Annotation method for analyzing genome centromere region deconstruction
Through the centroAnno method, heuristic strategy and string decomposition algorithm are used to solve the problem of recognizing centromer sequences of repeat monomers and HORs in the existing technology, and efficient and accurate detection of tandem repeat units and HORs is achieved, and analysis is improved efficiency.
Patent Information
- Application Number
- CN202410860232.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-28
- Publication Date
- 2024-11-08
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Existing centromere annotation tools are difficult to accurately identify repeat monomers and high-level tandem repeat units (HORs), and require prior information or provide detailed annotation results.
A comment method called centroAnno is proposed. The centroAnno sequence is analyzed through a heuristic strategy, and the adjacency matrix and sliding window strategy are used, combined with the string decomposition algorithm, and the tandem repeat units are automatically identified and HORs are inferred without relying on the prior template. .
Accurate detection of tandem repeat units and HORs is achieved, and the analysis efficiency is improved. It can quickly and accurately identify tandem repeat units in large-scale sequencing data, overcome the weaknesses of existing tools for identification, and is better than existing methods in human genome analysis. 10 times faster.
Smart Images

Figure CN118918952A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of biological gene sequences, and in particular to research requiring analysis of centromere region structures. Background Art
[0002] To better understand the structure of centromeres, we need to make accurate annotations, which is a basic step in analyzing their structure. The centromere region is rich in satellite DNA sequences, which are characterized by long tandem repeats. These sequences usually present a pattern of mmmm..., where m represents the repeating unit, called a monomer. The length of these monomers varies in different organisms. For example, the monomer length of the centromere region in humans is about 171 base pairs (bp), while it is 41bp in chicken. In some organisms, the centromere region may contain multiple types of monomers, with similarities ranging from 50% to 90%. These monomers are repeated in an organized manner to form higher-order tandem repeat units (HORs). For example, in the X chromosome of the human genome, the satellite DNA sequence follows a pattern similar to m1-m2-...-m11-m12-m1-m2...-m11-m12..., where m1-m2-...-m11-m12 is repeated multiple times to form a HOR consisting of 12 monomers. Centromere annotation focuses on identifying these monomers and HORs hidden within the centromere region.
[0003] A key issue facing centromere annotation is repeat annotation, and a series of tools have been designed specifically for this purpose, such as Tandem Repeat Finder (TRF), RepeatExplorer2, and TAREAN. Although these tools perform well in repeat annotation, they do not support the analysis of HORs. On the other hand, some tools are specifically designed to analyze HORs in centromere regions, such as Alpha-CENTAURI, CentromereArchitect, HORmon, and HiCAT. Although these tools can automate the analysis of HORs, they require known monomer templates, so their applicability in the analysis of centromere structures in some non-model organisms is limited. To address these challenges, TRASH and Satellite Repeat Finder (SRF) were proposed. TRASH can automatically analyze monomers from centromere sequences and infer HORs, but requires users to manually select monomers from the inferred monomer set as templates for HOR analysis, and SRF cannot provide detailed annotation results.
[0004] In general, there are two key challenges in the design of centromere annotation methods: one is to accurately identify the repeat monomers in the centromere sequence, and the other is to accurately analyze the HOR structure in the centromere sequence. Generally speaking, accurate identification of repeat monomers is an important prerequisite for analyzing the HOR structure. Summary of the invention
[0005] In this paper, we propose an annotation method (centroAnno) for analyzing centromere structure. CentroAnno first analyzes the repeating units of centromere sequences based on a heuristic strategy, which involves the determination of centromere sequences. When tandem repeating units are identified, centroAnno obtains an adjacency matrix based on the tandem repeating unit information and converts the HORs analysis problem into a graph theory problem of finding specific structures in the adjacency matrix. During this period, a sliding window strategy is combined to refine the structural analysis results of HORs.
[0006] Specifically, the present invention is achieved through the following technical solutions:
[0007] A centroAnno annotation method for parsing centromere regions comprises the following steps:
[0008] S1: Given a sequence, determine whether the sequence has tandem repeats. Subsequent analysis is only for sequences with tandem repeats.
[0009] S2: Determine the high-frequency k-mer in the sequence (a continuous string of length k in the sequence) and determine the distance between all high-frequency k-mers, which implies the length of the tandem repeat unit;
[0010] S3: Based on the distance obtained in S2, the set of tandem repeat units is obtained by directly intercepting subsequences in the sequence, and the sequence is decomposed based on this set. In this process, the sequence is decomposed into a list of blocks, each of which implies a tandem repeat unit;
[0011] S4: Based on the results obtained in S3, detect the incorrectly annotated areas in the sequence, where each incorrectly annotated block usually has a lower score.
[0012] S5: Re-inferring tandem repeat units in the detected incorrectly annotated regions and expanding the current tandem repeat set. Based on the expanded tandem repeat set, re-decomposing the sequence to obtain a more refined annotation result.
[0013] S6: Convert the annotation results obtained in S5 into an adjacency matrix and infer HORs based on the adjacency matrix.
[0014] Further, the S1 includes:
[0015] We count the frequency of each k-mer in the sequence. For a sequence of length n, there are a total of n-k+1 k-mers. K-mers with a frequency greater than 1 are called repeated k-mers. We count the proportion of repeated k-mers in all k-mers. When the proportion is greater than a threshold, the sequence will be judged as a tandem repeat sequence.
[0016] Furthermore, in S2, a distance list of all repeated k-mers is calculated. Two adjacent and identical k-mers induce a distance. Since repeated k-mers usually appear multiple times in a sequence, a distance list can be generated. If a k-mer appears n times in a sequence (n>1), then the distance list induced by the k-mer contains n-1 elements. We count the frequency of each distance, where high frequency and relatively small distances imply the length of the tandem repeat unit.
[0017] Further, the S3 includes:
[0018] S3-1: Using the distance d calculated in S2 to intercept the tandem repeat unit. Specifically, if the distance between adjacent and identical k-mers is d, the portion between the two k-mers is intercepted in the sequence, and the portion will be used as a potential tandem repeat unit.
[0019] S3-2: Based on all the tandem repeat units, we use a string decomposition algorithm to decompose the sequence into a block list. Specifically, the input of the string decomposition algorithm is a sequence and a set of potential tandem repeat units, and the output is a block list. Each block in the list has a label and a score, where the label indicates the tandem repeat unit number, and the score indicates the confidence that the block is a specific tandem repeat unit;
[0020] Further, the S4 includes:
[0021] S4-1: When the length of consecutive blocks exceeds a certain threshold (the default is 5000), we call these blocks a block region. We detect block regions with low scores, which suggest that we may have inferred an incomplete set of tandem repeat units. In the default setting, when a block score is lower than 85, it is recorded as a low-scoring block. For each low-scoring block region, we infer tandem repeat units again based on the S2 scheme;
[0022] S4-2: In the actual situation, there are also some blocks with relatively low scores scattered in various parts of the sequence, resulting in the failure to form a region of blocks with low scores. We extract these blocks separately and complete clustering. The consensus sequences generated by each cluster in the clustering result are used as new tandem repeat units;
[0023] Further, the S5 includes:
[0024] S5: Based on S4, we will obtain some new tandem repeat units. These tandem repeat units are used as the input of the string decomposition algorithm to decompose the sequence again. At this time, the decomposition result is used as the final annotation result of the tandem repeat units.
[0025] Further, the above method also includes the parsing step of HORs:
[0026] The first step: Construct an adjacency matrix. After S5 is completed, we will obtain a list of blocks. Each block in the list will be assigned a label, which will induce a label sequence. For example, a label sequence corresponding to a block list may be 1-2-3-1-2-3..., where '1' implies that the first block in the sequence corresponds to the 1st tandem repeat unit. Correspondingly, the first block also corresponds to a score, which implies the confidence of the first block as the 1st tandem repeat unit. A label sequence S with a length of n corresponds to an n×n adjacency matrix. For each element A[i][j] of the matrix, 1 < i, j ≤ n, when S[i] = S[j], A[i][j] = 1, otherwise it is 0.
[0027] The second step: Determine the specific structure in the adjacency matrix. Each diagonal in the matrix is called a wave. For an n×n adjacency matrix, it has n waves. Starting from the main diagonal, the length of each wave from left to right is n, n - 1,..., 2, 1. These waves are wave 1, wave 2,..., wave n in turn. For each wave i, if there are m consecutive elements with a value of 1 and m is not less than i, we say that these consecutive elements form a potential HOR region. The first i elements corresponding to these regions essentially correspond to a HOR. For example, for a simple block sequence 1-2-1-2, we will obtain a potential HOR region with a length of 2 on the 2nd wave of its corresponding adjacency matrix. The first two elements of this region correspond to "1-2", and "1-2" is a HOR.
[0028] Compared with the prior art, the advantages and beneficial effects of the present invention are:
[0029] The heuristic strategy proposed by the present invention effectively realizes the detection of tandem repeat units. Compared with the current mainstream centromere structure analysis tools, the present invention does not require prior information similar to tandem repeat unit templates to realize the analysis of HORs. In addition, compared with the current tandem repeat analysis tools, the strategy proposed by the present invention is more accurate. For example, compared with the most advanced tool TRASH, the present invention can more accurately detect the length of the tandem repeat unit, effectively overcoming the weakness of TRASH that two consecutive tandem repeat units are identified as a single tandem repeat unit.
[0030] On the other hand, the tool proposed by the present invention has a significant efficiency advantage over current tools. When analyzing the centromere region of the human genome, the present invention is 10 times faster than the current method while ensuring accuracy. This efficiency advantage and accuracy advantage provide a strong guarantee for directly analyzing tandem repeat units in huge sequencing data. BRIEF DESCRIPTION OF THE DRAWINGS
[0031] Figure 1 This is the workflow of centroAnno. a. Workflow diagram of centroAnno. b. Two methods for counting k-mers. c. The process of inferring candidate monomers. The k-mer distance refers to the number of bases between two identical k-mers on a linear sequence. d. The process of inferring high-confidence monomers from candidate monomers involves using the same string decomposition algorithm as StringDecomposer. Using edit distance as a metric, a density-based clustering algorithm is used to cluster all blocks, and a multiple sequence alignment algorithm is used to obtain the consensus sequence of each cluster as a high-confidence monomer. e. Bad blocks in the Block list. These bad blocks form bad regions. f. The process of inferring HORs. First, an adjacency matrix of the relevant block list is constructed. "1" represents a "hit", and HORs are mainly determined from the diagonal containing the most "hits".
[0032] Figure 2is a comparison of centromere analysis methods on 54 simulated datasets. a. Annotation rates of HOR structures in simulated datasets by different centromere analysis methods. b. The most frequent length of HOR inferred by all tools on simulated datasets. c. The most frequent length of HOR inferred by centroAnno on simulated datasets. d. The distribution of monomer lengths inferred by centroAnno and TRASH on simulated datasets. e. Annotation rates of monomers by centroAnno and TRASH on simulated datasets. f. Comparison of running time of repeated annotations by centroAnno and TRASH on simulated datasets. The figure shows the average running time of each method on all datasets. g. Comparison of memory usage of repeated annotations by centroAnno and TRASH on simulated datasets. h. Running time of HOR annotation by centromere analysis methods on all simulated datasets. The figure shows the average running time of each method on all datasets. i. Memory usage of HOR annotation by centromere analysis methods on all simulated datasets.
[0033] Figure 3 Comparison of centromere analysis methods on the human CHM13-T2T genome (v2.0). a. Repeat annotation rates of centroAnno and TRASH. b. Repeat lengths inferred by centroAnno and TRASH. c. The most frequent length of HOR inferred by the centromere analysis method. "0" indicates analysis failure. HORmon-v1.0 represents the HOR annotation results of HORmon in the human CHM13-T2T genome (v1.0). d. The actual length distribution of the most frequent HOR on chromosome 5 inferred by HiCAT. For convenience, only the first five HORs were used in the drawing process. e. The actual length distribution of the most frequent HOR on chromosome 15 inferred by HiCAT. For convenience, only the first five HORs were used in the drawing process. f. The centromere region contains αSatellite (αSat) and HumanSatellite 1 (HSat1). g. The repeat (monomer) length distribution inferred by centroAnno on the centromere region containing αSat and HSat1. h. Distribution of repeat lengths inferred by HiCAT on centromere regions containing αSat and HSat1. i. Distribution of run time of centromere analysis methods on the human CHM13-T2T genome. j. Distribution of memory usage of centromere analysis methods on the human CHM13-T2T genome.
[0034] Figure 4This is the centromere structure analysis of 4 species. a. Reported monomer lengths of 4 species. b. Distribution of monomer lengths inferred by centroAnno. "MonoM" indicates a monomer with a length of M inferred by centroAnno. c. Repeat length distribution of centroAnno on chromosome 3 of grape. d. RepBlock-MonoTemp consistency distribution of HiCAT on chromosome 3 of grape using monomer templates reported (Mono107) and inferred (Mono135) by centroAnno. e. Repeat length distribution of TRASH on chromosome Y of fish. f. RepBlock-MonoTemp consistency distribution of HiCAT on chromosome Y of fish using monomer templates reported (Mono512) and inferred (Mono524) by centroAnno. g. RepBlock-MonoTemp consistency distribution of centroAnno for three species. Shown as a line graph with the average value of 12 rice chromosomes. h. RepBlock-MonoTemp consistency distribution of centroAnno for grape. i. CentO counts in rice reported by
[20] and inferred by centroAnno. j. Repeat unit ratio estimated in rice. k. Heat map of the centromere region of grape chromosome 2 (top), and RepBlock-MonoTemp consensus distribution of centroAnno (bottom). l. Monomer ratio in Arabidopsis chromosomes. Monomer-x represents template x, and monomer-x' represents reverse complement. m. Monomer array of Arabidopsis chromosome 1. Different colors in the monomer array represent different monomers.
[0035] Figure 5 This is a summary of a method for structural annotation of centromere regions in genomes. Given a sequence with tandem repeats or high-order tandem repeat patterns, the first step is to obtain the centromere region sequence of the genome; the second step is to analyze the centromere region by a heuristic algorithm; the third step is to combine graph theory and optimization theory for data processing; the fourth step is to output the analysis results as a structural annotation report. Finally, the centromere region in the genome fragment is highlighted. DETAILED DESCRIPTION
[0036] The technical solution of the present invention is further described below in conjunction with embodiments.
[0037] Example 1
[0038] The annotation method for analyzing the centromere region in the genome of the present invention, the specific theoretical support, the inference method of the tandem repeat units in the centromere region and the inference method of HORs are as follows:
[0039] 1. Theoretical support:
[0040] 1. If a sequence is a tandem repeat sequence, especially a centromere sequence, due to its characteristics, it can be decomposed into a series of blocks, each of which represents a tandem repeat unit. For example, a centromere sequence S = S1S2S3..., where Si∈{A, T, C, G}, can be decomposed into ABAB..., which means that this sequence is composed of two tandem repeat units, A and B. From the perspective of evolution, A and B are evolved from the same "ancestral tandem repeat unit", and the similarity between A and B is about 50% to 90%. In particular, "AB" appears alternately in the sequence, making "AB" constitute a high-order tandem repeat unit, that is, a HOR.
[0041] 2. A provable proposition:
[0042] Proposition 1: Given two adjacent tandem repeat units A and B in the centromere sequence, assuming that the evolution rate is E and the length of the ancestral tandem repeat unit corresponding to A and B is L. Then the probability that A and B share a k-mer is not less than
[0043]
[0044] 3. String decomposition algorithm
[0045] Given a string R (corresponding to a centromeric (tandem repeat) sequence) and a set of strings Blocks (each block in Blocks corresponds to a monomer template), the goal of the string decomposition problem is to represent R using the alphabet of Blocks. We define a chain as any concatenation of blocks, and the optimal chain for R is defined as the chain with the highest score when performing a global alignment on R, which is the best representation of R among all possible chains. The goal of the string decomposition problem is to find the optimal chain for R.
[0046] Before describing the string decomposition algorithm, we first define the alignment graph. Given a block b and a string R, their standard alignment graph consists of all vertices (i,j) with 0≤i≤|b|, 0≤j≤|R|, where |x| represents the length of string x. Vertex (i,j) is connected to vertices (i+1,j), (i,j+1), and (i+1,j+1), representing insertion, deletion, and match / unmatch operations, respectively. Informally speaking, given a set of t blocks and a string R, their string decomposition graph (SD graph) consists of t standard alignment graphs, 'glued' together by their 0th row. We index the vertices in this graph using three indices (b for a block, i for a position in block b, and j for a position in string R), except for the vertices in row 0, which are simply indexed as (0,j) because all blocks share the same 'gluing' row 0. In addition, we add edges connecting each vertex (b,|b|,j) in the last row (each block b) with the vertex (0,j) in the 0th row. Formally, the SD graph is constructed on all vertices (b,i,j), where b is a block, 0≤i≤|b|, 0≤j≤|R|, assuming that for each j, all vertices (b,0,j) form a single vertex (0,j). The edge set of the SD graph includes the standard alignment edges in the standard alignment graph and the block switching edges described below. Vertex (b,i,j) is connected to vertices (b,i+1,j), (b,i,j+1), and (b,i+1,j+1), representing insertion, deletion, and match / mismatch operations, respectively (as in the standard alignment graph). The scores of these transformations are d (for insertion / deletion), r (for mismatch), and +1 (for match), respectively. In addition, vertex (b,|b|,j) is connected to vertex (0,j) of each block b in the block set by a zero-weight block switching edge. Each such edge models the transition from the aligned end of one block to the aligned beginning of the next block. Although the SD graph has directed cycles, the longest path in this graph (i.e., the path with the largest total weight of edges) is well-defined because all directed cycles have negative weights. We call the vertex (0,0) a source vertex and each vertex (b,|b|,|R|) a sink vertex of the SD graph.
[0047] The string decomposition algorithm is a wraparound dynamic programming algorithm for solving the string decomposition problem. The i-prefix of a block is defined as the string consisting of its first i symbols. Given a block b, we define a (b,i)-chain as the chain ending with the i-prefix of b [note that there may be multiple (b,i)-chains]. The dynamic programming algorithm for solving the string decomposition problem is based on computing a variable score score(b,i,j), defined as the score of the best global alignment between all possible (b,i)-chains and the j-prefix of the string R. We assume that the alignment is scored using parameters d for the penalty of insertion / deletion, r for the penalty of mismatch, and +1 for the reward of match. Similar to the fit alignment, i.e. finding a substring of string v that has the highest scoring alignment with the whole string w, the variable score score(b,i,j) is initialized to score(b,0,j)=0 for all b and j, and score(b,i,0)=i*d for all i. To compute score(b,i,j), the string decomposition algorithm uses dynamic programming to compute the length of the longest path to vertex (b,i,j) in the SD graph. The solution to the string decomposition problem is given by the longest path from a source to one of the sinks. The block b that maximizes this score represents the last block in the best chain. The other blocks in this chain are inferred by backtracking from the sink (b,|b|,|R|) to the source (0,0) and following the block switching edges defined by the backtracking path. Each path from a source to a sink in the string decomposition graph encodes an alignment of a set of blocks with respect to the input string R (the score of this alignment is equal to the length of the path). In turn, each such alignment is encoded by a path in the string decomposition graph. Therefore, the solution to the string decomposition problem is the longest path from a source to a sink in the string decomposition graph.
[0048] 2. Detection Algorithm of Tandem Repeat Units
[0049] Given a sequence, the detection of tandem repeat units is divided into three steps: the first is to determine whether the sequence has tandem repeats. When analyzing tandem repeat units directly in sequencing data, this scheme can quickly screen out centromere (tandem repeat) sequences. The second is to calculate the k-mer distance and determine the distance with high confidence, which implies the length of the tandem repeat unit. Finally, based on the distance, the potential tandem repeat units are cut off from the sequence, the string decomposition algorithm is used to decompose the sequence, and the final candidate tandem repeat units are determined based on the clustering scheme.
[0050] 1. Determination of tandem repeat sequences (such as Figure 1 (shown)
[0051] One intuition is that subsequences in tandem repeat sequences appear multiple times, which suggests that k-mers in tandem repeat sequences have a higher frequency. Based on this intuition, the present invention designs a heuristic strategy to realize the determination of tandem repeat sequences. Specifically, we count the frequency of each k-mer in the sequence, and calculate the ratio of repeated k-mers (frequency greater than 1). If the ratio is greater than a threshold r (the default is 0.1), then the sequence is preliminarily determined to be a tandem repeat sequence. In the subsequent analysis, the sequence will be decomposed into a series of blocks through the string decomposition algorithm. Each block has a score. We will count the average of all block scores. If the average is greater than a threshold (the default is 85), then the sequence is finally determined to be a tandem repeat sequence and the corresponding analysis result is output.
[0052] 2. Determine the length of the tandem repeat unit
[0053] We use a simple example to illustrate the rationality of the present invention's strategy for determining the length of tandem repeat units. Relatively small and frequent k-mer distances are likely to correspond to the typical length of a monomer. For example, consider a distance list [40, 170, 170, 170, 340, 340, 340, 340, 340]. Although 340 appears most frequently, intuition suggests that 170 is more likely to be the length of a monomer. The frequent occurrence of 340 may be attributed to two non-adjacent monomers sharing the same k-mer, and the fact that 170 is a factor of 340 further supports the possibility that 170 is the length of the monomer. The low frequency of 40 indicates that it may appear by chance and is unlikely to represent the length of the monomer. We start iterating from the first element of the distance list, and when a reliable distance is found, we terminate the iteration and infer this distance as the length of the monomer. Specifically, we model this problem as an optimization problem:
[0054]
[0055] Among them, Loss(ad i ) measures ad i As a penalty for repeat length, identity measures the similarity between two sequences, and f is the similarity threshold. i ), equal to i-1, as a penalty function, where penalty(ad i ) represents less than ad i The number of all elements, and min(1-p ij ,p ij ) means ad i It's ad j The cost of the factor. We define d to satisfy the confidence condition C N.B, where N represents the minimum number of supporting k-mer distances, and B represents the maximum number of supporting elements for that distance.
[0056] 3. Determination of the final tandem repeat unit
[0057] After determining the length of the tandem repeat unit, we directly cut out the potential tandem repeat units in the sequence. Based on these tandem repeat units, combined with the string decomposition algorithm, we decompose the sequence into a series of blocks. We cluster these blocks and finally get a series of clusters. We calculate the consistent sequence of each cluster and use these consistent sequences as the final tandem repeat unit and decompose the sequence again through the string decomposition algorithm to obtain preliminary annotation results. 2. Inference algorithm of HORs ( Figure 1 f)
[0058] Before inferring HORs, we first check the score of each block in the sequence decomposition result. When there are a large number of blocks with low scores, we extract these blocks and obtain the consensus sequences of these blocks through clustering. These consensus sequences are used as new tandem repeat units, and the sequences are decomposed again based on the sequence decomposition algorithm to obtain the final annotation result. Based on this result, we analyze the HORs.
[0059] Each element in the block list is assigned a monomer label, and we infer the HOR based on this label information. Specifically, if a tandem repeat sequence is decomposed into m1-m2...m n-1 -m n , we will each m i Treat it as a node, i∈{0,1,...,n-1}, and construct an adjacency matrix A.
[0060] Our algorithm identifies potential HOR regions and makes inferences based on label information. We first identify the wave with the largest “hit” count, using wave H , H∈{1,...,n-1}), the wave is considered to contain the HORs with the highest frequency. Then, we H Search for all "hit" regions that meet the following conditions: 1≤m≤t, 0≤j≤j+l≤nH-1, H≤k≤k+l≤n-1, l≥H-1.
[0061] Next, we create a list L of length n to record the annotation information. When an element in the list is annotated, its value is modified to 1. We iteratively annotate a HOR region and update L. Subsequently, we use a sliding window strategy combined with the following constraints to identify a single HOR that is not in the HOR region and update L.
[0062] Finally, through the above steps, we iteratively select from wave1, wave2, ..., wave N All HOR regions were identified and annotated in (N defaults to 50).
[0063] Embodiment 2:
[0064] Based on the centromere structure analysis algorithm provided in Example 1, we designed experiments to fully verify the algorithm.
[0065] centroAnno outperforms current methods on simulated datasets: We compare centroAnno with other centromere analysis tools on 54 simulated datasets ( Figure 2 ). In the absence of template information, the HOR annotation rate of centroAnno is comparable to that of HiCAT and HORmon. When template information is provided, centroAnno-mono (which means providing a template to centroAnno) performs better on most datasets. Although centroAnno has a HOR annotation rate of 0 on some datasets, its length recognition is accurate. Compared with TRASH, centroAnno is more reliable in identifying and annotating repetitive regions, and has obvious advantages in speed and memory usage. Overall, centroAnno performs well on both simulated and real datasets, and is an efficient, accurate, and low-resource utilization tool for centromere analysis.
[0066] centroAnno outperforms current methods on real human genome centromere regions: We extracted the centromere regions of all chromosomes from the human CHM13-T2T genome (v2.0) and used these regions to compare the performance of centroAnno with other centromere analysis tools ( Figure 3 ). We use CENM to represent the centromere region of chromosome M, such as CEN1 represents the centromere region of chromosome 1. CentroAnno performed well in repeat annotation, with the repeat annotation rate on all centromere regions exceeding 0.9, while the annotation rate of TRASH on CEN15 was less than 0.8. In addition, centroAnno accurately inferred the repeat length, which was mostly consistent with the actual length of 171bp. On the contrary, TRASH misinterpreted the repeat length in some regions.
[0067] There are some differences between centroAnno and other tools in terms of HORs annotation, especially the slightly different HOR lengths inferred on CEN5, CEN10, and CEN15. It is worth mentioning that the most frequent HOR length inferred by HiCAT may not always be the actual length. Sometimes we encounter situations where we cannot obtain pure HOR regions, containing unexpected repeat regions such as HSat1. In this case, centroAnno successfully detected all repeat types, while HORmon could not handle it and HiCAT missed the HSat1 repeat.
[0068] CentroAnno completed the analysis of the centromere region in an average of 190 seconds, which is 12 to 44 times faster than other methods. In terms of memory, centroAnno only needs 0.3GB on average, which is 0.03 to 0.6 times that of other methods. At the same time, centroAnno-mono is also better than centroAnno in time and memory. These results show that centroAnno performs well in centromere analysis, with high efficiency and accuracy, surpassing other tools.
[0069] centroAnno can effectively analyze the centromere regions of the genomes of multiple species: in addition to the human genome, we also collected T2T genomes of four species (Arabidopsis, rice, grape, and the Y chromosome of fish) and extracted the centromere regions of these genomes for analysis ( Figure 4 ). We first evaluated the performance of centroAnno for repeat annotation, and the results showed that its repeat annotation rate was over 0.9 in all centromeric regions, while TRASH performed slightly worse in some regions. Next, we used centroAnno to infer the monomers of chromosome 1 and chromosome Y of these species, and the results were consistent with previous reports. However, the monomer length inferred on grape chromosome 3 was inconsistent with previous reports. centroAnno inferred its length to be 135bp, while TRASH also inferred it to be 135bp. We input these two monomer lengths into HiCAT for further analysis, and the results showed that the RepBlock-MonoTemp identity generated by the 135bp monomer template was higher than that of the 107bp template, suggesting that the monomer length of grape chromosome 3 is more likely to be 135bp. In addition, the monomer length of the fish Y chromosome inferred by centroAnno was 512bp, which was different from the previous 524bp, but this difference was also supported by TRASH and HiCAT.
[0070] We also analyzed the distribution of RepBlock-MonoTemp identity on chromosomes of various species and found that centroAnno showed high RepBlock-MonoTemp identity on Arabidopsis and fish, while the value was less than 0.9 on specific chromosomes of rice and grape. In addition, we used centroAnno to determine the high-confidence blocks of the centromeric region of each rice chromosome and calculated the tandem repeats of each centromeric region. We also found that centromeric regions with lower RepBlock-MonoTemp identity generally showed lower annotation rates and lower self-similarity. Through further analysis, we found that some centromeric regions may lack self-similarity, such as the region on grape chromosome 2. We also explored the distribution of monomers in Arabidopsis and fish and found that the centromeric region of Arabidopsis is usually composed of 2 to 3 monomers, while the Y chromosome of fish is mainly composed of one monomer. Overall, centroAnno performed well in repeat annotation and monomer inference, and provided valuable information on the self-similarity and monomer distribution of centromeric regions.
[0071] On the basis of the above embodiments, the present invention continues to describe in detail the technical features involved therein and the functions and roles played by the technical features in the present invention, so as to help technicians in this field fully understand the technical solution of the present invention and reproduce it.
[0072] Finally, although this specification is described according to implementation methods, not every implementation method contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.
Claims
1. A method for structural annotation of centromere regions in a genome, characterized in that: The method comprises the following steps: S1: Obtain the centromere region sequence of the genome; S2: analyzing the centromere region by a heuristic algorithm; S3: Combining graph theory and optimization theory for data processing; S4: Identify and annotate the tandem repeat units in this region; S5: Determine higher-order tandem repeat patterns; S6: Outputting the analysis results as a structure annotation report.
2. A method for structural annotation of centromere regions in a genome according to claim 1, characterized in that: The heuristic algorithm includes the following steps: S1: Determine whether a given centromeric region has tandem repeats based on k-mer counts; S2: Identify the k-mers that appear more frequently in the region.
3. A method for structural annotation of centromere regions in a genome according to claim 1 or 2, characterized in that: The data processing steps include: S1: Determine the tandem repeat unit by optimization algorithm; S2: Determine higher-order tandem repeat patterns through graph theory.
4. A method for structural annotation of centromere regions in a genome according to any one of claims 1 to 3, characterized in that: The method does not require the user to provide any known prior information.
5. A method for structural annotation of centromere regions in a genome according to any one of claims 1 to 4, characterized in that: The method can be implemented in command line software of a Linux system.
6. A method for structural annotation of centromere regions in a genome according to any one of claims 1 to 5, characterized in that: The output of the method includes detailed annotation information of tandem repeat units and higher-order tandem repeat patterns.
7. A software system for implementing the method according to any one of claims 1 to 6, characterized in that: The software system is based on the Linux platform, receives input data through a command line interface, and outputs annotation results.
