1-2):GC1-10. doi:10.1016 / 0378-1119(95)00714-8.). The principle is to place the sequence itself in the horizontal and vertical positions of the dot-matrix map, and then perform its own sequence alignment using a set sliding window threshold. Because it considers similarity comparison, this method can present all repetitive sequences, including perfect and degenerate repetitive sequences.However, this method has the drawback of being unable to pinpoint the specific location of repeat sequences (it cannot obtain the boundaries of repeat sequences), thus failing to provide statistical evidence regarding repeat sequences within a sequence and hindering statistical inference studies. Furthermore, the tool lacks the ability to identify inverted repeat sequences. This results in a very narrow applicability of this method for identifying internal repeat sequences, failing to meet the real needs of biologists. Users can only determine the approximate location by visually observing the dot plot, which significantly limits biologists' research on repeats.

[0005] Therefore, an algorithm that can identify all repetitions within a sequence, including forward repetitions, reverse repetitions, short repetitions, long repetitions, perfect repetitions, and degenerate repetitions, should be developed. However, such a tool is not yet available.

[0006] Gene fusion may be a crucial step in creating new genes with novel functions. Fusion genes, formed from two or more genes, not only bypass the harmful steps of lengthy and inefficient site mutations, but also combine distantly related or unrelated functional domains through sequence rearrangement, readily generating new structural features and functions, thus propelling the adaptive evolution of species. Furthermore, gene fusion is an important genomic event in human cancer because its fusion gene products can drive cancer development and are potential prognostic tools or therapeutic targets for anti-cancer treatments. Currently, there is no perfect method for identifying and analyzing gene fusions. Summary of the Invention

[0007] The technical problems to be solved by the present invention are how to identify and locate similar sequence fragments between sequences and / or how to identify and locate repetitive sequences within a sequence and / or how to simultaneously identify and locate all repetitive sequences in a sequence and / or how to identify and locate all repetitive sequences in a sequence in batches and without limitation on the number of sequences and / or how to detect fusion genes.

[0008] To address the aforementioned technical problems, the present invention first provides a computer device, comprising a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to perform the following steps:

[0009] S1. Establish the permutation matrix: Establish the permutation matrix based on the sequence to be analyzed; the sequence to be analyzed is represented by sequence A and sequence B, and sequence A and sequence B are the same, both being the sequence to be analyzed.

[0010] S2. Score Matrix Creation: Establish a score matrix H for sequence A and sequence B. Score matrix H is an n+1 row and m+1 column matrix, where m and n are the lengths of the sequences to be analyzed. Set all elements in the first row and first column of score matrix H to 0. Then, score each element in score matrix H (excluding the first row and first column) based on the permutation matrix, filling in the score for each element in score matrix H (excluding the first row and first column) according to the following rules:

[0011]

[0012] In formula (3):

[0013] H ij This represents the score of the matrix element in the i-th row and j-th column; i is a natural number greater than or equal to 1 and less than or equal to n, and j is a natural number greater than or equal to 1 and less than or equal to m.

[0014] H i-1，j-1 The score of the matrix element in the (i-1)th row and (j-1)th column;

[0015] S3. Backtracking: Perform the following backtracking: S31. Initial backtracking: Start with the element A with the highest score in the score matrix H. ij Let's start by backtracking to element A. i-1j-1 This process is repeated until an element with a score of 0 is encountered, yielding the first backtracking result; S32, Second backtracking: The element B with the highest score outside the path of the first backtracking in the score matrix H. rc Let's start by tracing back to element B. r-1c-1 This process is repeated until an element with a score of 0 is encountered, at which point the second backtracking result is obtained.

[0016] S33, the highest-scoring element C outside the first backtracking path and the second backtracking path of the scoring matrix H. tu Let's start by backtracking to element C. t-1u-1 This process is repeated until an element with a score of 0 is encountered, resulting in the third backtracking result; this process is repeated until the v-th backtracking result is obtained, where v is a natural number greater than or equal to 1.

[0017] S4. Result Output: Based on the backtracking result, obtain the repeating fragment and the position of the repeating fragment on the sequence to be analyzed to obtain the repeating unit inside the sequence.

[0018] A ij A represents the element in the i-th row and j-th column of the score matrix H. i-1j-1 B is the element in the (i-1)th row and (j-1)th column of the score matrix H; rc B represents the element in the r-th row and c-th column of the score matrix H. r-1c-1Let C be the element in the (r-1)th row and (c-1)th column of the score matrix H, where r is a natural number greater than or equal to 1 and less than or equal to n, and c is a natural number greater than or equal to 1 and less than or equal to m; tu C represents the element in the t-th row and u-th column of the score matrix H. t-1u-1 This represents the element in the (t-1)th row and (u-1)th column of the score matrix H, where t is a natural number greater than or equal to 1 and less than or equal to n, and u is a natural number greater than or equal to 1 and less than or equal to m.

[0019] In the aforementioned computer device, the sequence to be analyzed may be a gene sequence or an amino acid sequence; the permutation matrix may be a permutation matrix Q1 or a permutation matrix Q2.

[0020] The permutation matrix Q1 is represented as:

[0021]

[0022] In formula (1):

[0023] a i Let b be the i-th sequence element in sequence A. j Let i be the j-th element in sequence B, where i is a natural number greater than or equal to 1 and less than or equal to n, and j is a natural number greater than or equal to 1 and less than or equal to m.

[0024] x is a i and b j The similarity score is a positive real number.

[0025] y is a i and b j The dissimilarity score is a negative real number.

[0026] The permutation matrix Q2 can be a BLOSUM matrix; the BLOSUM matrix can be a BLOSUM62 matrix.

[0027] In the aforementioned computer device, the units of m and n can be the length of a gene sequence or an amino acid sequence. The filling direction of the scoring matrix H can be from left to right or from top to bottom.

[0028] To address the aforementioned technical problems, the present invention also provides a method for identifying repeating sequences within a sequence, the method comprising the following steps:

[0029] C1) Establishing a permutation matrix: Establishing a permutation matrix based on the sequence to be analyzed; the sequence to be analyzed is represented by sequence A and sequence B, and sequence A and sequence B are the same, both being the sequence to be analyzed;

[0030] C2) Score Matrix Creation: Establish a score matrix H for sequence A and sequence B. Score matrix H is an n+1 row and m+1 column matrix, where m and n are the lengths of the sequences to be analyzed. Set all elements in the first row and first column of score matrix H to 0. Then, score each element in score matrix H (excluding the first row and first column) based on the permutation matrix, filling in the score for each element in score matrix H (excluding the first row and first column) according to the following rules:

[0031]

[0032] In formula (3):

[0033] H ij This represents the score of the matrix element in the i-th row and j-th column; i is a natural number greater than or equal to 1 and less than or equal to n, and j is a natural number greater than or equal to 1 and less than or equal to m.

[0034] H i-1，j-1 The score of the matrix element in the (i-1)th row and (j-1)th column;

[0035] C3) Backtracking: Perform the following backtracking: S31, Initial backtracking: Start from the element A with the highest score in the score matrix H. ij Let's start by backtracking to element A. i-1j-1 This process is repeated until an element with a score of 0 is encountered, yielding the initial backtracking result; S32, Second backtracking: The element B with the highest score outside the path of the initial backtracking in the score matrix H. rc Let's start by tracing back to element B. r-1c-1 This process is repeated until an element with a score of 0 is encountered, resulting in the second backtracking result; S33, the element C with the highest score outside the paths of the first and second backtracking in the score matrix H. tu Let's start by backtracking to element C. t-1u-1 This process is repeated until an element with a score of 0 is encountered, resulting in the third backtracking result; this process is repeated until the v-th backtracking result is obtained, where v is greater than or equal to 1; the backtracking includes at least the first backtracking, meaning that the backtracking can be performed at least once.

[0036] C4) Output: Based on the backtracking results, obtain the repeating segments and their positions on the sequence to be analyzed to obtain the repeating units within the sequence.

[0037] A ij A represents the element in the i-th row and j-th column of the score matrix H. i-1j-1 B is the element in the (i-1)th row and (j-1)th column of the score matrix H; rc B represents the element in the r-th row and c-th column of the score matrix H. r-1c-1Let C be the element in the (r-1)th row and (c-1)th column of the score matrix H, where r is a natural number greater than or equal to 1 and less than or equal to n, and c is a natural number greater than or equal to 1 and less than or equal to m; tu C represents the element in the t-th row and u-th column of the score matrix H. t-1u-1 This represents the element in the (t-1)th row and (u-1)th column of the score matrix H, where t is a natural number greater than or equal to 1 and less than or equal to n, and u is a natural number greater than or equal to 1 and less than or equal to m.

[0038] In the above method, the sequence to be analyzed can be a gene sequence or an amino acid sequence; the permutation matrix can be permutation matrix Q1 or permutation matrix Q2;

[0039] The permutation matrix Q1 can be expressed as:

[0040]

[0041] In formula (1):

[0042] a i Let b be the i-th sequence element in sequence A. j Let i be the j-th element in sequence B, where i is a natural number greater than or equal to 1 and less than or equal to n, and j is a natural number greater than or equal to 1 and less than or equal to m.

[0043] x is a i and b j The similarity score is a positive real number.

[0044] y is a i and b j The dissimilarity score is a negative real number.

[0045] The permutation matrix Q2 can be a BLOSUM matrix, and the BLOSUM matrix can be a BLOSUM62 matrix.

[0046] In the above method, the units of m and n can be the length of a gene sequence or an amino acid sequence. The filling direction of the scoring matrix H can be from left to right or from top to bottom.

[0047] To address the aforementioned technical problems, the present invention also provides a computer program product, which includes a computer program that, when executed by a processor, can implement the methods described above.

[0048] To address the aforementioned technical problems, the present invention also provides a computer-readable storage medium storing a computer program that enables a computer to perform the following steps:

[0049] S1. Establish the permutation matrix: Establish the permutation matrix based on the sequence to be analyzed; the sequence to be analyzed is represented by sequence A and sequence B, and sequence A and sequence B are the same, both being the sequence to be analyzed.

[0050] S2. Score Matrix Creation: Establish a score matrix H for sequence A and sequence B. Score matrix H is an n+1 row and m+1 column matrix, where m and n are the lengths of the sequences to be analyzed. Set all elements in the first row and first column of score matrix H to 0. Then, score each element in score matrix H (excluding the first row and first column) based on the permutation matrix, filling in the score for each element in score matrix H (excluding the first row and first column) according to the following rules:

[0051]

[0052] In formula (3):

[0053] H ij This represents the score of the matrix element in the i-th row and j-th column; i is a natural number greater than or equal to 1 and less than or equal to n, and j is a natural number greater than or equal to 1 and less than or equal to m.

[0054] H i-1，j-1 The score of the matrix element in the (i-1)th row and (j-1)th column;

[0055] S3. Backtracking: Perform the following backtracking: S31. Initial backtracking: Start with the element A with the highest score in the score matrix H. ij Let's start by backtracking to element A. i-1j-1 This process is repeated until an element with a score of 0 is encountered, yielding the first backtracking result; S32, Second backtracking: The element B with the highest score outside the path of the first backtracking in the score matrix H. rc Let's start by tracing back to element B. r-1c-1 This process is repeated until an element with a score of 0 is encountered, resulting in the second backtracking result; S33, the element C with the highest score outside the paths of the first and second backtracking in the score matrix H. tu Let's start by backtracking to element C. t-1u-1 This process is repeated until an element with a score of 0 is encountered, resulting in the third backtracking result; this process is repeated until the v-th backtracking result is obtained, where v is a natural number greater than or equal to 1; the backtracking includes at least the first backtracking, meaning that the backtracking can be performed at least once.

[0056] S4. Result Output: Based on the backtracking result, obtain the repeating fragment and the position of the repeating fragment on the sequence to be analyzed to obtain the repeating unit inside the sequence.

[0057] A ijA represents the element in the i-th row and j-th column of the score matrix H. i-1j-1 B is the element in the (i-1)th row and (j-1)th column of the score matrix H; rc B represents the element in the r-th row and c-th column of the score matrix H. r-1c-1 Let C be the element in the (r-1)th row and (c-1)th column of the score matrix H, where r is a natural number greater than or equal to 1 and less than or equal to n, and c is a natural number greater than or equal to 1 and less than or equal to m; tu C represents the element in the t-th row and u-th column of the score matrix H. t-1u-1 This represents the element in the (t-1)th row and (u-1)th column of the score matrix H, where t is a natural number greater than or equal to 1 and less than or equal to n, and u is a natural number greater than or equal to 1 and less than or equal to m.

[0058] In the aforementioned computer-readable storage medium, the sequence to be analyzed may be a gene sequence or an amino acid sequence; the permutation matrix may be a permutation matrix Q1 or a permutation matrix Q2;

[0059] The permutation matrix Q1 is represented as:

[0060]

[0061] In formula (1):

[0062] a i Let b be the i-th sequence element in sequence A. j Let i be the j-th element in sequence B, where i is a natural number greater than or equal to 1 and less than or equal to n, and j is a natural number greater than or equal to 1 and less than or equal to m.

[0063] x is a i and b j The similarity score is a positive real number.

[0064] y is a i and b j The dissimilarity score is a negative real number.

[0065] The permutation matrix Q2 can be a BLOSUM matrix, and the BLOSUM matrix can be a BLOSUM62 matrix.

[0066] In the aforementioned computer-readable storage medium, the units of m and n can be the length of a gene sequence or an amino acid sequence. The filling direction of the scoring matrix H can be from left to right or from top to bottom.

[0067] To address the aforementioned technical problems, the present invention also provides a method for detecting fusion genes, the method comprising the following steps:

[0068] D1) Detection of repetitive sequences within a gene: Use the method described above to detect repetitive sequences within the gene sequence to be tested;

[0069] D2) Repetitive sequence gene structure analysis: Gene structure analysis of the repetitive sequence was performed using gene structure analysis software to obtain the results of the repetitive sequence gene structure analysis;

[0070] D3) Determine the fusion gene: Based on the gene structure analysis results of the repetitive sequence, determine whether the repetitive sequence is a fusion gene.

[0071] In the above method, the gene structure analysis may include exon-intron structure analysis, functional domain analysis of the amino acid sequences encoded by the gene sequence, and motif analysis. The gene structure analysis software may be CFVisual (version 1.0) software or gene structure analysis software available on the SMART and MEME websites.

[0072] The following applications of the computer device, computer program product, and / or computer-readable storage medium described above also fall within the scope of protection of this invention:

[0073] P1. Application in the preparation of disease detection-related products;

[0074] P2. Applications in the development or preparation of disease-related drug targets;

[0075] P3. Application in the preparation of products for detecting fusion proteins.

[0076] Any of the following applications of the method described above also fall within the scope of protection of this invention:

[0077] P1. Application in the preparation of disease detection-related products;

[0078] P2. Applications in the development or preparation of disease-related drug targets;

[0079] P3. Application in the preparation of products for detecting fusion proteins.

[0080] The computer program products mentioned above can be software products that primarily implement their solutions through computer programs.

[0081] The computer-readable storage medium refers to a carrier for storing data, which may be magnetic tape, disk, floppy disk, optical disk, magneto-optical disk, ROM, PROM, VCD, DVD, hard disk, flash memory, USB flash drive, CF card, SD card, MMC card, SM card, Memory Stick, or xD card, etc.

[0082] Repeated sequences have been detected in almost all phyla, but explanations of their biological functions and mechanisms are scarce. This is mainly due to a lack of suitable tools. While some tools exist for identifying repeated sequences in biological sequences, limitations in algorithms prevent the determination of repeat boundaries and may hinder the identification of certain types of important repeats. This impedes biologists' research into the biological functions of repeated sequences. This invention provides a method to address these problems. The invention establishes a permutation matrix of the sequence to be analyzed, and fills the matrix with the permutation matrix to obtain a score matrix H. Each element H in the score matrix H... ij The value is the top left corner H of the element. i-1，j-1 The maximum value between the sum of the element value and the score of the permutation matrix at that element position and 0 is determined. Then, by backtracking by selecting the maximum value of the element value in the score matrix H, similar sequences or repetitive sequences within a sequence are detected and located between genes or amino acid sequences. Experiments have shown that by detecting repetitive sequences within the coding sequences of C4 genes and model organisms, and comparing with existing analytical methods and tools, the method of this invention can quickly detect and identify all repetitive sequences within a gene sequence, and batch design is very effective for analyzing large-scale gene sequences. At the same time, based on the method of this invention, the pro-region of the PME gene in plants has been found to have multiple repetitive sequences, and a fusion gene of the Arabidopsis aspartic protease family has been discovered. The method of this invention can also be applied to the identification of fusion genes in comparative genomics.

[0083] The method of this invention can identify and locate all types of repetitive sequences within a sequence, including nucleotide or amino acid sequences of genes; it can also identify similar sequence fragments between different sequences, also including nucleotide or amino acid sequences of genes. This method is the first to rapidly identify and locate an unlimited number of locally repetitive sequence fragments within a sequence in a single batch, which has broad application prospects for the repetitive analysis of massive biological sequences. Attached Figure Description

[0084] Figure 1 This is a schematic diagram illustrating the principle of the method for detecting similar sequences between sequences or all repetitive sequences within a sequence according to the present invention.

[0085] Figure 2 The diagrams show the scoring matrix backtracking schematic, dot matrix diagram, and flowchart of the method for detecting all repeats within a sequence based on the method of the present invention. (a) shows the scoring matrix and backtracking schematic of the method of the present invention based on the dynamic programming algorithm. (b) shows the dot matrix diagram and key parameters of the method of the present invention; the boundaries of repeating sequences within a sequence determined by the method of the present invention are represented by orange numbers and rectangles. (c) shows the principle flowchart of the method of the present invention for detecting all repeating sequences within a sequence.

[0086] Figure 3 The following are the results of the analysis of internal repeats within the coding sequence of the maize C4 gene using the method of the present invention, and a comparison with the results of repeat-match and Dotter software. (a) Statistical count of all repeat pairs with different homologies within the sequence identified by the method of the present invention; the vertical axis represents the number of repeat pairs within that homology interval, and the horizontal axis represents the intervals between 0 and 1 homologies divided at 0.05 intervals. (b) Venn diagram of perfect repeat sequence results identified by the method of the present invention and perfect repeat sequence results identified by repeat-match. (c) Comparison of dot plots drawn based on the method of the present invention, the position results of repeat-match, and the dot plot drawn by Dotter; the vertical axis represents the base position of the sequence itself in the vertical direction, and the horizontal axis represents the base position of the sequence itself in the horizontal direction.

[0087] Figure 4 Comparison of the number distribution, structure, and internal repetitive sequence density of different coding regions of Type-1 PME genes in representative plants. (a) Life tree and number distribution of Type-1 PME genes in representative plants. (b) Schematic diagram of the structure of Type-1 PME protein sequence. (c) Comparison of internal repetitive density of coding sequences of pro-region and PME domain of all Type-1 PME genes in representative plants, with the horizontal axis representing the repetitive density index obtained based on the method of this invention.

[0088] Figure 5 The fusion gene AT2G28220 was detected using the method of the present invention. (a) Schematic diagram of the process for detecting the fusion gene using the method of the present invention. (b) Internal repeat lattice map of a large segment within the coding sequence of AT2G28220 drawn using the method of the present invention. (c) Comparison of exon-intron structures of the aspartic protease tandem repeat cluster containing AT2G28220. (d) Comparison of motifs and domains of the aspartic protease tandem repeat cluster containing AT2G28220. (e) Chromosomal localization of the aspartic protease tandem repeat cluster gene containing AT2G28220, with the chromosome diagram represented by rectangles. (f) Evolutionary model of tandem repeats and gene fusion of the AT2G28220 gene in Arabidopsis thaliana. Ancestral gene represents the ancestral gene.

[0089] Figure 6 This diagram illustrates the scoring matrix and backtracking process for identifying positive repeat sequences within a nucleotide sequence using the method of this invention. The parameters in the permutation matrix are x = 5 and y = -4; the identity parameter for the repeat sequence is ≥ 80%, and the length parameter for the repeat nucleotides in the repeat sequence is ≥ 4.

[0090] Figure 7 This diagram illustrates the scoring matrix and backtracking process for identifying positive repeat sequences within a nucleotide sequence using the existing Smith-Waterman algorithm. In the permutation matrix, parameters x = 5 and y = -4; the identity parameter for repeat sequences is ≥80%; the length of repeat nucleotides in the repeat sequence is ≥4; and the gap penalty parameter is -3. Black circles indicate the overlapping starting positions of two backtracking paths.

[0091] Figure 8 This diagram illustrates the scoring matrix and backtracking process for identifying inverted repeat sequences within a nucleotide sequence using the method of this invention. The parameters in the permutation matrix are x = 5 and y = -4; the identity parameter for the repeat sequence is ≥80%, and the length parameter for the repeat nucleotides in the repeat sequence is ≥4.

[0092] Figure 9 This diagram illustrates the scoring matrix and backtracking process for identifying forward repeating amino acid sequences using the method of this invention. The permutation matrix is ​​BLOSUM62; the similarity parameter for repeating sequences is ≥70%, and the length parameter for repeating amino acids is ≥15. Detailed Implementation

[0093] The present invention will now be described in further detail with reference to specific embodiments. The given embodiments are merely illustrative of the invention and not intended to limit its scope. The embodiments provided below can serve as a guide for further improvements by those skilled in the art and do not constitute a limitation on the invention in any way.

[0094] Unless otherwise specified, the experimental methods used in the following examples are conventional methods, performed according to the techniques or conditions described in the literature in this field or according to the product instructions. Unless otherwise specified, the materials and reagents used in the following examples are commercially available.

[0095] Example 1: Establishment of the method for identifying similar sequences between sequences or repetitive sequences within a sequence based on dynamic programming algorithm according to the present invention.

[0096] 1. Algorithm principle and flow of the method of this invention

[0097] The specific principles and process of this invention based on the dynamic programming algorithm are as follows ( Figure 1 ):

[0098] Let the two sequences to be compared be gene sequences A and B. A and B (which may be identical or different) are represented as the horizontal and vertical sequences of a matrix, respectively. Where A = a1a2...a... n B = b1b2...b m,; a ​​represents each nucleotide in sequence A, n represents the length of sequence A (nt), a1 represents the first nucleotide of sequence A, and an represents the nth nucleotide of sequence A; b represents each nucleotide in sequence B, m represents the length of sequence B (nt), b1 represents the first nucleotide of sequence B, and bm represents the mth nucleotide of sequence B. When analyzing repetitive sequences within a specific sequence, A and B are considered the same sequence, in which case n = m.

[0099] 1.1 Establishing the permutation matrix

[0100] Obtain the permutation matrices of the two sequences A and B to be compared.

[0101] The permutation matrix can be represented as:

[0102] In formula (1): s(a j b j ) is a i b j The similarity score of the comparison, a i For the i-th nucleotide (sequence element) in sequence A, b j Let i be the j-th nucleotide (sequence element) in sequence B, where i is a natural number greater than or equal to 1 and less than or equal to n, and j is a natural number greater than or equal to 1 and less than or equal to m.

[0103] x is a i and b j Similarity score, which can be set manually, is usually a positive real number, such as "5"; y is a i and b j The dissimilarity score can be set manually, usually as a negative real number, such as "-4"; for example, the permutation matrix when sequence A is "5'-AGCT-3'" and sequence B is also "5'-AGCT-3'" is shown in Table 1:

[0104] Table 1. Examples of Permutation Matrices

[0105] A 5 -4 -4 -4 G -4 5 -4 -4 C -4 -4 5 -4 T -4 -4 -4 5

[0106] 1.2 Create and initialize the score matrix H for sequence A and sequence B.

[0107] The score matrix H has a size of n+1 rows and m+1 columns. Its first row and first column are initialized as follows:

[0108] H k0 =H 0l =0 (0≤k≤n,0≤l≤m) Equation (2);

[0109] In formula (2):

[0110] Hk0 —The scores of the matrix elements in the first column of the scoring matrix H;

[0111] H 0l —The score of the middle element in the first row of the score matrix H.

[0112] k represents row; l represents column.

[0113] The rows of the score matrix H, excluding the first row, are constructed based on sequence A from the 5' end nucleotide to the 3' end nucleotide. The columns of the score matrix H, excluding the first column, are constructed based on sequence B from the 5' end nucleotide to the 3' end nucleotide.

[0114] Scoring is performed on all rows except the first row and all columns except the first column of the score matrix H, from left to right and from top to bottom, according to the permutation matrix in step 1.1. The scores are then filled into each element of the score matrix H except the first row and all columns, according to the following rules:

[0115]

[0116] In formula (3):

[0117] H ij This represents the score of the element in the i-th row and j-th column of matrix H; i is a natural number greater than or equal to 1 and less than or equal to n, and j is a natural number greater than or equal to 1 and less than or equal to m.

[0118] H i-1，j-1 The score of the element in the (i-1)th row and (j-1)th column of matrix H, that is, the score of element (a i b j The top left element (a) i-1 b j-1 The score;

[0119] s(a i b j ) is a i b j The similarity score of the comparison is obtained from the permutation matrix; a i For the i-th element (e.g., a nucleotide) in sequence A, b j Let i be the j-th element (e.g., a nucleotide) in sequence B, where i is a natural number greater than or equal to 1 and less than or equal to n, and j is a natural number greater than or equal to 1 and less than or equal to m.

[0120] H ij The value is taken as H i-1，j-1 The maximum value among H and 0; if H ij A value of 0 represents a i and b j No similarity found, sequence alignment ends here.

[0121] 1.3 Backtracking

[0122] Starting from the element with the highest score in matrix H (when selecting multiple highest scores, prioritize the matrix element with the larger i and / or j values), backtrack along the path from the source of the score (in this invention, the score can only originate from the top left corner, i.e., the element currently in the i-th row and j-th column, denoted as matrix element (i,j), then the element in the top left corner of the current matrix element is the matrix element in the (i-1)-th row and (j-1)-th column, denoted as matrix element (i-1,j-1)). This process is repeated until a matrix element with a score of 0 is encountered, thus obtaining the repeated segment and its position in the sequence to be compared. The segment with the highest local (not the entire sequence) similarity (repeated segment) is generated in this process. A segment with the second highest similarity can be obtained by backtracking from the highest score position outside the highest similarity backtracking process; that is, after completing the first backtracking, backtracking begins from the highest score element outside the first backtracking region to obtain the second locally similar segment (repeated segment). Similarly, a third locally similar segment (repeated segment) is found. This process is iterated until all similar segments (repeated segments) longer than a certain length are found.

[0123] Specifically, the following backtracking is performed:

[0124] First backtracking: Starting from the element A with the highest score in the score matrix H ij Let's start by backtracking to element A. i-1j-1 This process is repeated until an element with a score of 0 is encountered, at which point the first backtracking result is obtained.

[0125] Second backtracking: The highest-scoring element B outside the path of the initial backtracking from the score matrix H. rc Let's start by tracing back to element B. r-1c-1 This process is repeated until an element with a score of 0 is encountered, at which point the second backtracking result is obtained.

[0126] Third backtracking: The highest-scoring element C outside the paths of the first backtracking and the second backtracking of the score matrix H. tu Let's start by backtracking to element C. t-1u-1 This process is repeated until an element with a score of 0 is encountered, resulting in the third backtracking result; this process is repeated until the v-th backtracking result is obtained, where v is a natural number greater than or equal to 1; the number of backtracking iterations depends on the length of the sequence to be analyzed, and includes at least the first backtracking iteration, meaning that the backtracking iterations are performed at least once.

[0127] Among them, A ij A represents the element in the i-th row and j-th column of the score matrix H. i-1j-1 B is the element in the (i-1)th row and (j-1)th column of the score matrix H. rcB represents the element in the r-th row and c-th column of the score matrix H. r-1c-1 Let C be the element in the (r-1)th row and (c-1)th column of the score matrix H, where r is a natural number greater than or equal to 1 and less than or equal to n, and c is a natural number greater than or equal to 1 and less than or equal to m. tu C represents the element in the t-th row and u-th column of the score matrix H. t-1u-1 This represents the element in the (t-1)th row and (u-1)th column of the score matrix H. t is a natural number greater than or equal to 1 and less than or equal to n, and u is a natural number greater than or equal to 1 and less than or equal to m.

[0128] 2. Based on the method of this invention, find all local repetitive sequences within a specific nucleic acid sequence.

[0129] 2.1 Finding positive repeat sequences within specific nucleic acid sequences based on the method of this invention

[0130] Using the following sequence as sequence M: 5'-GTAGAATTAGATGCTCGAT-3', the method of this invention is used to find all repeating sequences within sequence M, including forward repeating sequences, reverse repeating sequences, short repeats, long repeats, perfect repeats, and degenerate repeating sequences.

[0131] Construct the permutation matrix using the method described in step 1.1. This permutation matrix is ​​represented as:

[0132]

[0133] In equation (1): set the parameters in the permutation matrix as x = 5 and y = -4.

[0134] Using the method in step 1.2, determine the score matrix H. M ( Figure 6 Sequences A and B are both sequences M (nucleotide sequence 5'-GTAGAATTAGATGCTCGAT-3'). The identity parameter for the repetitive sequences is set to ≥80%, and the length of the repetitive nucleotides in the repetitive sequences is set to ≥4. The score matrix H is then processed according to the method described in step 1.3. M Perform backtracking iterations until all local (repetitive fragments) similar to sequence M (less than 19 in length) segments larger than a certain length (repetitive nucleotide length parameter ≥ 4) are found, thus obtaining all repetitive sequences within sequence M:

[0135] Because of the analysis matrix H MThe rows and columns of the sequence are all M, so the element with the highest score is the element in the lower right corner of the matrix (score of 95). The path back to this element is the diagonal of the matrix from the lower right corner to the upper left corner. That is, the entire sequence M itself is the longest perfect repetition (the sequence consistency is 100%, the length of the repeated sequence is the full length of sequence M, 19, and it is not considered a local repetition).

[0136] Then from matrix H M The element with the highest score (20) (i=5, j=11, denoted as element (5,11)) is traced back to the element at the previous position (top left corner) based on the source of the score (i-1, j-1), and this process is repeated until an element with a score of 0 is encountered (i=1, j=7, denoted as (1,7)). This yields the repeating fragment and a short, perfect repeat of the repeating fragment located on the test sequence where the nucleotides at positions 2-5 from the 5' end of the M sequence and positions 8-11 from the 5' end of the M sequence are internally identical, exhibiting 100% sequence identity. Figure 6 (As shown in the red box in the middle), the repeating sequence is TAGA;

[0137] Because of the analysis matrix H M The row and column sequences are both M, symmetric about the diagonal, and passed through the scoring matrix H. M The repetitive sequences within the obtained sequence M are symmetrically distributed, such as... Figure 6 The two black backtracking paths are symmetrically distributed along the diagonal line. The repeating sequence positions within the M sequence obtained by the two symmetrically distributed black backtracking paths are the same, both being nucleotides at positions 2-5 from the 5' end of the M sequence and nucleotides at positions 8-11 from the 5' end of the M sequence (TAGA). The segment with the highest local similarity (100% sequence identity) (repetitive segment) is generated in this process. Figure 6 (As shown in the black backtracking path in the middle);

[0138] The fragment with the second highest similarity (80% sequence identity) can be obtained by backtracking from the highest score (score 16) element position outside the highest similarity backtracking process. That is, after completing the first backtracking, backtracking starts from the highest score element outside the first backtracking region to obtain the second locally similar fragment (a degenerate repeat fragment within the M sequence). The fragment is located at nucleotides TAGAT at positions 8-12 from the 5' end of the M sequence and nucleotides TCGAT at positions 15-19 from the 5' end of the M sequence. Figure 6 (As shown in the blue backtracking path). Similarly, finding the third locally similar segment (a repeating segment with 75% sequence consistency) can be done using matrix H. MStarting from the highest-scoring element (11) in the third element, the backtracking begins. After the second backtracking, the backtracking continues from the highest-scoring element outside the first and second backtracking regions to obtain the third locally similar fragment (an internal repeating fragment within the M sequence). This fragment is located at nucleotides TAGA (positions 2-5 from the 5' end of the M sequence) and TCGA (positions 15-18 from the 5' end of the M sequence). Figure 6 (As shown in the purple backtracking path). This process is iterated until all similar segments with a repeating nucleotide length parameter ≥ 4 are found, i.e., positive repeat sequences.

[0139] 2.2 Finding inverted repeat sequences within specific nucleic acid sequences based on the method of this invention

[0140] To identify inverted repeat sequences, the sequence at one horizontal or vertical position is inverted, while the sequence at another position remains unchanged. For example, when searching for similar inverted repeat sequences in sequences A and B, the original 5' end of sequence A is chosen as the new 3' end, and the original 3' end is chosen as the new 5' end, resulting in a new sequence A with the 5' and 3' ends inverted. N After constructing the permutation matrix using the method in step 1.1, sequence A is obtained. N The score matrix H of B N Then, the backtracking steps in step 1.3 are executed, ultimately identifying A. N The reverse repeating sequence position of B is transformed to the position of sequence A for output. Again, taking sequence M as an example, the method of this invention is demonstrated to find the reverse repeating sequence within sequence M. The original 5' end of sequence M is taken as the new 3' end, and the original 3' end is taken as the new 5' end, resulting in a new sequence M with the 5' and 3' ends reversed. N Obtain sequence M N The score matrix H of M N ( Figure 8 Then, the backtracking steps in step 1.3 are executed, ultimately identifying the M. N The reverse repeating sequence position of M is transformed to the position of sequence M for output. The first backtracking path is as follows: Figure 8 The black arrows indicate that the nucleotides GTAGAATTAGATG at positions 1-13 from the 5' end of sequence M and GTAGAATTAGATG at positions 1-13 from the 5' end of sequence M are inverted repeat sequences within sequence M and their locations. Figure 8 (As shown in the red box).

[0141] 2.3 Obtaining the internal repeatability density file of a sequence based on the method of this invention

[0142] To better reflect the repetition in different regions within a sequence, a repetition density index ρ is defined based on the consistency between the scoring matrix and the dot plot, as shown in the following formula:

[0143]

[0144] In equation (4):

[0145] q—The total number of spatial backtracking paths in the scoring matrix;

[0146] l — the side length of the spatial square region (that is, the length of the sequence region).

[0147] The method of this invention can generate an internal repeatability density file for a sequence. Since sequences exhibit both forward and reverse repeating, the repeatability density file contains three values: forward repeatability density, reverse repeatability density, and total repeatability density. Forward repeatability density is calculated by dividing the total number of spatial backtracking paths in the sequence region under forward repeating conditions by the area of ​​that region. Reverse repeatability density is calculated by dividing the total number of spatial backtracking paths in the sequence region under reverse repeating conditions by the area of ​​that region. Total repeatability density is calculated by summing the total number of spatial backtracking paths under forward and reverse repeating conditions and dividing the sum by the area of ​​that region. This repeatability density index effectively reflects the internal repeatability of a sequence, and by comparing the repeatability density values ​​of different regions, the internal repeatability of different regions can be compared intuitively.

[0148] The method of this invention supports two key parameters: a minimum consistency / similarity parameter for repeated fragment pairs and a minimum repeat length parameter. By setting a threshold for the consistency / similarity parameter, the output of repeats with different degrees of sequence similarity can be controlled; by setting a threshold for the minimum repeat length parameter, the output of repeats with different lengths can be controlled. The repeat position file output obtained by the method of this invention is shown in Table 2. Using the output as shown in Table 2, biologists and other analysts can selectively analyze the repeats. For example, the corresponding repeat sequences can be extracted and submitted to BLAST or NCBI databases for preliminary biological significance studies. Since the executable program is written in C, it takes very little time (0.00000429 seconds) to identify all repeats in a 1000bp sequence (tested on a server with a CPU of 32 Intel(R) Xeon(R) Platinum 8269CY CPU T 3.10Ghz and CentOS Linux release 8.5.2111). Therefore, the method of this invention can quickly and effectively identify all repeating sequences within a sequence. By setting parameters, degenerate repeating sequences can be output. At the same time, statistical analysis can be performed based on the position of the obtained repeating sequences to obtain a visualization of the repeating sequences.

[0149] Table 2. Format of the results for outputting intra-sequence repeat position boundaries by the method of the present invention

[0150]

[0151] Note: Start 1, End 1, Start 2, and End 2 represent the start, end, start, and end positions in the vertical direction, respectively, of the dot matrix. When the analyzed sequence is an amino acid sequence, a Similarity result will be generated between the columns for the number of mismatches in the repeating sequence and the score of the repeating sequence.

[0152] 3. Comparison of sequence similarity analysis results between the method of this invention and the Smith-Waterman algorithm.

[0153] The Smith-Waterman algorithm is used to find all positive repeating sequences within the sequence M: 5'-GTAGAATTAGATGCTCGAT-3' from step 2.1. The parameters in the permutation matrix are set to x = 5 and y = -4. The score matrix H is determined using the same method as in step 2.1. M2 and backtracking path ( Figure 7 The Smith-Waterman algorithm was set to have an identity parameter ≥ 80% and a repeat nucleotide length parameter ≥ 4 for the repeat sequences. The results showed that the algorithm only yielded a perfect repeat sequence with an identity parameter of 100% for the repeat sequence corresponding to one backtracking path. Figure 7 The blue arrows in the middle indicate the repetitive sequences located along the backtracking path. These are sequences that perfectly match themselves and are not considered local repetitions. Furthermore, the second backtracking path largely overlaps. Figure 7 The black circles in the diagram indicate the overlapping starting positions of the two backtracking paths; while in step 2.1, the method of this invention was used to obtain more perfect and degenerate repeats within sequence M. It is evident that the Smith-Waterman algorithm cannot find internal repeating sequences within the sequence itself.

[0154] 4. The method of the present invention for the analysis of repeating sequences within an amino acid sequence.

[0155] Using the following amino acid sequence as sequence N: ARNDCQEGHILKMFPSTWYVEGHILKMFPSTRELK, all positive repeating sequences within amino acid sequence N are searched based on the method of this invention.

[0156] Construct the BLOSUM62 matrix (https: / / www.ncbi.nlm.nih.gov / IEB / ToolBox / C_DOC / lxr / source / data / BLOSUM62), and determine the permutation matrix according to the scoring rules of the BLOSUM62 matrix. Use the method in step 1.2 to determine the score matrix H. N ( Figure 9 ), where sequence A and sequence B are both sequences N. The similarity parameter for the repetitive sequences is set to simility ≥ 70%, and the length of the repetitive amino acids in the repetitive sequences is set to ≥ 15. The score matrix H is then processed according to the method in step 1.3. N Perform backtracking iterations until all similar fragments (repeated fragments) longer than a certain length (repeated amino acid length parameter ≥ 15) are found, thus obtaining all locally repeating sequences within amino acid sequence N:

[0157] Because of the analysis matrix H N The rows and columns of the sequence are all N, so the element with the highest score is also the element in the lower right corner of the matrix (score is 194). The path back to this element is the diagonal of the matrix from the lower right corner to the upper left corner. That is, the sequence N itself is the longest perfect repetition (the sequence consistency is 100%, which means that the sequence itself is a complete match with itself, and is not considered a local repetition).

[0158] Then from matrix H N The element with the highest score (56) (i=21, j=35, denoted as (21,35)) is traced back to the element at the previous position (top left corner) based on the source of the score (i-1, j-1), and this process is repeated until an element with a score of 0 is encountered (i=6, j=20, denoted as (6,20)). This yields the repeating fragment and the sequence of the repeating fragment located on the test sequence as follows: The sequence similarity between the amino acid residues EGHILKMFPSTWYVE (positions 7-21 from the amino terminus) and the sequence of the repeating fragment located on the test sequence as follows: EGHILKMFPSTRELK (positions 21-35 from the amino terminus) is 13 / 15 = 86.7%, and the identity is 11 / 15 = 73.3%. This is a shorter degenerate repeat. Figure 9 (as shown in the red box in the middle);

[0159] Because of the analysis matrix H N The rows and columns are all N, symmetrical about the diagonal, and passed through the scoring matrix H. N The analysis revealed that the repeating sequences within sequence N exhibited a symmetrical distribution, such as... Figure 9The two black backtracking paths are symmetrically distributed with the diagonal as the symmetry line. The positions of the repeating sequences inside the N sequence obtained by the two symmetrically distributed black backtracking paths are the same, namely the amino acid residues EGHILKMFPSTWYVE from the amino terminus of the N sequence from position 7 to 21 and the amino acid residues EGHILKMFPSTRELK from the amino terminus of the N sequence from position 21 to 35.

[0160] Therefore, based on the method of the present invention, it is possible not only to find and detect all repetitive sequences (including forward repetitive sequences, reverse repetitive sequences, short repetitives, long repetitives, perfect repetitives, and degenerate repetitive sequences) within a gene sequence, but also to find and detect all repetitive sequences within other sequences such as amino acid sequences.

[0161] 5. Implementation of the method of the present invention

[0162] To achieve efficient output, this invention uses C, currently the fastest low-level computer language, to implement the algorithm and allows it to be compiled into an executable program. The method of this invention is then coded into a computer program product that runs on the Unix / Linux platform.

[0163] By adjusting different parameter settings, the computer program product of this invention can form different analysis modules to analyze different types of sequences, such as gene nucleotide sequences or protein amino acid sequences. This executable program is called in batches using Python, and downstream analysis tools are then written in Python to analyze the identified repetitive sequences and perform data processing, statistical analysis, and visualization of the results. These downstream analysis tools specifically include: processing the location and number of repetitive fragments, drawing raster plots, drawing depth maps, calculating repetition density, and outputting detailed comparisons of specific repetitive fragments. All of these are processed and output in batches, and users can provide suggestions and optimizations for the development of all the code.

[0164] Example 2: Analysis examples of the method and computer program of the present invention and their comparative analysis with the prior art.

[0165] Taking the identification of internal positive repeat sequences in the coding sequence U08401.1 of the maize C4 gene (data download link: https: / / www.ncbi.nlm.nih.gov / nuccore / U08401.1 / ?report=fasta-) as an example, the analysis results of the method of this invention and the repeat-match method for identifying repeat sequences in the existing MUMmer toolkit are compared.

[0166] To ensure a fair comparison with the numerical results of repeat-match, the consistency parameter of the method in this invention is set to 0.0 (i.e., outputting all positive repeats with low sequence similarity), the minimum output repeat length parameter is set to 6, and the minimum output repeat nucleotide length parameter of repeat-match is also set to 6. Finally, the numerical results are statistically analyzed using Python and plotted graphically for comparison.

[0167] Analysis results of the method of the present invention show that the consistency of repeated segment pairs within the coding sequence of U08401.1 is between 0.45 and 1.0. Figure 3 In section a), the largest number of repeating fragment pairs were in the range of 0.65-0.7, reaching 6353 pairs. The smallest number of repeating fragment pairs were in the range of 0.9-0.95, with only 9 pairs. All repeat-match identification results were completely identical repeating fragment pairs with a consistency of 1.0. Venn diagram analysis showed that the 141 pairs of repeating fragment pairs identified by the method of this invention had a consistency of 1.0. Figure 3 The yellow-green histogram corresponding to (0.95, 1.0) of a, and the corresponding... Figure 3 The green circles in b represent all repeating segments that match the results of repeat-match; while the repeat-match method found 415 pairs of perfect repeating sequences other than 141 perfect repeating pairs. Figure 3 (The brown circle in b represents) The method of the present invention did not find the degenerate repeat sequence fragments, but found approximately 19,272 other pairs of degenerate repeat sequence fragments. This indicates that the principle of the method of the present invention based on the dynamic programming algorithm sacrifices some perfect repeats in order to find degenerate repeats. That is, among the other 415 pairs of perfect repeat sequences found by the repeat-match method, the method of the present invention found 19,272 pairs of degenerate repeat sequences. For example, some adjacent short perfect repeat sequences detected by repeat-match are regarded as a degenerate repeat sequence that exists as a whole by the method of the present invention.

[0168] The method of this invention and the results of repeat-match are plotted as dot plots, and compared with the dot plot results of Dotter, another method for displaying all repetitions within a sequence (dot plot sequence alignment). Figure 3 (c) Clearly, the overall pattern of internal repetition within the sequence is consistent. Figure 3 c) indicates the correctness of the algorithm of the present invention. Compared with other software, the Python computer program script edited based on the method of the present invention can extract all the repetitive sequence alignment results. Therefore, the method of the present invention successfully locates the boundaries of previously discovered large-fragment internal repetitions ( Figure 2 (B)

[0169] Example 3: Application of the method and computer program of the present invention for batch identification of internal repetitions in sequences

[0170] The method of this invention can be applied to batch identification of internal repeats in sequences. Large-fragment internal repeats in all coding sequences of the model plant Arabidopsis thaliana genome were identified using the method in Example 1 of this invention to illustrate the beneficial effects of batch identification (the identity parameter was set to 0.65, and the minimum output repeat length parameter repeatLen was set to 300).

[0171] The results showed that among the 27,416 coding sequences in Arabidopsis thaliana, the number of coding sequences containing large positive repeats (288) was greater than the number of coding sequences containing large negative repeats (4) (Table 3).

[0172] Table 3. Results of large intra-fragment duplication identification of all coding sequences in the Arabidopsis genome using the method of the present invention.

[0173]

[0174]

[0175] Functional annotation of these genes containing large repetitive segments was performed. Based on gene annotation classification using the TAIR database, these genes were found to belong to multiple gene families, such as the F-box family, the jacalin lectin family, and the disease resistance protein (TIR-NBS-LRR class) family. GO annotation using the agriGO v2.0 tool also indicated that they belong to multiple biological functions, such as "t-type cell wall organization," "carbohydrate derivative binding," and "intrinsic component of membrane." KEGG analysis using the KOBAS-i tool showed that these genes can participate in pathways such as ABC transporters, RNA transport, and RNA polymerase. Some of these fusion genes can be found within these large repetitive segments; for details, please refer to Example 5 below for an application example of the method of this invention.

[0176] Example 4: Application of the method of the present invention—Statistical proof of the repetitive nature within gene functional regions

[0177] Pectin methylesterase (PME) is an enzyme that catalyzes the demethylation of pectin in the cell wall and is widely found in plant cell walls. Type-1 PME in the PME gene family consists of an N-terminal pro-region and a C-terminal mature region (a conserved PME domain). The PME domain is the conserved functional domain of PME. The pro-region, however, is more divergent and generally has three biological functions: first, to help PME correctly locate in the cell wall; second, to help PME proteins fold correctly; and third, to inhibit PME enzyme activity. Evolutionary analysis shows that the pro-region is more active and variable than the PME domain. Previous studies have found that the pro-region of the Type-1 PME gene in foxtail millet contains more internal repetitions than the PME domain (Related literature: Ge W, Chen H, Zhang Y, Feng S, Wang S, Shang Q, Wu M, Li Z, Zhang L, Guo H, Jin Y, Wang X. Integrative genomics analysis of the ever-shrinking pectinmethylesterase (PME) gene family in foxtail millet (Setaria italica). Funct PlantBiol. 2022 Sep; 49(10):874-886.). This suggests that this may be the driving force behind the evolutionary characteristics of the pro-region. However, it remains unknown whether this is the case for other plants or for the entire plant kingdom as a whole.

[0178] Using the method in Embodiment 1 of this invention and its corresponding computer program product, representative plants in the plant kingdom (including: 7 dicotyledons (Brassica oleracea, Brassica rapa, Arabidopsis thaliana, Carica papaya, Populus trichocarpa, Vitis vinifera, and Solanum lycopersicum), 6 monocotyledons (Zea mays, Sorghum bicolor, Setaria italica, Brachypodium distachyon, Barley, and Oryza sativa), 1 basal angiosperm (Amborella trichopoda), 1 fern (Selaginella moellendorfii), and 1 bryophyte (Physcomitrella)) are analyzed. The coding sequences of the pro-region and PME domain of the Type-1 PME gene in *Chlamydomonas reinhardtii* and five green algae (*Chlamydomonas reinhardtii*, *Volvox carteri*, *Coccomyxa subellipsoidea* C-169, *Micromonas sp. RCC299*, and *Ostreococcus lucimarinus*) were analyzed (data downloaded from: https: / / figshare.com / articles / dataset / Coding_sequences_of_PME-domain_and_pro-region_of_Type-1_PME_in_representative_plants / 25160780) to identify all internal duplications. Figure 4 (a) Furthermore, the internal repeatability of the two regions was compared using the repeat density index developed by the method of this invention. The results showed that the coding sequence of the pro-region of the Type-1 PME gene in plants had a greater internal repeatability density than the coding sequence of the PME domain. Figure 4(c) and it was statistically significant (t-test, P<0.001). Therefore, the method of the present invention statistically demonstrates the nature of the Type-1 PME gene and shows that the repeat density index can be an aspect for evaluating the nature of a gene.

[0179] Example 5: An application example of the method of the present invention for inferring gene fusion through long internal repeats.

[0180] This embodiment demonstrates a method for identifying long repetitive sequences within a gene sequence and thereby inferring fusion genes using the method described in Embodiment 1 of the present invention. This illustrates the potential application capabilities of the method of the present invention. The principle of using the method of the present invention to infer fusion genes by identifying long repetitive sequences within a gene is as follows:

[0181] First, the method described in Example 1 of this invention is used to identify a batch of genes containing long repetitive sequences within genes; then, a dot matrix diagram is plotted for observation, and the long repetitive sequence is extracted; next, gene feature structure analysis is performed on the long repetitive sequence to determine that it is a sequence with complete gene characteristics, and the structural annotation of the gene is corrected to eliminate the phenomenon caused by genome annotation errors; finally, this invention can determine whether it is a fusion gene (…). Figure 5 (a)

[0182] Taking the internal repeat analysis of the coding sequences of the aspartic protease gene family in the model plant Arabidopsis thaliana as an example, the specific implementation steps are demonstrated as follows: First, using the method in Example 1 of this invention, long fragment internal repeats are identified in batches from the coding sequences of the four aspartic protease genes in Arabidopsis thaliana (data download link: https: / / figshare.com / articles / dataset / Raw_data_from_fusion_gene_analysis_by_LIR_inference / 25160993). (fileType=0 (0 indicates that the analyzed sequence is of DNA type), identityThr=0.85, repeatLen=600, DNA_Matrix=0 (0 indicates a permutation matrix where x equals 5 and y equals -4), mode=0,1 (0 indicates that forward repeats should be output, 1 indicates that reverse repeats should be output; both indicate that forward and reverse repeats are identified and output). The internal repeat position files of all target sequences are obtained. Then, a Python script is used to batch plot the internal repeat raster of the sequences.

[0183] The results showed that only the coding sequence of the AT2G28220 gene contained a large internal positive repeat sequence (repeat length 10¹⁴, identity 0.90927). Figure 5 (b). Next, the exon-intron structure of the gene was analyzed using CFVisual (version 1.0) software (related literature: Chen H, Song X, Shang Q, Feng S, Ge W. CFVisual: an interactive desktop platform for drawing gene structure and protein architecture. BMC Bioinformatics. 2022 May 13; 23(1):178. doi:10.1186 / s12859-022-04707-w.), SMART website (http: / / smart.embl-heidelberg.de / ), and MEME website (https: / / meme-suite.org / meme / tools / meme). Figure 5 c), functional domains of amino acid sequences and motifs Figure 5 Analysis of the large fragment (d) showed that the repetitive sequence within it possesses the characteristic domain of aspartic protease, proving that the AT2G28220 gene is a fusion gene. Furthermore, chromosome localization analysis using MapChart (version 2.32) software (related literature: Voorrips RE. MapChart: software for the graphical presentation of linkage maps and QTLs. J Hered. 2002 Jan-Feb; 93(1):77-8.) revealed that the AT2G28220 gene is located within a gene tandem repeat cluster ( Figure 5The presence of the tandem duplicated fusion gene indicates that it belongs to the tandem duplicated fusion gene type (related literature: Rogers RL, Hartl DL. Chimeric genes as a source of rapid evolution in Drosophila melanogaster. Mol Biol Evol. 2012 Feb; 29(2):517-29; Zhou Y, Zhang C, Zhang L, Ye Q, Liu N, Wang M, Long G, Fan W, Long M, Wing RA. Gene fusion as an important mechanism to generate new genes in the genus Oryza. Genome Biol. 2022 Jun 15; 23(1):130.).

[0184] Finally, this invention conducts a more in-depth evolutionary analysis, using methods previously employed by our laboratory (related literature: Wang X, Gowik U, Tang H, Bowers JE, Westhoff P, Paterson AH. Comparative genomic analysis of C4 photosynthetic pathway evolution in grasses. Genome Biol. 2009; 10(6):R68), to deduce the evolutionary model of this fusion gene. Figure 4 The specific model analysis results are as follows:

[0185] First, the original ancestral gene underwent three tandem replications, forming four tandem repeat genes. Then, the newly generated repeat gene underwent one tandem replication. Finally, this repeat gene fused with the ancestral gene, forming the fusion gene AT2G28220 that is currently detected. Thus, the method of this invention demonstrates its potential application in inferring gene fusions through long internal repeats, and can be used for gene fusion prediction and analysis.

[0186] The present invention has been described in detail above. Those skilled in the art will recognize that the invention can be practiced in a wide range of ways with equivalent parameters, concentrations, and conditions without departing from its spirit and scope, and without requiring unnecessary experiments. While specific embodiments have been provided, it should be understood that further modifications can be made to the invention. In summary, according to the principles of the invention, this application is intended to include any changes, uses, or improvements to the invention, including changes made using conventional techniques known in the art that depart from the scope disclosed herein.

Claims

1. A computer device comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to perform the following steps: S1. Establish the permutation matrix: Establish the permutation matrix based on the sequence to be analyzed; the sequence to be analyzed is represented by sequence A and sequence B, and sequence A and sequence B are the same, both being the sequence to be analyzed. S2. Score Matrix Creation: Create the score matrices for sequence A and sequence B. H The scoring matrix H for n +1 row and m A matrix with +1 columns, where m and n are the lengths of the sequence to be analyzed, and the scoring matrix... H Set all elements in the first row and first column of the score matrix to 0. H The rows excluding the first row and the columns excluding the first column are scored based on the permutation matrix, and the scores are then filled into the score matrix. H The fractions of each element except the first row and the first column are filled according to the following rules: Equation (3); In formula (3): H ij Representing the i line, number j The score of the matrix elements in the column; i greater than or equal to 1 and less than or equal to 1 n natural numbers, j greater than or equal to 1 and less than or equal to 1 m natural numbers, For the first i -1 line, number j The score of the matrix elements in column -1; S3. Backtracking: Perform the following backtracking: S31. Initial backtracking: From the score matrix... H The element with the highest score, A ij Let's start by backtracking to element A. i-1j-1 This process is repeated until an element with a score of 0 is encountered, yielding the first backtracking result; S32, Second backtracking: from the score matrix... H The element B with the highest score outside the path of the first backtracking. rc Let's start by tracing back to element B. r-1c-1 This process is repeated until an element with a score of 0 is encountered, resulting in the second backtracking result; S33, from the score matrix... H The element C with the highest score outside of the first backtracking path and the second backtracking path. tu Let's start by backtracking to element C. t-1u-1 This process is repeated until an element with a score of 0 is encountered, resulting in the third backtracking result; this process is repeated until the v-th backtracking result is obtained, where v is a natural number greater than or equal to 1. S4. Result Output: Based on the backtracking result, obtain the repeating fragment and the position of the repeating fragment on the sequence to be analyzed to obtain the repeating unit inside the sequence; A ij Represents the score matrix H The Middle i line, number j The element of column A i-1j-1 The score matrix H The Middle i -1 line, number j -1 column elements; B rc Represents the score matrix H The Middle r line, number c The element of column B r-1c-1 The score matrix H The Middle r -1 line, number c The elements of column -1 r It is a natural number greater than or equal to 1 and less than or equal to n. c It is a natural number greater than or equal to 1 and less than or equal to m; C tu Represents the score matrix H The element in row t and column u of C t-1u-1 Represents the score matrix H The element in the (t-1)th row and (u-1)th column of the array, where t is a natural number greater than or equal to 1 and less than or equal to n, and u is a natural number greater than or equal to 1 and less than or equal to m; The sequence to be analyzed is a gene sequence or an amino acid sequence.

2. The computer device according to claim 1, characterized in that: The permutation matrix is ​​either permutation matrix Q1 or permutation matrix Q2; The permutation matrix Q1 is represented as: Equation (1); In formula (1): ai For the first in sequence A i 1 sequence element, bj For the first in sequence B j 1 sequence element, i It is a natural number greater than or equal to 1 and less than or equal to n. j A natural number greater than or equal to 1 and less than or equal to m. x for ai and bj The similarity score is a positive real number. y for ai and bj The dissimilarity score is a negative real number. The permutation matrix Q2 is a BLOSUM matrix.

3. A method for identifying repeating sequences within a sequence, characterized in that: The method includes the following steps: C1) Establishing a permutation matrix: Establishing a permutation matrix based on the sequence to be analyzed; the sequence to be analyzed is represented by sequence A and sequence B, and sequence A and sequence B are the same, both being the sequence to be analyzed; C2) Score Matrix Creation: Construct score matrices for sequence A and sequence B. H The scoring matrix H for n +1 row and m A matrix with +1 columns, where m and n are the lengths of the sequence to be analyzed, and the scoring matrix... H Set all elements in the first row and first column of the score matrix to 0. H The rows excluding the first row and the columns excluding the first column are scored based on the permutation matrix, and the scores are then filled into the score matrix. H The fractions of each element except the first row and the first column are filled according to the following rules: Equation (3); In formula (3): H ij Representing the i line, number j The score of the matrix elements in the column; i greater than or equal to 1 and less than or equal to 1 n natural numbers, j greater than or equal to 1 and less than or equal to 1 m natural numbers, For the first i -1 line, number j The score of the matrix elements in column -1; C3) Backtracking: Perform the following backtracking: S31, Initial backtracking: From the score matrix H The element with the highest score, A ij Let's start by backtracking to element A. i-1j-1 This process is repeated until an element with a score of 0 is encountered, yielding the first backtracking result; S32, Second backtracking: from the score matrix... H The element B with the highest score outside the path of the first backtracking. rc Let's start by tracing back to element B. r-1c-1 This process is repeated until an element with a score of 0 is encountered, resulting in the second backtracking result; S33, from the score matrix... H The element C with the highest score outside of the first backtracking path and the second backtracking path. tu Let's start by backtracking to element C. t-1u-1 This process is repeated until an element with a score of 0 is encountered, resulting in the third backtracking result; this process is repeated until the v-th backtracking result is obtained, where v is a natural number greater than or equal to 1. C4) Output results: Based on the backtracking results, obtain the repeating segments and the positions of the repeating segments on the sequence to be analyzed to obtain the repeating units within the sequence; A ij Represents the score matrix H The Middle i line, number j The element of column A i-1j-1 The score matrix H The Middle i -1 line, number j -1 column elements; B rc Represents the score matrix H The Middle r line, number c The element of column B r-1c-1 The score matrix H The Middle r -1 line, number c The elements of column -1 r It is a natural number greater than or equal to 1 and less than or equal to n. c It is a natural number greater than or equal to 1 and less than or equal to m; C tu Represents the score matrix H The element in row t and column u of C t-1u-1 Represents the score matrix H The element in the (t-1)th row and (u-1)th column of the array, where t is a natural number greater than or equal to 1 and less than or equal to n, and u is a natural number greater than or equal to 1 and less than or equal to m; The sequence to be analyzed is a gene sequence or an amino acid sequence.

4. The method according to claim 3, characterized in that: The permutation matrix is ​​either permutation matrix Q1 or permutation matrix Q2; The permutation matrix Q1 is represented as: Equation (1); In formula (1): ai For the first in sequence A i 1 sequence element, bj For the first in sequence B j 1 sequence element, i It is a natural number greater than or equal to 1 and less than or equal to n. j For any natural number greater than or equal to 1 and less than or equal to m, x for ai and bj The similarity score is a positive real number. y for ai and bj The dissimilarity score is a negative real number. The permutation matrix Q2 is a BLOSUM matrix.

5. A computer program product, including a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of claim 3 or 4.

6. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program that causes a computer to perform the following steps: S1. Establish the permutation matrix: Establish the permutation matrix based on the sequence to be analyzed; the sequence to be analyzed is represented by sequence A and sequence B, and sequence A and sequence B are the same, both being the sequence to be analyzed. S2. Score Matrix Creation: Create the score matrices for sequence A and sequence B. H The scoring matrix H for n +1 row and m A matrix with +1 columns, where m and n are the lengths of the sequence to be analyzed, and the scoring matrix... H Set all elements in the first row and first column of the score matrix to 0. H The rows excluding the first row and the columns excluding the first column are scored based on the permutation matrix, and the scores are then filled into the score matrix. H The fractions of each element except the first row and the first column are filled according to the following rules: Equation (3); In formula (3): H ij Representing the i line, number j The score of the matrix elements in the column; i greater than or equal to 1 and less than or equal to 1 n natural numbers, j greater than or equal to 1 and less than or equal to 1 m natural numbers, For the first i -1 line, number j The score of the matrix elements in column -1; S3. Backtracking: Perform the following backtracking: S31. Initial backtracking: From the score matrix... H The element with the highest score, A ij Let's start by backtracking to element A. i-1j-1 This process is repeated until an element with a score of 0 is encountered, yielding the first backtracking result; S32, Second backtracking: from the score matrix... H The element B with the highest score outside the path of the first backtracking. rc Let's start by tracing back to element B. r-1c-1 This process is repeated until an element with a score of 0 is encountered, resulting in the second backtracking result; S33, from the score matrix... H The element C with the highest score outside of the first backtracking path and the second backtracking path. tu Let's start by backtracking to element C. t-1u-1 This process is repeated until an element with a score of 0 is encountered, resulting in the third backtracking result; this process is repeated until the v-th backtracking result is obtained, where v is a natural number greater than or equal to 1. S4. Result Output: Based on the backtracking result, obtain the repeating fragment and the position of the repeating fragment on the sequence to be analyzed to obtain the repeating unit inside the sequence; A ij Represents the score matrix H The Middle i line, number j The element of column A i-1j-1 The score matrix H The Middle i -1 line, number j -1 column elements; B rc Represents the score matrix H The Middle r line, number c The element of column B r-1c-1 The score matrix H The Middle r -1 line, number c The elements of column -1 r It is a natural number greater than or equal to 1 and less than or equal to n. c It is a natural number greater than or equal to 1 and less than or equal to m; C tu Represents the score matrix H The element in row t and column u of C t-1u-1 Represents the score matrix H The element in the (t-1)th row and (u-1)th column of the array, where t is a natural number greater than or equal to 1 and less than or equal to n, and u is a natural number greater than or equal to 1 and less than or equal to m; The sequence to be analyzed is a gene sequence or an amino acid sequence.

7. The computer-readable storage medium according to claim 6, characterized in that: The permutation matrix is ​​either permutation matrix Q1 or permutation matrix Q2; The permutation matrix Q1 is represented as: Equation (1); In formula (1): ai For the first in sequence A i 1 sequence element, bj For the first in sequence B j 1 sequence element, i It is a natural number greater than or equal to 1 and less than or equal to n. j A natural number greater than or equal to 1 and less than or equal to m. x for ai and bj The similarity score is a positive real number. y for ai and bj The dissimilarity score is a negative real number. The permutation matrix Q2 is a BLOSUM matrix.

8. A method for detecting fusion genes, characterized in that: The method includes the following steps: D1) Detection of repetitive sequences within a gene: Detecting repetitive sequences within the gene sequence to be tested using the method described in claim 3 or 4; D2) Repetitive sequence gene structure analysis: Gene structure analysis of the repetitive sequence was performed using gene structure analysis software to obtain the results of the repetitive sequence gene structure analysis; D3) Determine the fusion gene: Based on the gene structure analysis results of the repetitive sequence, determine whether the repetitive sequence is a fusion gene.

9. The use of the computer device of claim 1 or 2, the computer program product of claim 5, and / or the computer-readable storage medium of claim 6 or 7 in the preparation of a product for detecting fusion proteins.

10. The use of the method of claim 3, 4 or 8 in the preparation of products for detecting fusion proteins.

Images