A multi-configuration assembly method for plant mitochondrial genomes based on third-generation whole genome sequencing data
By combining third-generation whole-genome sequencing data with specific software, the challenge of assembling multiple configurations of the plant mitochondrial genome was solved, achieving efficient and low-cost precise assembly and resolving all possible configurations of the mitochondrial genome.
Patent Information
- Application Number
- CN202211009304.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-08-22
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2042-08-22
AI Technical Summary
Existing technologies are insufficient for effectively resolving the multiple configurations of plant mitochondrial genomes, resulting in fragmented assembly results and the inability to obtain complete mitochondrial genome sequences, thus hindering the development of related research.
Using third-generation whole-genome sequencing data, combined with Newbler v.3.0 assembly software and Bandage software, the assembly process of plant mitochondrial genome polymorphism was constructed by segmenting and correcting the third-generation sequencing data, utilizing the differences in coverage depth of the cell nucleus, chloroplasts and mitochondria, preserving and extending all possible connections until a closed loop is formed.
It achieves low-cost, high-efficiency, and accurate assembly of plant mitochondrial genomes, and can resolve all possible configurations of the mitochondrial genome, reducing the dependence on sample separation and high-depth sequencing.
Smart Images

Figure CN115449543B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of bioinformatics, and more specifically, relates to a method for assembling plant mitochondrial genome multi-configuration based on third-generation whole genome sequencing data. Background Art
[0002] With the rapid development of third-generation single-molecule sequencing technology, more and more plants have completed whole genome sequencing (WGS) and assembly. As of the end of February 2022, the NCBI genome database (https: / / www.ncbi.nlm.nih.gov / genome / ) contains about 2,000 plant nuclear genome sequences and more than 6,500 chloroplast genome sequences, while there are only 437 plant mitochondrial genomes, of which only 323 are terrestrial plants. As the main site of cellular aerobic respiration, plant mitochondria not only provide energy for complex physiological activities of cells, but also widely participate in important growth and development processes of plants. Studies have found that changes in the structure of plant mitochondrial genomes are closely related to important traits such as grain development, cytoplasmic male sterility (CMS) and seed aging. Therefore, accurate analysis of the complex structure of plant mitochondrial genomes and their formation mechanisms will not only help to clarify important scientific issues such as nuclear-cytoplasmic interaction mechanisms and plant evolution, but will also promote CMS sterile lines to play an important role in hybrid seed production and crop hybrid vigor utilization.
[0003] However, the plant mitochondrial genome contains a large number of horizontal gene transfer (HGT) sequences and repetitive sequences, which interfere with sequence extension during the assembly of the mitochondrial genome, resulting in fragmented assembly results and difficulty in obtaining a complete mitochondrial genome sequence. In recent years, the decline in sequencing costs and the advancement of assembly algorithms have spawned a massive amount of plant whole genome data and also promoted the rapid development of plant mitochondrial genome research. The number of plant mitochondrial genome sequences collected by NCBI each year has increased from 14 in 2011 to 70 in 2021, but it still lags far behind plant chloroplast and nuclear genomes. At present, there are two main strategies for assembling plant mitochondrial genomes: (1) mitochondrial DNA is isolated and purified before sequencing and assembling; (2) assembly directly from plant WGS data. However, the above two assembly strategies have a common defect, that is, when encountering multi-branch structures caused by recombination or HGT, Contigs are usually directly interrupted or only the path with the most reads support is selected to continue extending, while ignoring other possible connections. Many published plant mitochondrial genomes have only assembled a single master circle structure and have not resolved all possible configurations of the mitochondrial genome, which has seriously hindered the development of plant mitochondrial structure and functional genomics research.
[0004] In recent years, plant whole genome sequencing data has emerged in a blowout manner. How to obtain and utilize complete mitochondrial genome sequences from massive WGS data has become one of the key research directions of plant mitochondrial genomes in the fields of molecular biology, phylogenetics, and genetics. As a result, a large number of plant organelle genome assembly methods and related software for WGS data have been spawned. By analyzing the plant mitochondrial genomes included in the NCBI organelle database in the past five years (2017-February 2022), we found that there are currently 6 main plant mitochondrial genome assembly software, which are ranked from high to low in terms of frequency of use: SPAdes v.3.0 (48 species), Canu (20), Newbler (11), SMARTdenovo (9), NOVOPlasty (7) and GetOrganelle (7). In addition, a few plants use MIRA, Falcon, HGAP, Flye and other software to complete mitochondrial genome assembly. The following introduces 6 commonly used mitochondrial genome assembly software according to the sequencing platform used for sequencing data:
[0005] (1) Roche 454 sequencing platform: Newbler is an assembly software for Roche 454 sequencing data. With the update and iteration of sequencing technology, the 454 sequencing platform was shut down in 2016. Currently, most of the 454 data used by researchers to assemble mitochondrial genomes are sequenced before 2015.
[0006] (2) Illumina sequencing platform: SPAdes v.3.0 is generally only used to assemble Illumina sequencing data of isolated and purified plant mitochondria. However, the operation of isolating mitochondrial DNA is cumbersome and costly, and is not suitable for species that have completed WGS sequencing. NOVOPlasty and GetOrganelle are two assembly software specifically used to assemble organelle genomes from second-generation WGS data. Their assembly principles are similar. Both require a "seed" sequence from the target genome before assembly. The software then uses the "seed" to capture reads from mitochondria from WGS data, and continuously extends the captured reads in both directions until all mitochondrial reads are captured. Then SOAPdenovo2 or SPAdes v.3.0 is called for assembly. Finally, the contigs belonging to mitochondria are screened based on the assembly map information and contigs coverage information. A large number of studies have shown that although GetOrganelle and NOVOPlasty are extremely effective in assembling animal mitochondrial genomes and plant chloroplast genomes, due to the complex structure of plant mitochondrial genomes and the large number of repetitive sequences, these two software often produce many scattered Contigs fragments when assembling plant mitochondrial genomes. Additional Mate-pair sequencing or third-generation long-read sequencing is required to connect the Contigs fragments in order to obtain a complete mitochondrial genome sequence.
[0007] (3) PacBio or Nanopore sequencing platform: Canu and SMARTdenovo are two assembly softwares mainly used for nuclear genome assembly. Since WGS data also contain mitochondrial reads, the contigs belonging to mitochondria can be screened out from the nuclear genome assembly results based on the sequence alignment of closely related species. Finally, the contigs are extended using the original sequencing reads or Sanger sequencing to obtain the complete mitochondrial genome sequence. Although the third-generation sequencing reads are very long and can easily cross most repetitive sequences, when assembling the multi-branched structure of mitochondria, these two softwares only extend the path with the highest read support, while ignoring other possible paths. It is impossible to obtain all the sub-rings and isomers of the mitochondrial genome, and additional experiments or sequence alignments are still required to determine. In addition, due to the high error rate of the third-generation sequencing, in order to obtain accurate assembly results, high-depth sequencing is often required for self-correction, or Illumina sequencing data is used to assist in error correction. This is very costly for species that have not completed whole genome sequencing projects, especially gymnosperms with large nuclear genomes, and is not suitable for generalization to all plants.
[0008] In summary, the current research on plant mitochondrial genomics has the following main defects: (1) Plant mitochondrial DNA isolation and purification kits are expensive, have high requirements on sample quality and quantity, and require frequent manual operations, which is time-consuming and labor-intensive; (2) Assembly software based on Illumina sequencing data requires the mitochondrial genome sequence of closely related species as a reference, and cannot directly cross long repetitive sequences by extending short sequences, requiring other mate pair sequencing or long read sequencing to complete the assembly; (3) Due to the high error rate of single-molecule sequencing, assembly software based on PacBio / Nanopore sequencing data often requires high-depth sequencing or combined with Illumina sequencing to obtain a complete mitochondrial genome, which is costly; (4) When assembling recombinants mediated by repetitive sequences, existing assembly software usually directly interrupts contigs or only extends the path with the most read support, and cannot directly obtain the multi-ring and sub-ring structures of mitochondria. The most fundamental reason for the above problems is the lack of effective multi-configuration assembly methods for plant mitochondrial genomes, which seriously restricts the development of research based on plant mitochondrial genomes, such as phylogenetics, epigenomics (RNA editing, molecular regulatory mechanisms), and genetic engineering (cell fertility research, gene editing, and transgenics).
[0009] Related references and patents:
[0010] 1.Bi C, Qu Y, Hou J, et al.Deciphering the Multi-Chromosomal Mitochondrial Genome of Populus simonii[J].Frontiers in Plant Science, 1632.
[0011] 2. CN109411014B A method for assembling the whole genome of plant chloroplasts into circles based on second-generation sequencing
[0012] 3. CN112481413B Plant mitochondrial genome assembly method based on second-generation and third-generation sequencing technologies Summary of the invention
[0013] In view of the above problems existing in the prior art, the technical problem to be solved by the present invention is to provide a method for assembling multi-configuration of plant mitochondrial genome based on third-generation whole genome sequencing data, which can analyze all possible configurations of the mitochondrial genome.
[0014] In order to solve the above technical problems, the technical solution adopted by the present invention is as follows:
[0015] First, the plant samples were subjected to third-generation sequencing, where the data obtained by PacBio CLR and Nanopore sequencing had a high error rate (accuracy ~85%), while the data obtained by PacBio HiFi sequencing had a high accuracy (accuracy >99%).
[0016] For PacBio CLR and Nanopore sequencing data with higher error rates, it is necessary to use Illumina data of the same sample or its own third-generation high-depth data for error correction. The corrected data can be used for subsequent assembly. For PacBio HiFi sequencing data with higher accuracy, it can be used directly for subsequent assembly without error correction.
[0017] Newbler v.3.0 assembly software can retain the characteristics of all connections of reads and assemble the sequencing data obtained above, but this software is only suitable for 454 sequencing data. When there are reads of more than 30kb in the input third-generation sequencing data, the software cannot run. Therefore, before assembly, reads larger than 30kb in the third-generation sequencing data after error correction are split into different shift lengths (5k, 10k, 15k, 20k, 25k), and the split reads are renumbered and output to different files respectively.
[0018] Subsequently, Newbler v.3.0 was used to assemble the segmented original reads, and minimap2 was called to align the original reads to the Contigs obtained by the preliminary assembly. Samtools depth was used to obtain the average coverage depth of each Contig.
[0019] Then, the R language was called to draw the average sequencing depth distribution of Contigs after different reads were assembled. According to the different coverage depths of Contigs sequences of the nucleus, chloroplasts and mitochondria, the Contigs with longer mitochondrial Contig sequences were selected as seeds, and the connections between Contigs contained in the 454ContigGraph.txt file assembled by Newbler v.3.0 were extended. When the seed extension encountered multiple branches caused by repeated sequences, all possible connections were retained and continued to extend until a closed loop was formed or all existing plant mitochondrial protein-coding genes were included. Among them, the existing plant mitochondrial protein-coding gene library was downloaded from the NCBI organelle database, including protein-coding genes in all plant mitochondrial genomes, which was automatically updated once a week using Python crawlers.
[0020] If a closed circle is not formed or common protein-coding genes are not included after extending all available Contigs, it is necessary to merge the sequencing data with different synchronization cutoffs separately, assemble them again using Newbler v.3.0, and then repeat the above steps until the conditions are met.
[0021] During the extension process, the sequencing depth, length and connection relationship between different contigs were recorded using Perl scripts to obtain a Graphical Fragment Assembly (GFA) format file containing the connections between plant mitochondrial contigs.
[0022] The GFA file was opened with Bandage software, and connections with abnormal coverage or unable to form a ring structure were removed according to the assembly relationship diagram to obtain an accurate mitochondrial genome assembly relationship diagram. When the depth difference of contigs connection on the assembly relationship diagram is too large, considering gene horizontal transfer and repeated sequences, no processing is required.
[0023] Finally, according to the connection order and forward and reverse information of Contigs in the assembly relationship diagram, Bandage was used to extract the corresponding sequences from all Contigs sequences and connect them to obtain mitochondrial genome sequences of various configurations.
[0024] Compared with the prior art, the present invention has the following beneficial effects:
[0025] The present invention utilizes the characteristics of NewBler assembly software that can retain all possible connections, as well as the copy number differences between mitochondrial and chloroplast and nuclear sequences in the third-generation WGS data, and the advantage that the third-generation sequencing reads can cross repetitive sequences, to construct a set of plant mitochondrial genome multi-configuration assembly processes suitable for third-generation full-platform (PacBio / Nanopore / HiFi) sequencing data. This process does not require the separation of mitochondrial DNA from plant samples, and can directly use whole genome sequencing data to assemble mitochondrial genomes; in addition, when encountering multiple branches caused by repetitive sequences during the assembly of the mitochondrial genome, all possible connections can be retained and continued to extend until a closed loop is formed, so that all possible configurations of the mitochondrial genome can be analyzed, which has the advantages of low cost, high efficiency, and precise assembly compared to the existing technology for assembling plant mitochondrial genomes. BRIEF DESCRIPTION OF THE DRAWINGS
[0026] Figure 1 It is a schematic diagram of the process of the present invention;
[0027] Figure 2 Assemble the Contigs coverage and length distribution map for the present invention;
[0028] Figure 3This is the assembly map of the mitochondrial genome of Salix trifoliata in Example 1 of this method;
[0029] Figure 4 This is the assembly map of Phaseolus vulgaris mitochondrial genome in Example 2 of this method;
[0030] Figure 5 This is the assembly map of the mitochondrial genome of Populus simonii in Example 3 of this invention. DETAILED DESCRIPTION
[0031] The present invention is further described below in conjunction with specific embodiments.
[0032] Example 1: Multi-main ring structure (bean)
[0033] This example is a method for assembling plant mitochondrial multi-configuration based on three-generation whole genome sequencing data, and its steps include:
[0034] (1) DNA was extracted from fresh leaves of Phaseolus vulgaris, and then high-depth third-generation sequencing (>50×) was performed using the PacBio CLR sequencing platform to obtain clean reads;
[0035] (2) Using the Canu error correction module to correct the original clean reads with its own high-depth data, high-quality (>99%) third-generation sequencing data was obtained;
[0036] (3) Using a Perl script, the reads longer than 30 kb in the high-quality sequencing data generated in step (2) were broken into 5 kb, 10 kb, 15 kb, 20 kb, and 25 kb segments, respectively, to obtain five original third-generation sequencing data sets;
[0037] (4) Use Newbler v.3.0 to preliminarily assemble the five data sets generated in step (3) to obtain preliminarily assembled contigs;
[0038] (5) Use minimap2 to align the original reads to the contigs obtained by preliminary assembly, and use Samtools depth to obtain the average coverage depth of each contig;
[0039] (6) The R language was used to draw the average sequencing depth distribution of Contigs after different reads were assembled, and the coverage depth of Contigs sequences of bean nucleus, chloroplast and mitochondrial genome was different (e.g. Figure 2 ), select the Contig with longer mitochondrial Contig sequence as the seed, in this example, select Contig00001, Contig00002 and Contig00003 as the seeds for extension;
[0040] (7) Using the Perl script, according to the relationship between the Contigs contained in the 454ContigGraph.txt file assembled by Newbler v.3.0, extend the bean Contig00001, Contig00002 and Contig00003. When the seed extension encounters multiple branches caused by repeated sequences, retain all possible connections and continue to extend until a closed loop is formed. The loop includes six Contig sequences, namely Contig00001 (180.1kb), Contig00002 (98.3kb), Contig00003 (89.8kb), Contig16185 (10.1kb), Contig27218 (4.9kb), and Contig29982 (3.8kb). If a closed loop is not formed or common plant mitochondrial protein-coding genes are included after extending all available Contigs, it is necessary to merge the sequencing data of different phase shifts and cutoffs separately, assemble again using Newbler v.3.0, and then repeat the above steps until the conditions are met;
[0041] (8) During the sequence extension process of bean mitochondrial contigs, the sequencing depth, length and connection relationship between different contigs were recorded using Perl scripts to obtain a Graphical Fragment Assembly (GFA) format file containing the connection between plant mitochondrial contigs;
[0042] (9) Use Bandage software to open the GFA file and remove the connections with abnormal coverage or unable to form a ring structure according to the assembly relationship diagram to obtain an accurate mitochondrial genome assembly relationship diagram. When the difference in contig connection depth is too large, consider gene horizontal transfer and repeated sequences;
[0043] (10) Finally, according to the connection order and forward and reverse information of Contigs in the assembly relationship diagram, Bandage was used to extract the corresponding sequences from all Contigs and connect them to obtain the mitochondrial genome sequences of three different main ring configurations (such as Figure 3 );
[0044] (11) Assembly verification: The bean mitochondrial genome sequence obtained in step (10) was aligned to the existing plant mitochondrial protein-coding genes using BlastN, and the bean mitochondrial genome assembled in this example was confirmed to be complete based on the number of aligned genes. The existing plant mitochondrial protein-coding gene library was downloaded from the NCBI Organelle Database, including protein-coding genes in all plant mitochondrial genomes. The database was automatically updated once a week using a Python crawler.
[0045] Example 2: Single main ring structure (three-core willow)
[0046] This example is a method for assembling plant mitochondrial multi-configuration based on three-generation whole genome sequencing data, and its steps include:
[0047] (1) DNA was extracted from fresh leaves of Salix triandra, and then high-depth third-generation sequencing (>50×) was performed using the PacBio CLR sequencing platform to obtain clean reads;
[0048] (2) Same as step (2) of Example 1;
[0049] (3) Same as step (3) of Example 1;
[0050] (4) Same as step (4) of Example 1;
[0051] (5) Same as step (5) of Example 1;
[0052] (6) Calling R language to draw the average sequencing depth distribution diagram of Contigs after different reads are assembled, and according to the different coverage depths of Contigs sequences of the nucleus, chloroplast and mitochondrial genome of Salix trifoliata, the Contig with the longer mitochondrial Contig sequence is selected as the seed. In this example, Contig00001 is selected to extend the seed;
[0053] (7) Using Perl script, the connection relationship between Contigs contained in the 454ContigGraph.txt file assembled by Newbler v.3.0 was used to extend Contig00001 of Salix trifoliata. When the seed extension encountered multiple branches caused by repeated sequences, all possible connections were retained and continued to extend until a closed loop was formed. The loop includes two Contig sequences, namely Contig00001 (612.8 kb) and Contig00007 (122.4 kb):
[0054] (8) Same as step (8) of Example 1;
[0055] (9) Same as step (9) of Example 1;
[0056] (10) Finally, according to the connection order and forward and reverse information of Contigs in the assembly relationship diagram, Bandage was used to extract the corresponding sequences from all Contigs and connect them to obtain the mitochondrial genome sequence of S. trifoliata with a single main ring configuration (e.g. Figure 4 );
[0057] (11) Same as step (11) of Example 1.
[0058] Example 3: Multi-sub-ring structure (Populus trichocarpa)
[0059] This example is a method for assembling plant mitochondrial multi-configuration based on three-generation whole genome sequencing data, and its steps include:
[0060] (1) DNA was extracted from fresh leaves of Populus trichocarpa, and then high-depth HiFi third-generation sequencing (about 10×) was performed using the PacBio SequalII sequencing platform to obtain clean reads;
[0061] (2) Unlike PacBio CLR or Nanopore sequencing, which requires error correction, PacBio HiFi sequencing data has an accuracy of more than 99% and can be directly used for assembly without error correction;
[0062] (3) Same as step (3) of Example 1;
[0063] (4) Same as step (4) of Example 1;
[0064] (5) Same as step (5) of Example 1;
[0065] (6) Calling R language to draw the average sequencing depth distribution diagram of Contigs after different reads were assembled, and according to the different coverage depths of Contigs sequences of the nucleus, chloroplast and mitochondrial genome of Populus trichocarpa, the Contig with the longer mitochondrial Contig sequence was selected as the seed. In this example, Contig02522, Contig02535, Contig02688 and Contig03881 were selected as seeds for extension;
[0066] (7) Using Perl script, according to the connection relationship between Contigs contained in the 454ContigGraph.txt file assembled by Newbler v.3.0, the four Contig seeds of Populus trichocarpa were extended. When the seed extension encountered multiple branches caused by repeated sequences, all possible connections were retained and continued to extend until a closed loop was formed. The loop includes 7 Contig sequences, namely Contig02522 (199.6 kb), Contig02535 (199.3 kb), Contig02688 (185.7 kb), Contig03881 (113.1 kb), Contig04573 (82.5 kb), Contig06446 (23.5 kb) and Contig09427 (439 bp);
[0067] (8) Same as step (8) of Example 1;
[0068] (9) Same as step (9) of Example 1;
[0069] (10) Finally, based on the connection order and forward and reverse information of the Contigs in the assembly relationship diagram, Bandage was used to extract the corresponding sequences from all Contigs and connect them to obtain the mitochondrial genome sequence of Populus trichocarpa with a multi-subring configuration (e.g. Figure 5 );
[0070] (11) Same as step (11) of Example 1.
Claims
1. A method for assembling plant mitochondrial genome multi-configuration based on third-generation whole genome sequencing data, characterized in that: include: 1) Obtain third-generation plant whole genome sequencing data with an accuracy of >99%; 2) Using Perl scripts, reads longer than 30 kb in the sequencing data were broken down according to several different shifts to obtain several original third-generation sequencing data sets; 3) Using Newbler v.3.0, several original third-generation sequencing data sets generated in step 2) were preliminarily assembled to obtain preliminarily assembled contigs; 4) Use minimap2 to align the original reads to the preliminarily assembled Contigs, and use Samtoolsdepth to obtain the average coverage depth of each Contig; 5) Call R language to draw the average sequencing depth distribution diagram of Contigs after different reads assembly, and screen out mitochondrial Contigs according to the different coverage depths of Contigs sequences of plant nucleus, chloroplast and mitochondrial genome, and use several mitochondrial Contigs with longer sequence length as seeds; 6) Using Perl scripts, extend the mitochondrial seed Contig according to the connections between the Contigs contained in the 454ContigGraph.txt file assembled by Newbler v.3.
0. When the extension encounters multiple branches caused by repeated sequences, retain all possible connections and continue to extend until a closed loop is formed or all existing plant mitochondrial genome protein-coding genes are included; if a closed loop is not formed or all existing plant mitochondrial genome protein-coding genes are included after extending all available Contigs, merge the sequencing data of different phase shift cutoffs respectively, and assemble again using Newbler v.3.0 until a closed loop is formed or all existing plant mitochondrial genome protein-coding genes are included; 7) During the extension of the seed Contig sequence, the sequencing depth, length and connection relationship between different Contigs were recorded using a Perl script to obtain a GFA format file containing the connection between plant mitochondrial Contigs; 8) Use Bandage software to open the GFA file, remove the connections with abnormal coverage or unable to form a ring structure according to the assembly relationship diagram, and obtain an accurate mitochondrial genome assembly relationship diagram; 9) According to the connection order and forward and reverse information of Contigs in the assembly relationship diagram, Bandage is used to extract the corresponding sequences from all Contigs sequences and connect them to obtain the mitochondrial genome sequences of all possible configurations; 10) Assembly verification: Use BlastN to align the plant mitochondrial genome sequence obtained in step 9) to all existing plant mitochondrial protein-coding genes, and confirm that the assembled mitochondrial genome is complete based on the number of aligned genes; The step 1) specifically includes: using PacBio HiFi sequencing technology to obtain plant third-generation whole genome sequencing data with an accuracy of >99%; using plant genome sequencing data obtained by PacBio CLR and Nanopore sequencing, and using Illumina data of the same sample or its own third-generation high-depth data for error correction to obtain reads with an accuracy of >99%; There are 5 different step lengths in step 2), and the length difference between different step lengths is ≥ 5k, which are 5k, 10k, 15k, 20k, and 25k respectively; The existing plant mitochondrial protein-coding gene library is the protein-coding genes in all plant mitochondrial genomes downloaded from the NCBI organelle database.
