A method for assembling pan-structures of plant organelles and counting the frequencies of different conformations

Abstract

The present invention discloses a method for assembling a pan-structure of plant organelles and counting the frequencies of different conformations, and belongs to the field of bioinformatics. The method first performs HiFi sequencing on plant samples, and uses Jellyfish and Genomescope software to evaluate the size of the plant genome on HiFi data; then uses Flye to assemble the genome on the HiFi sequencing data to obtain the Contigs sequence and the initial assembly file of the Contigs; and then uses Blastn to align the conserved protein sequence of the plant mitochondrial genome and the plant chloroplast genome sequence to the sequences of all Contigs, etc., to achieve the assembly of the pan-structure of plant organelles and count the frequencies of different conformations. The results show that the method is superior to the existing organelle genome assembler in terms of assembly accuracy and completeness, and only 1×HiFi sequencing data is needed to obtain a complete plant organelle genome.

Description

Technical Field

[0001] The present invention belongs to the field of bioinformatics, and more specifically, relates to a method for assembling a pan-structure of plant organelles and counting the frequencies of different conformations. Background Art

[0002] Mitochondria not only act as "energy centers" in cells, but also play a key role in cell survival, development, and stress response. Precise regulation of mitochondrial function is essential for cell homeostasis and adaptability. So far, in the NCBI database (https: / / www.ncbi.nlm.nih.gov / genome / ), there are nearly 13,000 complete chloroplast genome data, while there are less than 680 complete plant mitochondrial genomes. This has seriously lagged behind the research progress of plant mitochondrial genomes. With the continuous improvement of sequencing technology and assembly methods, the plant mitochondrial genome is now developing rapidly.

[0003] However, due to the presence of a large number of repetitive sequences and horizontally transferred sequences in the plant mitochondrial genome, its structure is highly variable. According to the completed plant mitochondrial genome reports, there are a large number of polycyclic structures in onions, camphor, potatoes, etc. Although there are some mitochondrial genome assembly software (such as MitoHiFi and GSAT) that have proposed relatively mature solutions for single rings or simple polycyclic structures, problems such as complex polycyclic structures and the frequency of each ring structure in plant mitochondria are still unsolved. Therefore, this application proposes a pan-structure assembly method for plant organelle genomes that focuses on whole-genome HiFi sequencing data. Summary of the invention

[0004] 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 the pan-structure of plant organelles and counting the frequencies of different conformations.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is as follows:

[0006] A method for assembling a plant organelle pan-structure and counting the frequencies of different conformations, comprising the following steps:

[0007] 1) Perform HiFi sequencing on plant samples to obtain high-quality HiFi sequencing data;

[0008] 2) Use Jellyfish and Genomescope software to estimate plant genome size using HiFi data;

[0009] 3) Use Flye to assemble the HiFi sequencing data and obtain the initial assembly graph of Contigs, assembly_graph.gfa file;

[0010] 4) Use Blastn to align the conserved protein sequences of the plant mitochondrial genome and the plant chloroplast genome sequences to the sequences of all Contigs. Sort by Identity, Evalue, and Score in the alignment results of the mitochondrial conserved protein sequences and chloroplast genomes, and select the top five best aligned Contigs as seeds for the plant mitochondrial genome and chloroplast genome, respectively. For less than five Contigs, select all of them;

[0011] 5) According to the estimated plant genome size and HiFi data volume, the depth of plant genome Contigs is calculated; then, the depth of nuclear genome Contigs, the depth of mitochondrial genome Contigs and the depth of chloroplast genome Contigs are selected as the three input parameters required by the Gaussian mixture model sklearn.mixture in the scikit-learn library in Python, and the depth of genome Contigs, the depth of mitochondrial genome seeds and the depth of chloroplast genome seeds are marked, and the mitochondrial genome depth and chloroplast genome depth intervals are calculated;

[0012] 6) Use Python script to take mitochondrial genome seed as input parameter to implement DFS to complete the initial GFA format file of mitochondrial genome, and the same steps to complete the initial GFA file of chloroplast genome; the specific process is: first, dynamically scan the assembly_graph.gfa file output in step 3) to capture all Contigs adjacent to the seed; update the Contigs in the depth interval of mitochondrial genome to the seed list, and extract the connection relationship between adjacent Contigs in the assembly_graph.gfa file and the sequence of the Contig and write them into the initial GFA file of mitochondrial genome, and expand the dynamically added Contigs outward, and visit each layer of Contigs in turn until the length of the seed list no longer changes;

[0013] Using the above method, the initial GFA file of the chloroplast genome is calculated and output;

[0014] 7) For the initial GFA file obtained, use the Python script to delete the individually suspended abnormal Contigs and Bubbled Contigs to obtain a simplified GFA file;

[0015] 8) Use Python scripts to select and simplify repeated contigs that lead to multiple structures in GFA; store the contigs adjacent to the repeated contigs into a list and sort them by depth; store the list of contigs adjacent to each repeated contig into a dictionary in turn; perform full permutation on the dictionary to generate all possible conformational paths;

[0016] 9) Traverse and simplify the GFA according to each conformational path, and output the GFA path of a single conformation until the pan-structure of the organelle genome is obtained;

[0017] 10) Call the Python script to calculate the number of times each pair of connections in the conformation appears in the three conformations as the number of repetitions, and record it in a dictionary format. Extract the 2000bp sequence before and after each pair of connections in each conformation, use minimap2 to align the original HiFi sequencing data, and then use Samtools depth to calculate the average depth of the connection at 2000bp and assign it to the corresponding conformation. Finally, divide the average depth of 2000bp of all connections in the conformation by the number of repetitions and add up the average value as the frequency of the conformation.

[0018] In step 1), DNA was extracted from fresh Arabidopsis leaves and third-generation sequencing was performed using the PacBio HiFi sequencing platform to obtain high-quality HiFi sequencing data.

[0019] In step 2), Jellyfish was used to perform k-mer counting to obtain a spectrum, and then Genomescope2 was used to analyze the spectrum, and it was calculated that the Arabidopsis genome was 125M.

[0020] In step 3), the assembly_graph.gfa file contains Contigs sequences, Contigs depths and connection relationships between Contigs.

[0021] The process of step 5) is as follows: Divide the amount of assembled data by the genome size to obtain the depth of nuclear genome contigs; use nuclear genome contigs, mitochondrial contigs, and chloroplast contigs as the three input parameters required by the Gaussian mixture model sklearn.mixture in the scikit-learn library in Python. Fit the data under the assumption that there are three mixture components, and calculate the intervals corresponding to mitochondrial contigs and chloroplast contigs.

[0022] The deletion process in step 7) is as follows: the first step is to delete the abnormal contigs that are suspended alone in the initial GFA file, which are considered to be the influencing factors for the failure to form a ring; the second step is to delete the bubbled contigs in the initial GFA file.

[0023] Step 8) Obtain paths to three possible conformations.

[0024] In step 9), the Python script is called to loop through the paths of the three conformations, read the assembly_graph.gfa file, and write the connection information of each path into a separate conformation GFA file, and finally the organelle pan-structure of Arabidopsis thaliana is obtained.

[0025] Beneficial effects: Compared with the prior art, the present invention has the following advantages: The method of assembling the pan-structure of plant organelles and counting the frequencies of different conformations of the present invention is based on whole-genome HiFi sequencing, which is superior to the existing organelle genome assembler in terms of assembly accuracy and completeness, and only 1×HiFi sequencing data is needed to obtain a complete plant organelle genome. The Gaussian mixture model can more accurately determine the depth interval of the chloroplast genome and mitochondrial genome, which helps to filter the mixed sequences of the nuclear genome, and efficiently analyze all the structures and frequencies of organelles, which has good practicality. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a pan-structural organelle genome map. DETAILED DESCRIPTION

[0027] The present invention is further illustrated below in conjunction with specific examples. The examples are implemented under the premise of the technical solution of the present invention. It should be understood that these examples are only used to illustrate the present invention and are not used to limit the scope of the present invention. In the following examples, if there is no detailed method or process, the technical means used are conventional means well known to those skilled in the art. Example 1

[0028] A method for assembling pan-structures of plant organelle genomes based on whole-genome HiFi sequencing, the specific steps are as follows:

[0029] 1) Arabidopsis thaliana ( Arabidopsis thaliana ) DNA was extracted from fresh leaves and third-generation sequencing was performed using the PacBio HiFi sequencing platform to obtain high-quality HiFi sequencing data.

[0030] 2) The high-quality HiFi sequencing data were used to estimate the genome size using Jellyfish (v2.3.1) and Genomescope (v1.0) software. First, Jellyfish was used to perform k-mer counting to obtain the spectrum, and then Genomescope2 was used to analyze the spectrum, and the Arabidopsis genome was calculated to be approximately 125M.

[0031] 3) The HiFi dataset was preliminarily assembled using the genome assembly software Fyle (v2.9.3) with default parameters to obtain the initial assembly graph of contigs (assembly_graph.gfa file), which contains the contigs sequence, contigs depth, and the connection relationship between contigs.

[0032] 4) Extract the Contigs sequence from the assembly_graph.gfa file in step 3) using the awk command. Use the Python script to call Blastn to align the high-quality plant mitochondrial protein-coding genes and plant chloroplast genomes published by NCBI to the Contigs sequence, and then sort them according to Identity, Evalue and Score, and automatically select the 5 best aligned Contigs as the seeds of the mitochondrial genome and chloroplast genome. If there are less than 5 Contigs, the program will select all of them.

[0033] 5) According to the genome size obtained by evaluating step 2) and the amount of HiFi sequencing data in step 1), the depth of the plant nuclear genome Contig is calculated. First, the amount of assembled data is divided by the genome size to obtain the depth of the nuclear genome Contigs. Considering that the depth of chloroplast Contigs is greater than the depth of mitochondrial Contigs, and the depth of mitochondrial Contigs is greater than the depth of nuclear genome Contigs, this embodiment uses the depth of nuclear genome Contigs, the depth of mitochondrial Contigs, and the depth of chloroplast Contigs as the three input parameters required for the Gaussian mixture model sklearn.mixture in the scikit-learn library in Python. Fit the data under the assumption that there are 3 mixed components, and calculate the intervals corresponding to the depth of mitochondrial Contigs and the depth of chloroplast Contigs.

[0034] 6) Use Python script to implement the breadth-first search algorithm (DFS) to calculate and output the initial GFA (Graphical Fragment Assembly, GFA) format file of the mitochondrial genome.

[0035] First, dynamically scan the assembly_graph.gfa file output from step 3) to capture all Contigs adjacent to the seed. Update the Contigs in the depth interval of the mitochondrial genome to the seed list, extract the connection relationship between adjacent Contigs in the assembly_graph.gfa file and write the sequence of the Contig into the initial GFA file of the mitochondrial genome, and expand the dynamically added Contigs outward, visiting each layer of Contigs in turn until the length of the seed list no longer changes.

[0036] The method is the same as above, calculate and output the initial GFA file of the chloroplast genome.

[0037] 7) Call the Python script to simplify the initial GFA file of the Arabidopsis organelle genome obtained in step 6) to obtain a simplified GFA file.

[0038] The first step is to delete the abnormal contigs that are suspended alone in the initial GFA file, which is considered to be the influencing factor for the failure of ring formation;

[0039] In the second step, the bubbled contigs in the initial GFA file were deleted to obtain a simplified GFA file of the Arabidopsis mitochondrial genome with 18 contigs, where the length and depth information of each contig are shown in Table 1.

[0040] Table 1 Characteristics of Arabidopsis thaliana assembled contigs

[0041]

[0042] 8) Repeated Contigs are the main cause of multiple conformations. A Python script was used to determine whether it was a repeated Contig based on the depth of the Contigs and the number of adjacent Contigs. Two repeated Contigs were identified in the simplified GFA file of Arabidopsis, namely Contig00395 and Contig00452. At the same time, a Python script was used to scan the Contigs that were connected to the repeated Contigs, store them in the corresponding path list and sort them according to the depth. The path list corresponding to Contig00395 and Contig00452 was stored in the dictionary, and a full arrangement was performed to obtain the paths of the three possible conformations.

[0043] 9) Call the Python script to loop through the paths of the three conformations, read the assembly_graph.gfa file, and write the connection information of each path into a separate conformation GFA file, and finally obtain the organelle pan-structure of Arabidopsis thaliana ( Figure 1 ).

[0044] 10) Call the Python script to calculate the number of times each pair of connections in the conformation appears in the three conformations as the number of repetitions, and record it in a dictionary format (Contig00229-234: 1). Extract the 2000bp sequence before and after each pair of connections in each conformation, use minimap2 to align the original HiFi sequencing data, and then use Samtools depth to calculate the average depth of 2000bp at the connection and assign it to the corresponding conformation. Finally, the average depth of 2000bp at all connections in the conformation is divided by the number of repetitions and the average is accumulated to obtain the frequency of the conformation (Table 2).

[0045] Table 2 Frequency of conformations

[0046]

Claims

1. A method for assembling the pan-structure of plant organelles and counting the frequencies of different conformations. It is characterized in that The following steps are involved: 1) Perform HiFi sequencing on plant samples to obtain high-quality HiFi sequencing data; 2) Use Jellyfish and Genomescope software to estimate plant genome size using HiFi data; 3) Use Flye to assemble the HiFi sequencing data and obtain the initial assembly graph of Contigs, assembly_graph.gfa file; 4) Use Blastn to align the conserved protein sequences of the plant mitochondrial genome and the plant chloroplast genome sequences to the sequences of all Contigs; sort the mitochondrial conserved protein sequences and chloroplast genome alignment results according to Identity, Evalue and Score, and select the top five best aligned Contigs as seeds for the plant mitochondrial genome and chloroplast genome, respectively. For less than five Contigs, all are selected; 5) Calculate the depth of plant genome Contigs based on the estimated plant genome size and HiFi data volume; Then, the nuclear genome contigs depth, mitochondrial genome contigs depth and chloroplast genome contigs depth were selected as the three input parameters required by the Gaussian mixture model sklearn.mixture in the scikit-learn library in Python, and the genome contigs depth, the depth of the mitochondrial genome seed and the depth of the chloroplast genome seed were marked, and the mitochondrial genome depth and chloroplast genome depth intervals were calculated; 6) Use Python script to take mitochondrial genome seed as input parameter to implement DFS to complete the initial GFA format file of mitochondrial genome, and the same steps to complete the initial GFA file of chloroplast genome; the specific process is: first, dynamically scan the assembly_graph.gfa file output in step 3) to capture all Contigs adjacent to the seed; update the Contigs in the depth interval of mitochondrial genome to the seed list, and extract the connection relationship between adjacent Contigs in the assembly_graph.gfa file and the sequence of the Contig and write them into the initial GFA file of mitochondrial genome, and expand the dynamically added Contigs outward, and visit each layer of Contigs in turn until the length of the seed list no longer changes; Using the above method, the initial GFA file of the chloroplast genome is calculated and output; 7) For the initial GFA file obtained, use the Python script to delete the individually suspended abnormal Contigs and BubbledContigs to obtain a simplified GFA file; 8) Use Python scripts to select and simplify repeated contigs that lead to multiple structures in GFA; The contigs adjacent to the repeated contigs are stored in a list and sorted by depth; the lists of contigs adjacent to each repeated contig are stored in a dictionary in turn; the dictionary is fully arranged to generate all possible conformational paths; 9) Traverse and simplify the GFA according to each conformational path, and output the GFA path of a single conformation until the pan-structure of the organelle genome is obtained; 10) Call the Python script to calculate the number of times each pair of connections in the conformation appears in the three conformations as the number of repetitions, and record it in a dictionary format; extract the 2000bp sequence before and after each pair of connections in each conformation, use minimap2 to align the original HiFi sequencing data, and then use Samtools depth to calculate the average depth of 2000bp at the connection and assign it to the corresponding conformation; finally, divide the average depth of 2000bp of all connections in the conformation by the number of repetitions and add up the average value to obtain the frequency of the conformation.

2. The method for assembling the pan-structure of plant organelles and counting the frequencies of different conformations according to claim 1, It is characterized in that In step 1), DNA was extracted from fresh Arabidopsis leaves and third-generation sequencing was performed using the PacBio HiFi sequencing platform to obtain high-quality HiFi sequencing data.

3. The method for assembling the pan-structure of plant organelles and counting the frequencies of different conformations according to claim 1, It is characterized in that In step 2), Jellyfish was used to perform k-mer counting to obtain a spectrum, and then Genomescope2 was used to analyze the spectrum, and it was calculated that the Arabidopsis genome was 125M.

4. The method for assembling the pan-structure of plant organelles and counting the frequencies of different conformations according to claim 1, It is characterized in that In step 3), the assembly_graph.gfa file contains Contigs sequences, Contigs depths and connection relationships between Contigs.

5. The method for assembling the pan-structure of plant organelles and counting the frequencies of different conformations according to claim 1, It is characterized in that The process of step 5) is as follows: divide the amount of assembled data by the genome size to obtain the depth of nuclear genome contigs; use the depths of nuclear genome contigs, mitochondrial contigs, and chloroplast contigs as the three input parameters required by the Gaussian mixture model sklearn.mixture in the scikit-learn library in Python; fit the data under the assumption that there are three mixture components, and calculate the intervals corresponding to the mitochondrial contigs and chloroplast contigs.

6. The method for assembling the pan-structure of plant organelles and counting the frequencies of different conformations according to claim 1, It is characterized in that The deletion process in step 7) is as follows: the first step is to delete the abnormal contigs that are suspended alone in the initial GFA file, which are considered to be the influencing factors for the failure to form a ring; the second step is to delete the bubbled contigs in the initial GFA file.

7. The method for assembling a plant organelle pan-structure and counting the frequencies of different conformations according to claim 1, It is characterized in that Step 8) Obtain paths to three possible conformations.

8. The method for assembling the pan-structure of plant organelles and counting the frequencies of different conformations according to claim 7, It is characterized in that In step 9), the Python script is called to loop through the paths of the three conformations, read the assembly_graph.gfa file, and write the connection information of each path into a separate conformation GFA file, and finally the organelle pan-structure of Arabidopsis thaliana is obtained.

Images