WO2012171213A1 - Method and system for assembly of genome - Google Patents

Abstract

The present invention provides a method and a system for assembly of genome. The method comprises the steps of shearing the genome to Fosmid clones using Fosmid as the carrier; dividing the Fosmid clones into various pooling libraries, shearing the Fosmid clones of each pooling library and sequencing, getting the reads and performing short sequence assembly to obtain the scaffold of each pooling library, then removing the invalid bases from the scaffold to get the scaftig of each pooling library; splicing the scaftigs to get the scaffold of the genome based on the results of the scaftig alignment, in which the results of the repeatitive sequences alignment have been masked. The method and system for assembly of genome provided by the present invention combines whole-genome shotgun sequencing with Fosmid to Fosmid for sequencing, and hierarchical assembly has also been used to improve the accuracy of the assembly.

Description

 Genomic assembly method and system

 The present invention relates to the field of bioinformatics technology, and more particularly to a genome assembly method and system. Background technique

 At present, the genome assembly project uses the Whole Genome Shotgun sequencing (WGS) as the mainstream design scheme. It mainly uses the DNA inserts of different lengths for double-end sequencing according to the specific characteristics of the genome repeat sequences. The accuracy of the single base and the integrity of the genome can be guaranteed if the average sequencing depth of the genome is sufficient. With the maturity and popularity of Next-generation sequencing (NGS), sequencing costs are greatly reduced, and genome-wide shotgun sequencing based on second-generation sequencing technology has become the mainstream of sequencing for various genome projects.

 However, for complex genomes, it has high heterozygosity (heterozygous, that is, a state in which different alleles exist at one or more sites on a homologous chromosome) and various problems such as repetitive sequences, the above solution Susceptible to these problems, the assembly results can not meet the standard, resulting in data analysis and difficult to install, does not apply to complex genomes. Summary of the invention

 One technical problem to be solved in one aspect is to provide a genomic assembly method and system that can improve the accuracy of assembly.

According to an aspect of the present invention, a genomic assembly method is provided, comprising: disrupting a genome into a Fosmid clone sequence using Fosmid as a vector; dividing a Fosmid clone sequence into each pooling library, and performing each of the sample libraries Sequencing; disrupting and sequencing the Fosmid clone sequence of each mixed sample library Reading short sequence assembly to obtain the skeleton sequence of each mixed sample library; removing the invalid bases in the skeleton sequence to obtain the skeleton shearing contig sequence of each mixed sample library; splicing the skeleton shear based on the alignment result between the skeleton cut contig sequences The contig sequence is cut to obtain the skeletal sequence of the genome, wherein the alignment results in the region of the repeat sequence are masked.

 According to an embodiment of the method of the present invention, the skeletal-shear contig sequence based on the alignment result of the skeletal-shear contig sequence obtains the skeletal sequence of the genomic sequence, wherein the alignment result in the masked repeat region includes: The skeleton cut contig sequence is subjected to pairwise alignment; the repeat sequence region is determined according to the alignment result, and the alignment result in the repeat region region is masked; the skeleton cut contig sequence is spliced based on the alignment result to obtain the skeleton sequence of the genome.

 According to an embodiment of the method of the present invention, the method further comprises: converting the invalid ^^ in the skeleton sequence of the genome to be valid by using the pair relationship between the readings.

 According to an embodiment of the method of the present invention, the disrupting the genome into a Fosmid clone sequence using Fosmid as a vector comprises: randomly breaking the genome into a Fosmid clone sequence using Fosmid as a vector, the total length of the Fosmid clone sequence being several times greater than The length of the genome.

 According to one embodiment of the method of the invention, each sample library comprises from 10 to 90 Fosmid clone sequences, preferably from 30 to 90 Fosmid clone sequences.

According to an embodiment of the method of the present invention, the short sequence assembly of the sequencing sequences of each mixed sample library obtains the skeleton sequences of each mixed sample library including: For each mixed sample library: the sequencing sequence is sequentially cut out to have a length of K Short sequence K-mer; store the K-mer into the hash table to form the vertices of the De Bruin diagram; the successive K-mers on the sequencing sequence are connected to form the edge of the De Bruin diagram; all sequencing sequences are All processed to obtain the entire De Bruin diagram; remove the path caused by sequencing errors and heterozygous sites in the De Bruin diagram; connect the linear K-mer paths to form the first-level contig; Comparing to the first-level contig sequence, the relative position and orientation relationship between the contig sequence are established according to the inter-read relationship, and the skeleton sequence of each mixed library is formed. According to one embodiment of the method of the present invention, multiple sets of different K values are used for each mixed sample library, and contigs and skeleton sequences under different K values are generated for each mixed sample library; for each: library, from different K Among the assembly results of the values, the contig of N50 and the skeleton sequence are picked.

 According to an embodiment of the method of the present invention, the skeleton cut contig sequence of each mixed sample library is spliced based on the comparison result to obtain a skeleton sequence of the genome, including: clustering the skeleton cut contigs based on the comparison result to obtain a contig Clustering; Sorting the sequences in the contig cluster according to the context, and outputting a uniform contig sequence; using the inter-read pair relationship to connect the contig sequence to the skeleton sequence.

 According to another aspect of the present invention, a genome assembly system is provided, comprising: a sample-mixing library generating unit, configured to break a genome into a Fosmid clone sequence using Fosmid as a vector, and divide the Fosmid clone sequence into each mixed sample library (ooling Library); a sequencing unit for interrupting and sequencing the Fosmid clone sequence of each sample library; a first-stage assembly unit for short-sequence assembly of the obtained reads of each sample library to obtain a skeleton of each sample library Sequence; a cut contig acquisition unit for removing invalid bases in the skeleton sequence to obtain a skeleton cut contig sequence of each sample library; a second assembly unit for alignment between skeletal cut contig sequences As a result, the skeletal splicing contig sequence was obtained to obtain the skeletal sequence of the genome, and the alignment result in the region of the repeat sequence was masked.

 According to an embodiment of the system of the present invention, the secondary assembly unit comprises: a comparison module, configured to perform a pairwise alignment of all skeleton shearing contig sequences; and a sequence processing module for determining a repetition according to the comparison result The sequence region, the alignment result in the repeated sequence region is masked; the splicing module is configured to splicing the skeleton cut contig sequence based on the comparison result to obtain the skeleton sequence of the genome.

According to an embodiment of the system of the present invention, the system further comprises: a hole filling unit, configured to convert the invalidity in the skeleton sequence of the genome into having the pair relationship between the readings According to one embodiment of the system of the invention, the total length of the Fosmid cloned sequence is several times greater than the length of the genome; or each mixed sample library comprises 10 to 90 Fosmid cloned sequences.

 According to an embodiment of the system of the present invention, the primary assembly unit pairs each sample library: the sequencing sequence is sequentially cut out of a short sequence K-mer of length K; the K-mer is stored in a hash table to form a German The apex of the Bruin diagram; the successive K-mers on the sequencing sequence are connected to form the edge of the De Bruin diagram; all the sequencing sequences are processed to obtain the entire De Bruin diagram; Paths caused by sequencing errors, heterozygous sites; linear K-mer paths are joined to form a first-order contig; sequenced sequences are aligned to the first-order contig sequence, based on the inter-read relationship The relative position and orientation relationship between the contig sequence forms the skeleton sequence of each mixed sample library.

 According to an embodiment of the system of the present invention, the first-level assembly unit adopts multiple sets of different K values for each mixed sample library, and the ^^ mixed sample libraries generate overlapping groups and skeleton sequences under different K values; The mixed sample library picks out the contigs and skeleton sequences of N50 from the assembly results of different K values.

 According to an embodiment of the system of the present invention, the splicing module clusters the skeleton cut contigs based on the comparison result to obtain contig group clustering; sorts the sequences in the contig cluster cluster according to the context, and outputs a unified contig Sequence; contig sequence sequences are joined as skeleton sequences using inter-read pair relationships.

 The genomic assembly method and system provided by the invention adopts the method of combining whole genome shotgun and Fosmid to Fosmid for sequencing, and performs hierarchical assembly, thereby improving assembly accuracy. DRAWINGS

1 is a view showing an embodiment of the genome assembly method of the present invention; FIG. 2 is a flow chart showing another embodiment of the genome assembly method of the present invention; and FIG. 3 is a view showing another embodiment of the genome assembly method of the present invention. Flow chart of an embodiment; Figure 4 shows an example of the alignment result of two sequences;

 Figure 5 shows an example of a repeat sequence in a sequence and its alignment result;

 Figure 6 shows an example of determining a re-arranged area;

 Figure 7 is a diagram showing a masking strategy and a masking strategy before modification; Figure 8 is a structural diagram showing one embodiment of the genome assembly system of the present invention; Figure 9 is a diagram showing a genome assembly system of the present invention. A partial structural view of an embodiment. detailed description

 The invention will now be described more fully hereinafter with reference to the accompanying drawings

 In one embodiment of the present invention, a hierarchical assembly method and system based on second-generation sequencing technology is provided, which is performed by a genome-wide shotgun method combined with Fosmid to Fosmid for sequencing of complex genomes. .

 Figure 1 shows a 3⁄4⁄2 diagram of one embodiment of the genome assembly method of the present invention. As shown in Figure 1, in step 102, the genome is disrupted to Fosmid cloned sequence using Fosmid as a vector.

 Step 104: Divide the Fosmid clone sequence into each pooling library, and sequence each mixed sample library. For example, a mixture of Fosmid clone sequences from a sample library is broken into small fragments for double-end sequencing.

 Step 106: Perform short sequence assembly on the reading of each mixed sample library to obtain a skeleton sequence of each mixed sample library;

Step 108: Remove the invalid bases in the skeleton sequence to obtain the skeleton cut contig sequence of each mixed sample library. For example, suppose the skeleton sequence is "ATTGCNNNGGAC", where N represents an invalid base, and the corresponding skeleton-cut contig includes "ATTGC" and "GGAC" ₀

Step 110, splicing the skeleton shear based on the alignment result between the skeleton cut contig sequences The contig sequence is cut to obtain the skeletal sequence of the genome, wherein the alignment result in the region of the repeat sequence is masked. All skeleton cut contig sequences are subjected to pairwise alignment; the repeat region is determined according to the alignment result, the alignment result in the repeat region is masked, and then the skeleton cut contig sequence is spliced based on the alignment result to obtain the genome Skeleton sequence.

 The genomic assembly method provided by the invention breaks the genome into Fosmid DNA fragments and divides them into a mixed sample library, and performs hierarchical assembly after sequencing, thereby improving assembly accuracy.

 Figure 2 is a flow chart showing another embodiment of the genome assembly method of the present invention. As shown in Fig. 2, in step 202, the whole genome is randomly interrupted into segments of a predetermined length, for example, 32 Kb.

 Step 204, constructing a Fosmid library.

 In step 206, all Fosmid libraries are randomly combined to form a mixed sample library, for example, in groups of 30 to 90, and put together to form a mixed sample library.

 Step 208: Randomly interrupt the small segment in the inserted sample in the mixed sample library, and insert a small segment into the library. For example, the length of a small segment is 300 bp to 500 bp.

 Step 210: Sequencing small fragments of each mixed sample library to obtain sequencing data in units of mixed sample libraries.

 Step 212, - level assembly: The inserts in the mixed sample library are regarded as virtual genomes, and the first-level assembly is performed in units of each mixed sample library. The assembly at this stage uses the splicing software SOAPdenovo, which is short based on the De Bruin map. Sequence assembly. A longer Contig (contig) sequence that does not include invalid bases is assembled by PE reads, and then assembled by Contig to obtain longer scaffolds (skeletal sequences) that may include invalid bases, and remove all invalid bases in scaffold. , get the scaftig (skeleton cut contig) sequence between Contig and scaffold. Put all the scaftigs obtained from the sample library together for secondary assembly.

Step 214, secondary assembly: mainly using a splicing method based on comparison, using one All the sample libraries obtained from the stage assembly are scaftig, put them together, compare them in pairs, assemble the second-level Contig, and even the secondary scaffold.

 Figure 3 is a flow chart showing still another embodiment of the genome assembly method of the present invention. As shown in Figure 3, in step 302, a sample library consisting of Fosmid clone sequences is sequenced to obtain reads.

 According to one embodiment of the present invention, the Fosmid to Fosmid protocol is used to randomly break the whole genome with Fosmid as a Fosmid clone sequence of similar length to Fosmid. The total length of all Fosmid clones is several times longer than the whole genome, 1⁄2 length, to ensure coverage of each 1⁄2 of the genome as much as possible. For example, according to the Lander-Waterman model (Reference Genomic Mapping by Fingerprinting Random Clones: A Mathematical Analysis* ERIC S.

LANDER AND MICHAEL S. WATERMAN, Genomics. 1988 Apr;2(3):231-9. ), for Fosmid with a length of 35kbp, 8 times the depth of genome coverage can theoretically ensure that each base on the genome is minimally covered. once.

 For example, 10~90 Fosmid clone sequences were used as an ooling library. After the mixed sample library was established, the Fosmid clone sequence was pooled and sequenced to obtain a double-end sequencing sequence, ie read. For example, high-throughput sequencing can be performed using Illumina GA Solexa sequencing technology. Solexa sequencing technology belongs to next-generation sequencing technology (second generation). Its core idea is sequencing by synthesis or ligation, SBS & SbL. Bridge PCR is performed on small cells by using single molecule arrays. reaction. The new reversible blocking technique allows for the synthesis of only one base at a time, without the need to label fluorophores, and then captures the light using the corresponding laser fluorophores to read the information.

 Step 304, based on the assembly of the De Bruin diagram. After obtaining the sequencing sequence of each sample library, each sample library can be spliced. This stage is called first stage assembly.

The read (sequencing sequence) is sequentially subjected to a short sequence of length K, called K-mer, and K-mer overlaps K-1 bases before and after. Store K-mer to hash table In the middle, the vertices of the De Bruin diagram are formed; the K-mers that are successively read and read think that the two K-mers are connected to form the edge of the De Bruin diagram. After all the readings have been processed, the entire De Bruin diagram can be obtained. The paths caused by sequencing errors and heterozygous sites are removed, and the linear K-mer paths are connected to form the first-level contigs. . These K-mer bases are joined to form a first-order contig sequence. The readings are then compared to the contig sequence, and the relative position and orientation relationship between the contig sequence is established according to the paired end of the reading, thereby forming a skeleton sequence of the first level. Finally, use the read pair relationship to make a hole.

 For example, the assembly at this stage can be spliced using the mosaic software SOAPdenovo, which was independently developed by the Huada Gene Research Institute. The short sequence assembly based on the De Bruin diagram is used to obtain the first-level skeleton sequence ( scaffold ) of each sample library. Assembly software reference literature Li, R. et al. De novo assembly of human genomes with massively parallel short read sequencing. Genome Res (2009). This software is available free of charge on the web at http://soap.genomics.org.cn/soapdenovo.htmlo

 Step 306, obtaining a skeleton cut contig. When the skeleton sequence is assembled in the first stage, a part of the repeating sequence base cannot be assembled to the skeleton sequence. For each sample library, after the first stage assembly is completed, all invalid bases N of the skeleton sequence are removed, and the skeleton cut contig (scaftig) of the sample library is obtained.

 Using the data of each sample library, the skeleton-shear contig sequence assembly in the second-stage assembly is performed. This assembly mainly uses a splicing method based on alignment. The second level of overlapping group sequence assembly process specific process includes:

 Step 308, the skeleton-cut contig sequence pair-wise alignment:

The relationship between the sequences is established by comparison. The skeleton cut contigs are pairwise and the two skeleton cut contig sequences of each comparison result can be calculated as a similarity score, and the results are sorted in descending order. At the same time for each skeleton shear The contig sequence is cut to create a list to record all other skeletal cut contig sequences that are aligned with it. For example, the skeleton cut contig sequence a has an alignment relationship with b, c, and a 2 (the number of comparison relationships) b, c is recorded in the list.

 The sequence alignment indicator indicates that the two skeleton-shear contig sequences have the same base. Assume that sequence A is ATTTCGTGA and B is TTGGTGACTG. The comparison result can be expressed as:

 ATTTCGTGA TTGGTGACTG

 The underlined bases are bases that are not aligned, and the remaining bases are the aligned ones. The alignment result is, for example, as shown in Fig. 4, the vertical line portion is the portion on the alignment, and the oblique line portion is the portion on the unaligned.

 Step 310, Repeat boundary detection There is a repetitive sequence in the genome, and the same part of the comparison result may be misspelled if it contains a repeat sequence. Taking Figure 5 as an example, b is a repeated sequence. If the boundary of b is not distinguished, an error of a-b-e or d-b-c may occur in the splicing result. The results were screened to eliminate the alignment in the repeat region to prevent misspelling.

A repeating region in the skeleton cut contig sequence is determined by a branch detected by an overlap relationship. Using the characteristics of the non-repetitive sequence region on both sides of the repeated sequence region (ie, a and d are different in FIG. 5, c and e are not the same), and the inter-sequence positional relationship can be determined by the correlation between the inter-sequence alignment results, and Find most of the potential repeat region. Thereby, the alignment result of the same base completely in the repeat region can be masked. Taking FIG. 6 as an example, the alignment relationship of the sequences A, B, and C is as shown in the figure, wherein the lattice area represents the area on the two sequence alignments, and the oblique line area represents the area on the unaligned. The results of the comparison between A, B and C indicate that the relative positions between A, B and C are as shown in Fig. 6. Suppose you need to judge whether the comparison between A and B can be used for subsequent splicing. You can use the comparison between A and C and between B and C. Compare the results. As shown in Figure 6, the front end of A is not aligned with the front end sequence of C. It can be considered that the boundary of the repeated sequence is on the boundary of the comparison result of A and C; the right end B is similar to the case of C, and the repeat sequence can be considered. The boundary at the right end is at the boundary of the alignment result of B and C. By creating such a partial map, the repeat sequence boundaries (the areas indicated by the two vertical dashed lines in Figure 6) can be discerned. For A, B, which is completely within the boundaries of the repeat sequence, it is filtered and not used in subsequent stitching.

 Step 312, Contig assembly:

 Establish a comparison set that constitutes a contig, use a certain scoring strategy to treat each skeletal sequence as a contig, and cluster the contigs by comparison. One alignment can splicing two contigs into one Larger contigs, after continuous clustering, generate large contigs, ie contig clusters.

 This step establishes a set of comparison results that make up the contig. Each alignment result is scored according to the same base ratio (similarity) between sequences, and the higher the number of identical bases, the higher the score. The alignment results are sorted from large to small, and the alignment results and sequences with high similarity and inconsistent sequence ratio below a certain threshold are merged into the contig cluster structure.

 Step 314, Consensus:

 According to the comparison result set generated in the above step, the inconsistency between the bases is removed, and the contig sequence is generated. Read the contig clustering result of the previous step. For each contig cluster, the corresponding sequence can be known. The sequences in the contig cluster are sorted according to the context, and all sequences can be synthesized according to the inter-sequence alignment. For a uniform sequence. The unified sequence output at this time is the second-order contig sequence.

 Steps 316 and 318, shielding part of the sequence base, and constructing a skeleton sequence. After the overlapping group sequence is obtained, the second-level scaffold splicing can be performed, and the contigs are connected into a skeleton sequence by using a paired-end relationship. For example, this stage can be achieved using the mosaic software SOAPdenovo developed by the Huada Gene Research Institute.

In addition, the repeat sequence masking strategy in this method can be based on complex genomes. The situation is improved by masking a sequence and its associated connections to mask one of the connections, thereby extending the skeleton sequence. The original shielding strategy is shown in Figure 7 (A). Once a contig that is connected upstream or downstream of a contig (contig 1) (as illustrated in Figure 7 downstream) is found to overlap in position (such as The contigs 2, 3 in Fig. 7 calculate the distance between the contigs by reading the inter-read relationship and reading the insert size of the corresponding library, by which the position overlap can be calculated. This contig 1 and its associated connections (as shown) are then masked. The strategy of repairing (as shown in Fig. 7(B)) is that only the farthest contigs that extend the path and their connections are retained after the position overlap occurs (as in the case of the figure, it is assumed that the contig 2 is not connected subsequently, The contig 3 can be connected further. At this time, the connection of the contig 1 to the contig 2 is selected.

 Compared to the original version, using the modified version of SOAPdenovo can increase the skeleton sequence N50 by about 10%.

 Step 320, fill the hole. After the skeleton sequence is completed, the inter-read pair relationship can be used to fill the invalid base N in the skeleton sequence. For example, the software KGF developed by the Huada Gene Research Institute can be used to fill the hole, and the SOAPdenovo supplement hole software can also be used. GapCloser does this work, GapCloser is available for free at soap.genomics.org.cn.

 A specific application example of the method of the present invention is provided below. In this example, the oyster genome sequencing assembly is implemented. The specific steps are as follows:

 (1) Fosmid mixed sample data processing

 1), remove the linker sequence in the original data (Raw Data) to obtain the sequence file after removing the linker sequence.

 2) Calculate the yield of bases in the sequence of the above-mentioned removal of the linker, discard the sample library with too low output, and reduce the influence of the low-volume sample-mixing library on the assembly result.

3), the read and reference sequences after the joint is removed are compared with the short sequence alignment software SOAPaligner, and the corresponding pair and single alignment results are obtained. Reference Li, R. et al. SOAP2: an improved ultrafast tool for short read alignment. Bioinformatics (2009). This software is available for free from the Internet at http://soap.genomics.org.cn/soapaligner.html o The reference sequence can be a genome sequence obtained using the genome-wide shotgun method.

 4), from the previous step SOAPaligner aligns the results to calculate the actual insertion fragment length of the sequence, and generates an insert length file for subsequent assembly to build the assembly configuration file.

 5), separately construct the assembly configuration file from each of the sample library sequence insert lengths obtained in the previous step. Constructing an assembly profile separately for each sample library reduces the variance of the configured insertion segment length and the actual insertion segment length to avoid the impact on the accuracy of the assembly results due to excessive variance.

 6), using the configuration file generated in the previous step, each mixed sample library is assembled separately with SOAPdenovo software. Each assembly is assembled: the library uses multiple sets of different k values, and ^^ mixed sample libraries generate different k values. Assembly results for a series of contigs and skeleton sequences.

 7) For each sample library, pick the N50 best contigs and skeleton sequences from the assembly results of different k values, which will be used for subsequent processing.

 8), the skeleton sequence of each sample library generated in the previous step is filled with KGF software to generate a new skeleton sequence after the hole is filled.

 9), correct the skeleton sequence of each mixed sample library after the hole is generated in the previous step, and generate the corrected skeleton sequence.

 10), the corrected skeleton sequence generated by each of the sample libraries in the previous step is scored by qualScore software, and the score of each base of the corrected skeleton sequence is obtained, and the ^^ score file is obtained.

11), the skeleton sequence corrected by all the sample libraries and the corresponding base score results are integrated. A backbone sequence of about 5.3 G in size can be obtained with an N50 of about llkbp. 12): All the invalid bases N of the sequence are removed, and a 5.3G skeleton-shear contig is obtained, and the N50 is about 5 kbp.

 (2) splicing contig sequence

 1) The skeleton cut contig sequence set of the first-stage spliced output is used as the input sequence conversion ID

 2) Establish a suffix tree index on the input sequence. This step will facilitate the next seed comparison, and the sequence can be quickly indexed to quickly find the corresponding subsequence.

 3) Seed comparison of the input sequences. This step will perform a rapid seed alignment, selecting similar, shorter-length subsequences between the two sequences to determine a certain similarity between the two sequences, thus screening for possible alignment results. The sequence pairs are handed to the exact alignment of the next step.

 4) Establish a sequence search library for the input sequence. The sequence created in this step is mainly for quickly finding sequences and compressing sequences.

 5) Align the sequences, this step will be an exact alignment to establish the alignment between the sequences.

 6) Detect the repeat region. This step will create a series of subgraphs based on the comparison results from the previous step to identify the repeat region and remove the alignment results completely in the repeat region.

 7) Score each alignment result according to the same base ratio between sequences. The higher the score, the higher the score. The alignment results are sorted from large to small, and the alignment results and sequences with high similarity and inconsistent sequence ratio below a certain threshold are merged into the contig structure.

 8) Remove base inconsistency and generate a second-level contig sequence.

 Obtain a contig file of approximately 656M in size, with an n50 of approximately 9kbp.

 (3) splicing the second-level skeleton sequence

1) Skeleton sequence stitching data preparation, this step mainly generates some preliminary files for SOAPde. use. 2) The skeleton sequence is spliced, and the second-level skeleton sequence is established according to the paired-end information. First, the readings of different inserts are compared to the contig sequence, and then the anterior-post relations between the contig sequences are determined according to the double-end read information on both contig sequences, and the skeleton sequences are arranged.

 A skeleton sequence file of approximately 777M in size can be obtained, with an n50 of approximately 46174 bp.

3) Filling the hole, using only one end or only a part of the double-ended reading is aligned to the contig, and the other end can be positioned in the "N" region of the skeleton sequence according to the size of the inserted segment, so that the skeleton sequence can be Invalid bases are converted to valid bases.

 At this point, the final assembly result is obtained.

 The genomic assembly method provided by the embodiments of the invention has the advantages of high genome coverage, long N50 splicing sequence and high accuracy. For example, in the above application example, about 99.2% of the assembly results can be read and compared, and the total length of the assembly results is very close to the estimated genome length with high accuracy. The oyster mosaic result is about 850 Mbp, the genome is estimated to be 800 Mbp, and the genome coverage is high.

 The assembled contigs or skeleton sequences are arranged from large to small. When the cumulative length just exceeds 50% of the total length of the entire assembly sequence, the size of the last contig or skeleton sequence is N50, and the N50 is sequenced. The integrity of the importance is important.

Oyster genome sequencing projects, the second stage of the contig sequence and the second stage resulting framework sequences length N50 higher than the backbone sequence contig sequence and shotgun obtained, the other points ¹ J reaches 9kb and 50kbp _o

Figure 8 is a block diagram showing one embodiment of the genome assembly system of the present invention. As shown in FIG. 8, the system 800 includes: a mixed sample generating unit 81, which interrupts the genome into a Fosmid clone sequence using Fosmid as a carrier, and divides the Fosmid clone sequence into each pooling library; the sequencing unit 82, The Fosmid clone sequence of each sample library is interrupted and sequenced; a first assembly unit 83 is configured to perform short sequence assembly on the obtained reads of each sample library to obtain a skeleton sequence of each sample library; The contig group obtaining unit 84 removes the invalid bases in the skeleton sequence to obtain the skeleton cut contig sequence of each mixed sample library; the second assembly unit 85, splicing the skeleton shear based on the alignment between the skeletal cut contig sequences The contig sequence obtains the skeletal sequence of the genome in which the alignment results in the region of the repeat sequence are masked. Wherein, the total length of the Fosmid cloned sequence is several times longer than the length of the genome; each mixed sample library may include 10 to 90 Fosmid clone sequences.

 According to one embodiment of the present invention, a primary assembly unit for each sample library: the sequencing sequence is sequentially cut out of a short sequence K-mer of length K; the K-mer is stored in a hash table to form Debruin The vertices of the graph; the successive K-mers on the sequencing sequence are connected to form the edge of the de Bruin diagram; all the sequencing sequences are processed to obtain the entire De Bruin diagram; the de Bruin diagram is removed by sequencing error a path caused by a heterozygous site; a linear K-mer path is joined to form a first-order contig; the sequencing sequence is aligned to the first-order contig sequence, and an contig sequence is established based on the inter-read relationship The relative position and orientation relationship between them forms the skeleton sequence of each mixed sample library. The first-level assembly unit can use multiple sets of different K values for each sample library, and generate contigs and skeleton sequences with different K values for each sample library; for each sample library, assembly from different K values Among the results, a contig of N50 and a skeleton sequence were picked.

Figure 9 is a partial structural view showing an embodiment of the genome assembly system of the present invention. According to an embodiment of the present invention, the secondary assembly unit 95 in the genome assembly system comprises: a comparison module 951 for performing pairwise alignment of all skeleton shearing contig sequences; a sequence processing module 952 for ratiometric The result is determined by determining the repeat region, and the alignment result in the repeated sequence region is masked; the splicing module 953 is configured to splicing the skeleton cut contig sequence based on the comparison result to obtain the skeleton sequence of the genome. According to an embodiment of the invention, the genome assembly system further comprises a hole-filling unit 96 for converting the invalidity in the skeleton sequence of the genome into an effective embodiment of the system according to the invention by using the pairwise relationship between the readings, the splicing module Based on the alignment result, the skeleton The contigs are cut and clustered to obtain contig clustering. The sequences in the contig cluster are sorted according to the context, and a unified contig sequence is output. The contig sequence is connected into the skeleton sequence by using the inter-read pair relationship.

 For the function of the various devices or units in Figures 8 to 9, reference may be made to the description of the corresponding portions in the above embodiments of the method of the present invention, which will not be described in detail herein for the sake of brevity.

 The description of the present invention has been presented for purposes of illustration and description. Many modifications and variations will be apparent to those skilled in the art. The embodiment was chosen and described in order to explain the invention and the embodiments of the invention

Claims

Rights request A method of assembling a genome, comprising: The genome was disrupted to Fosmid cloned sequence using Fosmid as a vector; Divide the Fosmid clone sequence into each sample library; The Fosmid clone sequences of each mixed sample library were interrupted and sequenced, and the obtained reads were subjected to short sequence assembly to obtain skeleton sequences of the respective mixed sample libraries; Removing the invalid bases in the skeleton sequence to obtain the skeleton cut contig sequence of each mixed sample library; The skeletal sequence of the genomic sequence is obtained by splicing the skeletal cut contig sequence based on the alignment between the skeletal cut contig sequences, wherein the alignment results in the repeat region are masked.  2 . The method according to claim 1 , wherein the framing the skeletal contig sequence based on the alignment result between the skeletal shear contig sequences obtains a skeletal sequence of the genomic region, wherein the ratio in the region of the mask repeat region The results include: Performing a pairwise alignment of all skeleton shearing contig sequences; Determining the repeat region based on the alignment result, masking the alignment result in the repeated sequence region; The skeleton sequence of the genome is obtained by splicing the skeleton cut contig sequence based on the alignment result.  3. The method according to claim 1, further comprising: Invalid 1⁄2 in the skeletal sequence of the genome is converted to an effective base using the pair relationship between reads.  4. The method according to claim 1, wherein the disrupting the genome into a Fosmid clone sequence using Fosmid as a vector comprises: The genome was randomly interrupted with Fosmid as a Fosmid clone sequence, and the total length of the Fosmid clone sequence was several times longer than the length of the genome. 5. The method of claim 1 wherein each of the sample libraries comprises 10 ~ 90 Fosmid clone sequences.  The assembly method according to claim 1, wherein the Fosmid clone sequence of each sample library is interrupted and sequenced, and the obtained sequence is assembled by short sequence to obtain a skeleton sequence of each sample library: For each mixed sample library: The sequencing sequence is sequentially cut out of a short sequence K-mer of length K; Store the K-mer in a hash table to form the vertices of the De Bruin diagram; The successive K-mers are ligated in the sequencing sequence to form the edge of the de Bruin diagram; all sequencing sequences are processed to obtain the entire De Bruin diagram; Removal of the St. Bruce caused by sequencing errors and heterozygous sites; Connecting linear K-mer paths to form a first-level contig; The sequencing sequence is aligned to the contig sequence of the first stage, and the relative position and orientation relationship between the overlapping group sequences are established according to the inter-read relationship, and the skeleton sequence of each mixed sample library is formed.  7. The assembly method according to claim 6, wherein Multiple sets of different K values are used for each mixed sample library, and contigs and skeleton sequences under different K values are generated for each mixed sample library; For each sample library, the contigs and skeleton sequences of the N50 were picked from the assembly results of different K values.  8. The assembly method according to claim 2, wherein the skeletal-shear contig sequence of each of the sample-mixing libraries is spliced based on the comparison result to obtain a skeleton sequence of the genome, including: The skeletal cut contigs are clustered based on the comparison result to obtain contig clustering; the sequences in the contig cluster are sorted according to the context, and a unified contig sequence is output; The contig sequence is concatenated into a skeleton sequence using an inter-read pair relationship.  9. A genome assembly system, comprising: Mixed sample generation unit for interrupting the genome to Fosmid using Fosmid as a vector Cloning the sequence, dividing the Fosmid clone sequence into each sample library; a sequencing unit for interrupting and sequencing the Fosmid clone sequence of each sample library; a first-level assembly unit for short-sequence assembly of the read samples of each mixed sample library to obtain a skeleton sequence of each mixed sample library; a cut contig acquisition unit for removing invalid bases in the skeleton sequence to obtain a skeleton cut contig sequence of each sample library; A secondary assembly unit for splicing the skeletal contig sequence based on the alignment between the skeletal shear contig sequences to obtain a skeletal sequence of the genome, wherein the alignment results in the region of the repeat sequence are masked.  10. The system of claim 9, wherein the secondary assembly unit comprises: a comparison module, configured to perform a pairwise alignment of all skeleton cut contig sequences; a repeat sequence processing module, configured to determine a repeat sequence region according to the comparison result, and mask the alignment result in the repeated sequence region; A splicing module for splicing the skeleton cut contig sequence based on the alignment result to obtain a skeletal sequence of the genome.  11. The system according to claim 9, further comprising: a hole-filling unit for converting the invalid base group in the skeleton sequence of the genome into an effective pair by using the pair relationship between the readings  12. The system of claim 9, wherein the total length of the Fosmid clone sequence is several times greater than the length of the genome; Or Each sample library includes 10 to 90 Fosmid clone sequences. 13. The assembly system according to claim 9, wherein the primary assembly unit pairs each of the sample libraries: sequentially sequencing the sequence to extract a short sequence K-mer of length K; storing the K-mer To the scatter table, forming the vertices of the De Bruin diagram; The sequential K-mers are ligated in the sequencing sequence to form the edge of the de Bruin diagram; all the sequencing sequences are processed to obtain the entire De Bruin diagram; the de Bruin diagram is removed by sequencing error, heterozygous a path caused by a point; a linear K-mer ^ St. is joined to form a first-order contig; the sequencing sequence is aligned to the contig sequence of the first level, and a relative contig between the ensemble sequences is established according to the inter-read relationship The position and orientation relationship form the skeleton sequence of each sample library.  14. The assembly system of claim 13 wherein: The first-level assembly unit adopts multiple sets of different K values for each mixed sample library, and generates contigs and skeleton sequences under different K values for each mixed sample library; for each mixed sample library, from different K values The contigs and skeleton sequences of N50 were picked out from the assembly results.  The assembling method according to claim 10, wherein the splicing module clusters the skeleton cut contigs based on the comparison result to obtain contig clustering; and aligns the sequences in the contig cluster according to the context Sorting, outputting a uniform contig sequence; ligating the contig sequence into a skeleton sequence using the inter-read pair relationship.