TW201243117A - Method and system of assembling DNA reads with paired-end sequencing - Google Patents

Abstract

The present invention discloses a method of assembling DNA reads of paired-end sequencing to determine the target DNA sequences. The method comprises the steps of: inputting sequencing reads including pair-end reads; indexing the head/tail subsequences of the reads; producing initial fragment for extending; determining the reads having a distance within a specific range with respect to its paired read to be foreseeing reads; adding one foreseeing read to extend the initial fragment whenever after aligning one candidate read which is paired with the foreseeing read; and completing the process of assembly when no suitable reads for extending.

Description

201243117. VI. Description of the invention: [Technical field of the invention] The present invention relates to a DNA combination method using a paired-end sequencing sequence, in particular, a data analysis system and method for gene sequencing is a genetic engineering technique. field. [Prior Art] The genome (deoxyribonucleic acid, abbreviated as DNA) is a human genetic material that contains all gene sequences and non-gene sequences. A gene sequence is a sequence in a dirty molecule that exhibits physiological functions. In order to study the analysis of the genome, the genome must be sequenced to determine the base phase for identification and analysis. To properly analyze the genome's current general recognition is that the genome must be fully sequenced, that is, the genome is broken into small fragments, and sequencing reads are generated, and then analyzed according to sequencing reads. After analyzing the physiological functions of the gene fragments, the sequencing reads are recombined to establish a genetic map. The current methods of gene sequencing are based on the "Genenomics Maniac Craig Venter" invented in 1977, "the whole genome of the bird robbing sequencing, developed. This chemical sequencing method Known as the "smashing method", it is necessary to cut the long leg of the leg into several segments to sequence each segment, and then re-splicing the fragment sequence into a sequence of the entire DM. For example, Rydogen. "Nan〇fluidic" chip technology used by Munak and US Genomics. This technique can sequence approximately 2 base pairs per minute. However, the sequencing sequence obtained by the bird robbing method is quite time consuming in gene recombination and cannot be effectively reorganized for repeated reading. 201243117 In order to facilitate the combination of the base sequence and the problem of the sequencing and recombination of the aforementioned bird robbing method, developers have developed a well-known paired-end DNA SeqUencing technology. There have also been many recombination techniques that have been applied to paired end reads, such as the US Public Publication 20090233291 (Paired end sequencing) ^, according to the World Patent Cooperation Treaty, the open WQ bribe / Q7_ double-end sequencing (double The -ended sequencing method, and reference [9] are well known paired-end sequencing technologies. Although current gene sequencing technologies have produced many suitable paired end sequences, no technique for efficiently combining sequencing sequences has been developed. According to the commonly known paired-end DNA sequendng technique (shown in Figure 1), a sequencing primer site is added to both ends of the sequence when the sequencing procedure of the DNA sequence to be tested is performed. The adapter is Al, A2. In the determination of the base, the base sequence is synthesized by taking the sequence Sf to be tested and its reverse complement sequence SRC as a template chain. Therefore, the sequence information of each of the two test bases on the two sides of the sequence Sf can be simultaneously measured. Therefore, the short reads rl, r2 produced by sequencing the paired end genes are paired, as shown in Figure 2. The reverse complement of each read sequence and its paired reads is a sequence fragment located at the two ends of an insertion to be tested. Because the exact length of the inserted sequence is unknown, the distance between the two readings of the pairing is also unknown. The data produced by the sequencing program are as follows (13210606 for short sequences):

Sequence 1-1: TCCTGTATATTCTAAACTTAGAGATTGTTCAT Sequence 2-l:CATAAACATCTTTATAAAATACTAATAGAAAG

Sequence 1-2: AAAGGAGAGAACGTCGTCGTTTTCGTCGAAGT 201243117

Sequence 2-2: ACAACCCTAACTCTTTTTTTTTTGGCTATTGT

Sequence 1-13210605: TCTTCCGCCGTCGCAACTTTACCCAACGCCGC Sequence 2-13210605: ACCGCAAAAGCAAGATGATTCATTGTGTATCC

Sequence 1-13210606: CCTGGATCACAGCATCCACACGCACAAATATC Sequence 2-13210606: CCAATGGATTCTTTCTTTACTAACAATATCGA Conventional gene sequencing de novo assembly method There are three main types: (1) De Bruijn graph (2) overlap-arrangement- Overlap -Layout -Consensus (3) Greedy extension algorithm In the new generation of short-sequence splicing programs, most of them use the algorithm of De Bruijn graph, including Abyss[l], Velvet[2], Allpaths [3], Euler [4] and SOAPdenovo [5]. In De Bruijn's diagram, there are some nodes and directed edges of the connected nodes. Each node represents a word of length K (called k-mers), and the words of adjacent nodes overlap by κ-1 bases. All sequences are represented as De Bruijn diagrams called 'roadmaps'. Using the De Bruijn graph from the beginning of the combined method algorithm is to merge adjacent nodes into one larger node. If the result of the merging forms a complex node' then find the Eulerian path in the De Bruijn diagram as the last sequence that can be merged. The sequencing combination method of the conventional De Bruijn graph, because the node of the De Bruijn graph is that the k-mers structure ^ k-mers structure is only a partial base of the sequencing read sequence, rather than reading all the bases. Therefore, the information at the end of the match cannot be found at De 201243117.

Bruijngraph shows the wealth. Therefore, the system of the (10) sequencing combination method is constructed after the $ 〇 ) , , 才 才 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 The scaffolding operation is to arrange the contigs into longer sequence scaffolds. Although the paired-end sequencing technology can clearly show the results of the two sides of the sequence. The financial sector is regarded as a technology that can improve the correct combination rate and coverage rate. However, the traditional three main base-sequence combination methods, as shown in Table 2 of the Annex II, are the _ pairs of paired end_sequence_in sea_. The construction method is to construct the scaffold by using the paired end information after the contig is generated. Since almost half of the sequences in human genomes are repetitive, it is necessary to combine repetitive target phases when processing the sequencing data of human genomes or other organisms. We will use Figure 5 to illustrate the sequence repeat problem. The target gene sequence in the example of Fig. 5 (8) is a gene sequence consisting of 5 sub-sequence fragments ARBRC. Among them, R is a gene sequence fragment that appears twice, and A, B, and C are different gene sequence fragments. As shown in Figure 5 (9), when we want to extend the subsequence AR, we will find that the read sequence spanning (R, B) and spanning (R, c) can continue to the right of the subsequence AR. For example, AGATaccga is a read sequence spanning (R, B), and GGGGAAAAAT is a read sequence spanning (R, C). For example, we will insert the read sequence GGGGAAAAAT on the right side of the subsequence ar, which will form the wrong subsequence. Similarly, as shown in Figure 5(c), if we connect the read sequence AGATAACGGA to the right of the subsequence BR, the error subsequence of the BRB will be formed. In the field of gene sequencing, the combination of repeated sequences has always been a problem of notorious. In the light of the above-mentioned conventional techniques, the present invention has developed a completely new method called a read foreseeing sequencing combination technique architecture. Read foreseeing is a mechanism for recording the paired sequence information used in the sequence in the current combination when the sequence is combined, and then performing a read-ahead check. This mechanism uses pairing end information for extension, bridging, and repeatsensing. Therefore, the method newly created by the method of the present invention is called EBAR, which is an abbreviation of English Extension, Bridging And Repeat sensing. In order to clarify the technical terms first created in the novel technology of the present invention, such as reading preamble, abnormal matching, vague matching and divergence extension. Therefore, the Chinese and English nouns related to the technique of the extension-bridge-repetitive detection of the present invention are as shown in the attached table. [Invention] The object of the present invention is to provide a new method and system for combining paired end readings. . The present invention alters the manner in which scaffolding is traditionally performed using paired end information. The technique of the present invention is referred to as a method and system for read f〇reseeing. The present invention fully utilizes the information at the end of the pair to handle the sequence rep〇blem that would be encountered when reading the splicing combination. The technical means to achieve the aforementioned objectives is to construct a technical structure for read foreseeing. The read sequence used in the subsequent sequence defines the paired read sequence that has not yet been connected as a Foreseeable Reads (FRs), and the subsequence is continued in the foreseeable read order. The above described objects and other features of the present invention will become more apparent from the aspects of the appended claims. In the description of these embodiments 'for the sake of brevity of explanation, different sequence lengths' and different index key lengths are used in different examples. 201243117 [Embodiment] I. The basic technical architecture of the present invention:::1:7 is not a technique for read foreseeing of the invention = structure = real action. Figure 6 shows a schematic diagram of the paired read order distribution in the sequence. Figure 7 shows how to apply the read order look-ahead to determine the correct contiguous position of the read sequence. When the subsequence AR is continued, since η, γ2, θ is the read order of the bit on ar, the (4) read-reading read order (reseeable Reads' FRs) is defined. Its towel, as shown in Figure 2, reads the sequenced dragon read by the base-sequence system, and its order is (rl r2>; but when combined, the actual reading of the invention is read, and each read is The reverse complement sequence c2, cl ' of the sequence rl r2 and its partner sequence is also illustrated in Figures 3 and 4 (H, c2) and (r2, cl) 'Figure 4 (a) shows the invention One pair of paired ends rl, c2, as shown in Figure 4(b), utilizes the pair of mating ends rl, c2 to combine contigs a, C2 to create a contig bracket. Figures 5 through 7. In the example, it can be found that the read sequence r4 and rll can be connected to the right side of the sub-sequence AR, but only the read order γ4 belongs to the foreseeable read order. Therefore, as shown in FIG. 7(8), the present invention continues the sub-sequence ar with the read order r4. On the contrary, as shown in FIG. 7(b), when the sub-sequence ARBR is continued, the read sequence rii is found to be the read-read sequence 'to extend the sub-sequence ARBR into the correct ARBRC sequence. The architecture is feasible, but not all of the foreseeable readings found can be used to generate the correct extended sequence fragments. The architecture must solve two problems: (1) anomaly of pairing and (2) ambiguity of pairing. The first problem that needs to be solved when the invention is read and implemented is that it is abnormally paired. The problem in Figure 8 illustrates the problem of anomalous 201243117 pairing under the read-predictive architecture. When the repeat sequence is short and the insertion order is long, the paired read sequence may span two or more. The sequence of the overlay is as shown in Fig. 8 (r2, r5). When the subsequence AR is connected, the read sequence r4, r5, and r6 are all foreseeable. Therefore, the read sequence r5 can be continued in the subsequence ar. On the right side, an erroneous extension is formed. A similar situation occurs when the subsequence ARBR is continued. If the read sequence r4 is still in the foreseeable read set, the read sequence r4 can continue to the right of the subsequence arbr, forming an erroneous extension. To solve this problem, the solution developed by the inventors is to detect a reasonable pairing distance. If the length of the inserted sequence is about 2 base pairs (base pak, bp), we can speculate that the paired reading is in the sequence. of The distance is between 1 〇〇 and 3 〇〇 bp. In the example of Fig. 8, for example, the paired reading order (rl, r4), (r2, r5), (γ3, γ6) shown in Fig. 8 (; a) The distances are 160 bp, 145 bp and 180 bp respectively. When an abnormal pairing occurs, the distance between the paired readings is abnormally high or abnormally low. For example, as shown in Fig. 8(b), the reading sequence r5 is connected to the subsequence. When the AR is on the right side, the pairing read order (r2, r5) has a pairing distance of only 25 bp 'so we can infer that an abnormal pairing occurs, that is, although the reading order is a foreseeable reading order, it should not be connected to the sub-sequence AR right. side. A similar situation is shown in Figure 8(c). If the read sequence r4 is continued to the right of the sub-sequence ARBR, the pairing read order (rl, r4) has a pairing distance of 375 bp, so we can infer that an abnormal match occurs. Although the read sequence r4 is a foreseeable read sequence, it should not be contiguous to the right side of the subsequence ARgR. The second problem that needs to be solved when implementing the read-ahead architecture of the present invention is ambiguity of pairing. In the example in Figure 9, the problem of ambiguous pairing occurs under the read-ahead architecture. As shown in Fig. 9 (8), when the repeat sequence is long, the order of the sequence segments that cannot be judged before and after is determined. That is to say, it is impossible to judge which of the sub-sequences ARB or ARC is correct. In Figure 9(b), when the sub-sequence 201243117 is listed as AR, the ^^ is deducted and the pre-reading sequence can be legally connected to the sub-sequence AR. It is impossible to judge which ARB or ARC is the correct group. σ The inventor calls it a branching extensi〇n. Figure 10 illustrates the situation when a divergent extension occurs. When the subsequence AR is continued, one possible consecutive read sequence χΐ, χ2, χ3,·.., χ1〇 is found. It can be seen from the example of Fig. 10 that χι and χ3 are foreseeable readings. However, simultaneous X1 and Χ3 form two different extension fragments: ACTTaacgt and CCCTGGCCA. Therefore, when the system checks for a divergence extension, it is necessary to abandon the connection so as not to combine the wrong sequence. DETAILED DESCRIPTION OF THE INVENTION i The system of the present invention, as shown in Figs. 11 to 16, is a gene sequence combining system of the present invention, which is used to sequence a gene sequencing system. A collection of short sequences of the gene pairing ends, spliced into the target gene sequence. A specific embodiment includes a data reader 11, an index database 12, a pairing database 13, a start fragment generator 14, an indexer 15, a preview read checker 16, and an extender. 17. A bridge 18, a fault tolerant read sequence checker 19, a contig group constructor 2, and an output interface 21. The foregoing elements are described in detail below. A data reader 11 (which may be an input interface) for reading a plurality of readings 11 from a file stored in a database or a memory (a plurality of readings may be generated by sequencing by a genetic sequencing system) , including a paired read sequence, which is paired-end, or a longer sequence is modeled as a paired type 〇nate-pair) to form a read set and given Each of the input reads a numbered data, and each N bases before and after each reading is used as an index key value data 201243117.m, and the index key value data m is stored in the index database η. The numbered data m of the read order is stored in the paired database 13. The data reader u includes a read order use record array 114 for recording whether the read order has been queued into the sequence contig. Among them, the definition of the continuation bit in a sequence is read, and the smear field is not read by the __ material, and the domain is - foreseeable reading set. The start fragment generator 14 is configured to generate at least one start sequence segment 141, the start sequence segment 141 for use by the subsequent extender π and the bridger 18 to extend or bridge other read sequences. The indexer 15 is configured to select a sequence in the detection window 150 as the index key value 151 (as shown in FIG. 12 &) when the system extends a sequence of segments, to the index database 12 Querying the number data of the reading sequence corresponding to the index key value data 151 corresponding to the index key value 151, and converting the number data into the base string of the reading sequence, and providing the base string to the starting fragment The generator 14, the extender 17, and the preview read checker 16 use the index query of the possible read sequences ri, r2, r3, r4 (as shown in Figure 12(b)). The read sequence checker 16' is foreseen to dynamically record the foreseeable read sequence of the current extended sequence. Referring to FIG. 11, 12(b) and FIG. 14, the start fragment generator η and the extender 17 will submit the read sequence 161, 162, 163 of the candidate to be extended to the preview read checker 16 ( 1*1 into 13/4) and its position is not in the foreseeable reading order (!1, £2, 〇) foreseeing the reading order checker 16 in addition to checking the candidate reading order 161, 162, 163 (1*1/23, Whether it is currently in the set of read orders, and calculate whether the distance between it and the paired reading is 201243117. Whether it is reasonable within a predetermined range. The extender 17 and the bridge 18' are actually responsible for finding extensions. The current sequence of foreseeable reads 162, 163. When the extender 17 cannot find a valid foreseeable read sequence 162, the bridger 18 is used to find the foreseeable read sequence m3 that can extend the current sequence. Although the extender 17 and the bridge are used. 18 successfully establishes an extended segment of the current sequence, and then uses the fault-tolerant read sequence checker 19 to check whether the selected read sequence has a sequencing error and whether the check is extended or not, that is, when the sequence is checked for two different foreseeable reads. In the sequence, two different extensions are formed, which is the difference. Extending, at this point, it is necessary to abandon the connection 'to avoid assembling the wrong sequence. The contig constructor 20 is used to extend the extender 17 and the bridge 18, and the selected read by the fault-tolerant read sequence checker 19 is checked. The sequence 191 is arranged in an overlapping manner to construct a contig 2 〇 1. The sequence of the sequence contig 201 is output to the file 211 through the output interface 2 for use. H. The method of the present invention Please refer to FIG. 10 to FIG. 16 for a specific embodiment of the sequencing read sequence combination method of the present invention. The method of the present invention includes the following steps. Step S2〇1: From data The reader n reads a plurality of readings (ie, reads a plurality of readings generated by sequencing by the genetic sequencing system) to form a reading sequence, respectively giving each input of the input a number, and establishing each The index of the cake is stored and stored in (4) 丨龙库 U, and the pairing information of each reading order is established and stored in the pairing database 13. 12 201243117 Step S211: As shown in FIG. 11 and FIG. Start fragment generator 14 The sequence fragment 141 is started for use by the subsequent extender π and the bridge 18. A preferred embodiment of the method for generating the start sequence fragment of the present invention first utilizes the start fragment generator 14 from the data reader 11 In the read set that is read, a pair of paired readings 113 that have not been used in other sequence contigs are found, and the paired readings are used as a starting segment SegA and SegB. In the self-reading set, it is checked whether there is a paired reading 16 when the matching reading ι61 (rl, r2) which can extend the starting segments SegA and SegB respectively is used as the reading order 140 of the primary selection. The initial read sequence 161 (ie, the pair read sequence rl and ^) is used to extend the start segments SegA and SegB, respectively. When a read sequence in the self-reading set can be found on the start segments SegA and SegB at the same time, the left and right position order of the start segments SegA and SegB is determined, and the two start segments are joined by the read sequence. SegA and SegB' are in turn joined into a start sequence segment 141. Steps S221 and S222: are left-and-right symmetrical extension calculation programs, and the lengths of the respective bases on the two sides of the current sequence or the start sequence segment 丨41 are respectively taken as a detection window 142 by the extender 17. The fifteenth and sixteenth examples of the present invention illustrate the following procedure by extending to the right. Steps S231 and S232: are steps of extending the sequence of the extender 17, and are explained in conjunction with the examples of Figs. 11, 13, and 15. The extender 17 uses the aforementioned detection window 142 to find a contiguous candidate read sequence, and then queries the preview read sequence checker 16 to see if there is a candidate read order rl, r2, r3, r4 belongs to the foreseeable read order 162. Figure 15 (fl, f2, f3). If so, it is determined that the candidate read order and the preview read order fl, f2, (b) are 201243117. The forward read sequence 162 on at least one side of the start sequence fragment 41 can be extended and used to establish a read order 162. A longer extended sequence fragment 171. Steps S251 and S252: When the extender 17 does not find the look-ahead read sequence 162, then the bridger 18 queries the preview read sequence checker 16 for a set of look-ahead reads 163. Bases of length N on the side of these foreseeable readings 163 are then used as anchors 164, respectively. It is then found in the detection window 142 whether or not there is a base string of the anchor sequence 164. The program is shown in Figure 16. An array of anchors 164 having a length K is found in the detection window 142. At this time, there will be three situations. The third case 疋 When the detection window 142 cannot find the anchor sequence 164, it is determined that the preview read sequence 163 is not ranked at this position. The second case is that the anchor point sequence 164' is found in the detection window 142, but after the position of the alignment error sequence 164 is aligned, the residual test base and the test base of the detection window 142 are not coincident, then the short sequence of the preview read sequence 163 is determined. It is not in this position. In the third case, the error sequence 164 is found in the detection window 142, and after the alignment, the remaining test base is also consistent with the test basis of the detection window 142, then the preview read sequence 163 is determined to be extendable, and the read-ahead is read. The sequence 163 short sequence continues or establishes the extended sequence segment 181. Steps S26i and S262: The foreseeable reading sequence 162 and the job reading sequence 163 found from the extender 17 or the bridger 18 are sequentially used to establish the extended sequence segment (8). The fault-tolerant read sequence checker 19 then finds the read order that can be arranged in this extended sequence segment (8). ^ The order of reading will be arranged to check whether there is a sequencing error, the pairing distance is different, and the reading order of the detecting wire 142 is not selected, and the _ passer is selected as the reading 201243117. Steps S271 and S272 are screened by step S261. In the selected reading list, the fault-tolerant reading checker 19 further checks whether there is a divergence extension. If there is no disagreement, it is determined that all of the selected readings 191 can be added to a target sequence contig "and the paired data 192 of the reading in the selected reading 191 is returned to the preview reading checker 16 to update the current foreseeable reading. Ordered collection. At the same time, the pairing information of the paired reading that has been used is deleted from the pairing database. In step S291, when the two sides of the extended sequence segment 181 cannot continue to add a new read sequence, all the selected read sequences 191 are overlapped by the contig group constructor 20 into a sequence contig (c〇ntig) 201. . The output interface 21 outputs the most determined test sequence for each position of the sequence overlap group 201 to the slot case to become the base sequence 211 of the final combined target gene. Although the present invention has been disclosed in the above preferred embodiments, it is not intended to limit the invention, and the present invention may be modified and retouched in the winter without departing from the spirit and scope of the invention. The scope of protection is subject to the definition of the scope of the patent application. Figure 1 is a schematic diagram of a sequencing procedure of a conventional paired end; Figure 2 is a schematic diagram of a sequencing readout generated by a conventional paired end; Figure 3 is a schematic diagram of a conventional paired sequencing read position relationship; (a) and (b) are schematic diagrams of conventional contig scaffolds; Figures 5(a), (b) and (c) are schematic diagrams illustrating conventional repeat sequence problems; A schematic diagram of a read-predictive architecture using a read sequence; 15 201243117. Figure 7 (8) and (b) are schematic diagrams of a repeating sequence for the foreseeable reading of the invention; Figures 8(8)(b) and (6) are illustrative Schematic diagram of the abnormal pairing and solution of the invention; Figures 9(8) and (b) are schematic diagrams illustrating the ambiguous pairing of the present invention; 10 is a schematic diagram of the branching extension and solution of the present month; BRIEF DESCRIPTION OF THE DRAWINGS FIG. 12 is a schematic diagram of a read sequence f of the present invention; FIG. 13 is a flow chart of a combination of Wei and end phases of the present invention; FIG. , (9), (6) and (the gorge illustrate the invention FIG. 15 is a schematic diagram of an extender and an extended operation of the present invention; and FIG. 16 is a schematic diagram of the operation of the bridge and bridge of the present invention. Annex 1: References. One is the comparison between the prior art and the technique of the present invention using paired-end sequencing information, and Table 1 is a comparison of Chinese and English nouns for the novel gene sequencing combination technique of the present invention.

Sequencing Reader Combination System 10 Read Order 110, 113, 161, 162, 163, 191 Data Reader 11 Paired Data 1U Index Correspondence Data 112 Read Order Use Record Array H4 Index Database 12 Paired Database Start Fragment Generator ηStart Sequence Fragment 141 201243117 Detection Window 142, 150 Indexer 15 Index Key Value 151 Preview Reader Checker 16 Anchor Sequence 164 Extender 17 Extend Sequence Segment 171, 181 Bridge 18 Fault Tolerant Reader Checker 19 Pairing Data 192 Contig group constructor 20 sequence contig 201 output interface 21 base sequence of target gene 211 17

Claims (1)

201243117 VII. The scope of the application for patents · The method of sequel to the order of life, the collection of the sequence of the _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ The reading sequence of the complex array pairing; establishing the index data of the reading sequence, and the pairing number data of the pairing reading sequence; using the 4-order generation-starting sequence, the segment is selected from the _ _ _ _ _ _ Waiting for the sequence to extend the sequence; wherein, in the reading order of the candidate reading order and the pairing distance is within a preview, the reading order is a foreseeable reading order; and each time after the opening (four) paragraph 阙-hetero_order, The pre-reading sequence is then continued to extend the sequence. 2. The method of claim 1, wherein the paired read sequence is a reverse complement of the read sequence and the paired read sequence. 3. The method of claim 1, wherein the length of at least one side of the sequence segment of the sequence that has been succeeded is taken as a detection window, and the candidate window is found by the detection window and the Whether the candidate read order is the foreseeable read order. 4. The gene sequence combination system according to claim 3, wherein the candidate read-side length unicorn is used as a fine sequence, and the sequence is searched in the sequence, and the presence or absence of the sequence is the same as the sequence of the error. The base string of the test, when found, and after aligning the candidate read sequence with the reward window, the remaining button base of the candidate sequence is also coincident with the hiding base of the detection window, and the riding candidate sequence. 18 201243117 5. Method as claimed in claim 3 or 4 Read order and check for sequencing errors, and the reading order of the detection window is inconsistent. 'Where, the candidate contiguous to the sequence is filtered out of the pairing distance abnormality or the method as described in claim 3 or 4, wherein the paired reading is continued in the sequence to check whether there is a divergence extension If you stop connecting. 7. The method of claim 1, wherein a pair of read sequences that have not been connected are found from the set of readings, and the paired readings are used as two starting segments, and the other candidate is The reading sequence extends the two starting segments separately, and the starting sequence can be generated by simultaneously joining the reading sequence of the two starting segments and joining the two starting segments. 8. The method of claim 1, wherein the record array is used in a read order to record whether the candidate read sequence has been connected to the sequence. 9. The method of claim 1, wherein the pairing number data of the read sequence of the received phase is deleted. 10. The method of claim 1, wherein when the two sides of the sequence are unable to continue to add a new read sequence, then all the _ read orders are overlapped by their positions into a sequence of contigs and output Sequence contigs are available for utilization.