TWI420007B - System and method of assembling dna reads - Google Patents

Description

Combined system and method for gene sequencing sequences

The invention relates to a combined system and method for gene sequencing sequences, in particular to a data analysis method for DNA sequencing. A short string of nucleotide base sequences generated by De novo assembly DNA sequencing. The erroneous short nucleotide base sequence strings are spliced into the correct long gene sequence.

The combination technique of gene sequencing sequences is a puzzle technique applied to gene sequences. The sequence generated by gene sequencing, as in the following example (26421212 sequences):

Sequence 1: TCCTGTATATTCTAAACTTAGAGATTGTTCAT;

Sequence 2: CATAAACATCTTTATAAAATACTAATAGAAAG;

Sequence 3: AAAGGAGAGAACGTCGTCGTTTTCGTCGAAGT;

Sequence 4: ACAACCCTAACTCTTTTTTTTTTGGCTATTGT;

...

Sequence 26421209:

TCTTCCGCCGTCGCAACTTTACCCAACGCCGC;

Sequence 26421210:

ACCGCAAAAGCAAGATGATTCATTGTGTATCC;

Sequence 26421211:

CTGGATCACAGCATCCACACGCACAAATATC;

Sequence 26421212:

CCAATGGATTCTTTCTTTACTAACAATATCGA.

The combination of the above-mentioned gene sequencing sequences and the splicing of the ordinary jigsaw puzzle are mainly:

(1) The combination of gene sequencing sequences is one-dimensional string splicing. The fragment of the gene sequencing data is a string containing only four (A, G, C, T) nucleotide bases, and the splicing is to overlap the different sequence fragments according to their identical parts.

(2) The combination of gene sequencing sequences is a huge number of fragment splicing. The number of pieces of common image puzzles may be 300 or 500. 1000 pieces of puzzles are difficult to splicing. The number of sequences to be spliced by the combination of gene sequencing is enormous, often ranging from 1,000,000 to 1,000,000,000, and even more sequence fragments can be produced as technology advances.

To splicing such a large number of sequencing sequences, a very large amount of memory is often required to remember the contig products produced during the process. Moreover, because the process of gene sequencing may produce sequence errors. Therefore, how to judge errors in the sequence is also an important problem to be solved.

Traditional gene sequencing from the beginning of the combination method includes the following three:

(1) Overlap-Alignment-Consensus;

(2) De Bruijn graph (De Bruijn graph); and

(3) Greedy extension algorithm.

The following three conventional methods are further explained by the splicing examples of five gene sequences:

R1 CCCTTCCAAC;

R2 ATTTAATCCC;

R3 TTAATCCCTT;

R4 TTCCAACAGC; and

R5 AACAGCCGCC

(1) Overlap-arrangement-consistent method is the most traditional way. It consists of three phases:

Stage 1: Overlapping the two-by-two gene sequencing sequences to find out their distance. For example, r2 and r1 can align the middle 3 bases, and record d(r2, r1) = -3, as shown in the following table.

For example, r1 and r2 can align the middle 8 bases and record d(r2, r3) = -8, as shown in the following table.

Stage 2: Overlapping the two-by-two gene sequencing sequences to find out their distance. A directed graph of all gene sequences is established by the distance between gene sequences, as shown in Figure 1.

Stage 3: Find the order of the consistent relationship in the directed graph (such as the minimum expansion tree method), as shown in Figure 2. In Fig. 2, the sequence of sequence stitching can be obtained by the minimum expansion tree method by overlapping r1, r2, r3, r4, r5 from left to right. The overlap of these five gene sequences is called a contig, and the final combined sequence can be obtained from each identical base in this contig: ATTTAATCCCTTCCAACAGCCGCC, as shown in FIG.

(2) De Bruijn graph: The sequence is composed of one node per k, as shown in Fig. 4. Combining different gene sequences with the same nodes, the results shown in Figure 5 can be obtained. The De Bruijn graph merges the adjacent nodes in Figure 5 into one larger node. Because the examples in the figure form only one sequence, merging adjacent nodes will eventually form a single node, as shown in Figure 6. If the result of the merge forms a complex node, then the Eulerian path is finally searched for as the last sequence that can be merged. Patent applications of the De Bruijn diagram are disclosed in U.S. Patent Nos. 7,071,324, 7,034,143, 6,865,491, 6,689,563, and 5,683,881.

(3) Greedy extension algorithm: Select a gene sequence such as r1: CCCTTCCAAC to see if the postfix is a prefix of another word. If so, it will be overlapped. For example, TTCCAAC is the suffix of r1 and is the prefix of r4: TTCCAACAGC. So merge r1 and r2 into a contig (1, 4): CCCTTCCAACAGC. The suffix AACAGCC of the contig (1, 4) is the prefix of r5: AACAGCCGCC, so the contigs (1, 4) and r5 are merged into a contig (1, 4, 5): CCCTTCCAACAGCGCC. The contigs (1, 4, 5) do not have any suffixes that are the beginnings of other sequences, so they stop. Change the sequence from r2: ATTTAATCCC to start. The suffix of r2, TTAACCCC, is the prefix of r3: TTAATCCCTT, so the combination of r2 and r3 becomes a contig (2, 3): ATTTAATCCCTT. Finally, the contigs (2, 3) and the contigs (1, 4, 5) are combined to obtain ATTTAATCCCTTCCAACAGCCGCC.

All of the above three traditional methods require continuous merge operations. The contigs of relatively short gene sequences are combined into a contig of relatively long gene sequences. The merging process requires a large amount of memory to store the temporary results of the splicing process. However, when the amount of data is large, a large amount of memory is often required to store the result of the splicing process. Even as many as hundreds of Giga's memory can be carried out. Therefore, when the gene sequence data is large, it is often limited by the limitation of the memory, and the splicing action cannot be completed. Moreover, when there are errors in the gene sequence, they often cannot be combined.

Furthermore, there are many techniques for the recombination or analysis of gene sequences, such as the Patent No. I326431 of the Republic of China, the No. 7,809,509 project of the United States, and the technical contents published by references [1] to [10] of Annex 1. However, the prior art that has been seen so far has not been found to be a leader in the art of the present invention.

The object of the present invention is to provide a system and method for combining gene sequencing sequences to solve two problems caused by the conventional method: (1) the combined action requires a large amount of memory; and (2) the error in the allowable sequence The bases can also be combined.

In order to solve the above problem 1, the technical means of the present invention provides a bidirectional extension combination method for splicing individual gene sequences to form a target gene sequence. This technique was developed to extend the individual gene sequences simultaneously to the two sides of a sequence of genes to be joined. Because it is extended to the two sides of the gene sequence to be ligated, we can arbitrarily select a gene sequence to be connected (for a gene sequence or a contig that is composed of several gene sequences) to start the two-way two-way The action of extending the sequence of other candidate genes is extended. Finally, other gene sequences located on the same contig can be found and combined into a target gene sequence (ie, a longer contig that is composed of more gene sequences).

To solve the above problem 2, the technical means of the present invention provides a screening mechanism for fault-tolerant sequences to find the correct sequence. Sequences resulting from gene sequencing may have the following two errors:

(1) Error in base pairing mismatch: A certain base in the original gene sequence is incorrectly paired into other bases. For example, ACATTAAGCCTT is the original gene sequence, and the sequence generated by gene sequencing treatment is AGATTAAGCCTT. That is, the second base C is mismatched into G.

(2) Error in insert/deletion: The gene sequence generated by gene sequencing treatment has extra bases or a certain base number in the original gene sequence. For example, ACATTAAGCCTT is the original gene sequence, and the 4th to 5th bases are consecutive bases T. If the sequence generated by gene sequencing treatment is ACATAAGCCTT, one T is reduced at the same position as the original gene sequence, which is called a base deletion error. Conversely, if the gene sequence generated by the gene sequencing treatment is ACATTTAAGCCTT, one more T at the same position than the original gene sequence, this case is called a base insertion error.

The fault-tolerant sequence filter of the present invention finds the correct gene sequence of the contiguous gene sequence among all the candidate gene sequences, and allows the erroneous bases in the gene sequence.

In order to make the objects and other features of the present invention more comprehensible, some preferred embodiments are described below, and are described in detail with reference to Figures 7 through 16 of the accompanying drawings. In the description of these embodiments, different sequence lengths are used in different examples for the sake of simplicity of explanation. And different index key lengths.

I. System of the invention

As shown in FIG. 7, the gene sequence combining system 10 of the present invention is used to splicing a collection of gene sequences into a target gene sequence. One specific embodiment includes an input interface 11, an indexer 12, and a bidirectional extension. A combiner 13, a fault tolerant sequence filter 14, an contig constructor 15 and an output interface 16. The foregoing elements are described in detail below.

The input interface 11 reads a plurality of gene sequences (which may be a plurality of gene sequences generated by a gene sequencing system) from the file 110 stored in the database or the memory to give the input gene sequence number and establish The left index structure and the right index structure of the gene sequence. In a specific embodiment of the invention, the input interface 11 reads the gene sequence twice. The first time the gene sequence is input, the N bases before and after the gene sequence are taken as the index key value data 111, and the index key value data 111 is placed in the indexer 12 for storage. Index key values can be stringed or converted to numeric representations. The input interface 11 has a sequence usage record array to record whether the sequence has been queued into the contig.

The indexer 12 is configured to store index value data 111 of the plurality of gene sequences for finding a candidate gene sequence that may be contiguous on two sides of a waiting gene sequence. It can be an index array placed in the memory, or an index file placed on the hard disk, or a remotely located database for inputting the sequence of genes indexed through the input interface 11 (ie, The short base sequence 130), and the multiplex candidate gene sequence 122 corresponding to the index.

a bidirectional extension combiner 13 for extending the selected gene sequence determined by the fault tolerant sequence filter to at least one side of the sequence of the waiting gene and extending into a longer gene sequence until the base sequence of the target gene is determined . In the embodiment of the present invention, the M base lengths on the two sides of the waiting gene sequence (or the currently combined contig sequence) are respectively taken as an extension test window 21/22 by using the bidirectional extension combiner 13 , and the present invention is The length of a gene sequence is taken as the length of the extended test window, and the two extended test windows 21/22 are respectively used to find the gene sequence which can be attached to the extended test window as the candidate gene sequence. In a specific embodiment of the present invention, the bidirectional extension combiner 13 extends the test window from a length of 1 to divide the gene sequence in the extended extension test window into a new index key 131 and a fault tolerance comparison area 132, as shown in FIG. Shown. The new index key 131 is used to query the indexer 12 for possible candidate gene sequences. The fault tolerant alignment region 132 is provided to the fault tolerant sequence filter 14 for comparing the selected gene sequences that are correct for extension.

The fault-tolerant sequence filter 14 is configured to determine that the candidate gene sequence is a selected gene sequence that can be flanked on two sides of the waiting gene sequence. Based on the fault-tolerant alignment region 132 and the multiple candidate gene sequence 122 input by the bidirectional extension combiner 13, please refer to the screening process shown in Figures 13 and 14 to leave the selected gene with position and correct extension. Sequence 141 is given to contig constructor 15.

The contig constructor 15 arranges the selected gene sequences 141 in overlapping positions, and constructs a contig 151. This contig 151 is output to the file 161 through the output interface 16.

II. Method of the invention

Referring to Figures 7 to 16, a specific embodiment of the gene sequence combining method of the present invention includes the steps described below.

Step S201: A plurality of gene sequences are input from the input interface 11, a number of all gene sequences input is given, and a left index structure and a right index structure of the gene sequence are established and stored in the indexer 12.

Step S211: The input interface 11 finds an unused sequence from the sequence using the recording array, and first compares with the adjacent sequence to determine the correctness of each base, and the sequence is used as the two-way combiner 13 for bidirectional The extended initiation of the gene sequence 112 is initiated. Since a single gene sequence may be erroneous, several consecutively adjacent multiple gene sequences can be used as the starting pair of gene sequences. Looking for a plurality of consecutively adjacent gene sequences, the index of the displacement is used to find a sequence of genes that are identical to each other, and first overlaps into a sequence of the starting gene sequence. Among them, if adjacent gene sequences are inconsistent with each other, they cannot be used as two sides of the bidirectionally extended start-to-end gene sequence fragment.

Steps S221 and S222 are left-and-right symmetrical operation programs, which will be described here with reference to the right embodiment. The bidirectional extension combiner 13 takes out the lengths of the M bases on the two sides of the to-be-connected gene sequence 20 as the left extension test window 21 and the right extension test window 22, respectively, and shifts the test window 21/22 from the length 1 to the displacement. The gene sequence in the test window 21/22 is divided into a new index key 131 and a fault tolerant alignment area 132, a new index key 131 is used to query the indexer 12 for a possible candidate gene sequence 122, and the fault tolerance comparison area 132 is provided to the fault tolerance. Sequence filter 14 is used to compare selected gene sequences that are correct for extension.

As shown in FIG. 9, the present invention uses the two sides of the waiting gene sequence 20 (or the gene sequence fragment of the currently completed group) as a right extension test window and a left extension test window, respectively. As shown in FIG. 10, the present invention slides the left extension test window 21 and the right extension test window 22 to find a candidate gene sequence 122 that can be continued on both sides of the currently known to-be-connected gene sequence 20.

Steps S231 and S232 are left-right symmetric operation programs, and FIG. 12 is an example of generating a candidate gene sequence extending to the right. The comparison reference sequence pattern in the comparison window 23 is CACAGCAGTAAGTTTCCAATATATGGT. In this sequence, CACAGCA is used as the index key, and GTAAGTTTCCAATATATGGT is the area for fault tolerance comparison. All gene sequences whose left index key is CACAGCA are found from indexer 12. These gene sequences are also divided into indexing and error-tolerant alignment regions. Comparing the aligned reference sequence patterns of the extended window 23 and the candidate gene sequence 122, calculating the number of different bases. If the number of different bases is less than a threshold T, the gene sequence is selected as a possible extended candidate gene sequence. .

Steps S241 and S242 are left-right symmetric operation programs, and the possible extended candidate gene sequences generated by the previous step must be further tested for sequencing errors. In the method of the present invention, all the identified candidate gene sequences are arranged in an overlapping manner according to the positions at which they are likely to be extended, and the ratio of the ACGT bases at each position is calculated, that is, the bases at the same position after the arrangement of different candidate gene sequences are counted. To determine whether it is a sequence error resulting from sequencing or the candidate gene sequence is not a sequence attached to this position. For a single sequence, if the base at a certain position is different from the base at the same position of other gene sequences, there are two cases: the first case is that the gene sequence is the correct candidate gene sequence, but occurs. Sequencing errors in base pairing errors; the second case is that this gene sequence is not a candidate gene sequence that can be ligated at this position. Figures 13 and 14 illustrate these two scenarios, respectively. In Fig. 13, the sequences r1, r2, r3, r4, and r6 each have 1 to 2 bases which are inconsistent with other candidate gene sequences. However, when they overlap, the wrong base at each position does not exceed a certain percentage, such as 1/5. At this point, the wrong base is considered a sequencing error for base pairing errors. Further, in Fig. 14, the sequences r1, r2, r3, r4, r5, and r6 each have 1 to 2 bases which are inconsistent with other candidate sequences. When they overlap, r1, r2, and r3 have an incorrect base at the same position exceeding a certain ratio, such as 1/5. The base at this position is A, and the base at this position is T compared to other gene sequences, and thus the gene sequences such as r1, r2, and r3 are determined not to be candidate gene sequences attached to this position. Steps S241 and S242 also detect whether a base insertion or deletion sequencing error has occurred. Sequencing error detection for base insertion or deletion is shown in Figures 15 and 16.

Steps S251 and S252 are left and right symmetrical determination procedure steps. If the previous step generates some candidate gene sequences, which can be added to the right side of the known to-be-connected gene sequence, step S221 is performed again. If the previous step produces some candidate gene sequences that can be appended to the left side of the known to-be-supplied gene sequence, step S222 is performed again.

In step S261, when the new gene sequence cannot be added on both sides of the waiting gene sequence, all the found extensible selected gene sequences are overlapped into a contig according to their positions. The bases most determined at each position of the contig are output to be the combined target sequence.

9 and 10 are diagrams showing an embodiment of sequence combining of the bidirectional extension combiner of the present invention. This example illustrates the primary method by which the present invention seeks a population of gene sequences that can be spliced together. A small start-to-end gene sequence is extended to the two ends to find the gene sequence that can be placed in place.

Figure 11 is a simplified embodiment of a fault tolerant sequence filter of the present invention. This embodiment illustrates the relationship between a fault tolerant sequence filter and an extended test window. The extension test window is the alignment sequence that initiates the two sides of the waiting gene sequence. The bidirectional extension combiner shifts the extended test window and divides the gene sequence in the extended test window into index keys and fault tolerant alignment regions.

As shown in Figure 12, the present invention finds a method of fault tolerant alignment of candidate gene sequences that can be used to extend a gene sequence.

As shown in Figures 13 and 14, screening of candidate gene sequences is shown to detect sequencing errors in base pairing errors. Figure 13 shows the sequencing error scenario in which base pairing errors occurred, and Figure 14 shows the sequencing error scenario for non-base pairing errors.

Figures 15 and 16 show screening of candidate gene sequences to detect sequencing errors in base insertion or deletion. The detection of errors in base insertion or deletion is performed by converting the originally aligned sequence patterns into differential sequence patterns. The alignment reference pattern ref in the extended test window is converted to the differential sequence pattern dref by scanning the gene sequence. Successive identical bases are considered to be single bases. For example, the reference reference pattern ref=GTAAGTTTCCAATATATGGT, the differential sequence type dref=GTAGTCATATATGT, that is, two consecutive AA bases in ref are only represented as a single A base in dref. Similarly, two consecutive TTT bases in ref are only represented as a single T base in dref. In the screening of candidate gene sequences, the alignment of the candidate gene sequence r1 is a reference pattern of GTAAAGTTTCCAATATATGGT, and the differential sequence pattern is dr1=GTAGTCATATATGT. The alignment of the two differential sequence patterns (dref, dr1) is identical, so the alignment reference pattern of r1 is replaced by er1=GTAAGTTTCCAATATATGGT. Thus, r1 is considered to be an extensible selected gene sequence that can be ligated at this position.

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 modified 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.

10. . . Gene sequence combination system

11. . . Input interface

110,161. . . file

111. . . Index key value data

112. . . Start-up gene sequence

12. . . Indexer

122. . . Candidate gene sequence

13. . . Bidirectional extension combiner

130. . . Short base sequence

131. . . Index key

132. . . Fault tolerance comparison area

14. . . Fault tolerant sequence filter

141. . . Selected gene sequence

15. . . Overlapping group builder

151. . . Contiguous group

16. . . Output interface

twenty one. . . Left extension test window

twenty two. . . Right extension test window

Figure 1 is a conventional directed graph;

2 is a schematic diagram showing the order of arrangement in a conventional directed graph;

Figure 3 is a schematic diagram of a conventional combination of sequences in an overlapping-arranged-consistent manner;

Figure 4 is a schematic view of a conventional De Bruijn;

5 is a schematic diagram of a large node merged by adjacent nodes in a conventional De Bruijn diagram;

Figure 6 is a schematic diagram of a sequence of a conventional De Bruijn diagram;

Figure 7 is a schematic view showing an embodiment of a gene sequence combining system of the present invention;

Figure 8 is a flow chart showing an embodiment of a gene sequence combination method of the present invention;

Figure 9 is a schematic view showing the left and right extension test window of the bidirectional extension combiner of the present invention;

10 is a simplified embodiment of a sequence combination of a bidirectional extension combiner of the present invention;

Figure 11 is a simplified embodiment of a fault tolerant sequence filter of the present invention;

12 is a schematic diagram of a method for finding a candidate sequence that can be used to extend a sequence according to the present invention;

Figure 13 is a schematic diagram showing the sequencing error of screening and detecting whether a base pairing error occurs in a candidate sequence according to the present invention;

14 is another schematic diagram of sequencing errors of a candidate sequence and detecting whether a base pairing error has occurred;

15 is a schematic diagram of sequencing errors of a candidate sequence and detecting whether a base insertion or deletion has occurred; and

Figure 16 is a schematic diagram showing the sequencing error of a candidate sequence and the detection of whether a base insertion or deletion has occurred.

Annex I: References.

10‧‧‧Combination system of gene sequences

11‧‧‧Input interface

110,161‧‧‧Files

111‧‧‧ Index key value data

112‧‧‧Starting the waiting gene sequence

12‧‧‧ indexer

122‧‧‧candidate gene sequence

13‧‧‧Two-way extension combiner

131‧‧‧ index key

132‧‧‧ Fault-tolerant comparison area

14‧‧‧Fault-tolerant sequence filter

141‧‧‧Selected gene sequence

15‧‧‧Overlapping group builder

151‧‧‧ contig

16‧‧‧Output interface

Claims (12)

A gene sequence combination system for splicing a plurality of gene sequences generated by a gene sequencing system to determine a base sequence of a target gene, the system comprising: an indexer for storing the plurality of gene sequences Index value data for finding a candidate gene sequence that may be contiguous on two sides of a waiting gene sequence; a fault-tolerant sequence filter for determining that the candidate gene sequence is contiguous in the candidate gene a selected gene sequence on the two sides of the sequence; and a bidirectional extension combiner for extending the selected gene sequence determined by the fault tolerant sequence filter to at least one side of the sequence of the to-be-connected gene to extend into a longer The gene sequence until the base sequence of the target gene is determined. The gene sequence combination system of claim 1, wherein the bidirectional extension combiner comprises two extension test windows respectively corresponding to two sides of the to-be-connected gene sequence, wherein the two extension test windows are respectively searched from the indexer The gene sequence that can be attached to the extension test window serves as the candidate gene sequence. The gene sequence combination system according to claim 1, wherein the fault-tolerant sequence filter detects whether there is a sequence error of gene sequencing from the candidate gene sequence. The gene sequence combination system according to claim 3, wherein the sequence error detected by the fault-tolerant sequence filter includes an error in base pairing error and an error in base insertion or deletion. The gene sequence combination system according to claim 4, wherein the detection of the error of the base pairing error is performed by arranging a plurality of the candidate gene sequences according to possible extension positions, and counting the bases of the same position after the different sequence alignment To determine whether it is a sequence error resulting from sequencing or that the candidate gene sequence is not a sequence attached to this position. The gene sequence combination system according to claim 4, wherein the detection of the erroneous detection of the base insertion or deletion is performed by converting the originally aligned sequence patterns into the differential sequence patterns. The gene sequence combination system of claim 6, wherein the differential sequence type is to represent the same contiguous base in the gene sequence as a single base. The gene sequence combination system according to claim 4, wherein when the fault-tolerant sequence filter performs the detection of the error of the base insertion or deletion, when the difference between the selected gene sequence and the extended test window is successful, The excess base inserted in the selected gene sequence or the deleted missing base will be replaced by the correct number of bases. The gene sequence combination system according to claim 1, wherein the indexer uses a partial fragment of the gene sequence as an index key to classify all the gene sequences, and when combined, according to the index The key value is used to retrieve the gene sequence for use. A gene sequence combination method for splicing a plurality of gene sequences generated by a gene sequencing system to determine a base sequence of a target gene, comprising: step (A) providing the system according to claim 1; Step (B) inputting the plurality of gene sequences and indexing, storing the index value data in the indexer; step (C) generating a candidate gene sequence from the unused sequence; and step (D) The extension combiner takes out a predetermined base length on each side of the sequence of the to-be-connected gene as a left extension test window and a right extension test window, and respectively shifts the left extension test window to the left and a right shift by a predetermined length. Extending the test window to the right, dividing the gene sequence in the left extension test window and the right extension test window after each displacement into a new index key and a fault tolerance comparison area, and querying the indexer with the new index key a possible candidate gene sequence; step (E) determining, by the fault-tolerant sequence filter, the candidate gene sequence is a selected gene sequence contiguous on two sides of the sequence of the to-be-connected gene; step (F) The stretcher combines the selected gene sequence determined by the fault-tolerant sequence filter to extend to a longer gene sequence on at least one side of the sequence of the to-be-connected gene, and when the extension is successful, repeats step (D) to ( F) until the two sides of the sequence of the to-be-connected gene are unable to continue; and all the selected extensible selected gene sequences are overlapped into a contig according to their positions, and the most likely base at each position of the contig is output The base sequence of the target gene is formed. The gene sequence combination method according to claim 10, wherein the step (B) inputs the plurality of gene sequences via an input interface, assigns a number to each of the gene sequences, and establishes a left index structure and a right index of the gene sequence. structure. The gene sequence combination method according to claim 10, wherein the step (C) uses a plurality of consecutive adjacent gene sequences as the starting candidate gene sequence fragment, and the consecutive adjacent plurality of gene sequences are used. The index key of the displacement finds a gene sequence that is identical to each other and overlaps into the fragment of the starting gene sequence.