Plasmid sequencing and sequence correction and screening method based on nanopore sequencing and use thereof

By employing a nanopore-based plasmid sequencing method, combined with multiple sequence alignment and cluster analysis, the problems of low throughput, high cost, and long time in plasmid sequencing have been solved. This method enables efficient and low-cost high-throughput plasmid sequencing and sequence correction screening, making it suitable for large-scale industrial applications.

WO2026232812A1PCT designated stage Publication Date: 2026-11-12BIOINTRON BIOLOGICAL INC +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
PCT/CN2025/143227
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2025-05-09
Filing Date
2025-12-17
Publication Date
2026-11-12

AI Technical Summary

Technical Problem

Existing technologies for plasmid sequencing suffer from problems such as low sequencing throughput, high cost, long time, and complex data analysis. In particular, third-generation sequencing platforms have low single-sequencing accuracy and low throughput, while second-generation sequencing cannot achieve long-fragment sequencing and lacks mature data analysis algorithms.

Method used

We employ a plasmid-based sequencing method based on nanopore sequencing, combined with multiple sequence alignment algorithms and cluster analysis models. Through data preprocessing, barcode matching, sequence alignment analysis, and modular design, we achieve high-throughput sequencing and sequence correction screening, including quality control, barcode matching, sequence alignment, and cluster analysis. We optimize the processing flow to improve accuracy and efficiency.

Benefits of technology

It enables high-throughput sequencing of up to 10,000+ plasmids, significantly improving screening efficiency and throughput, reducing costs, and is suitable for large-scale industrial plasmid screening. It can handle complex DNA structures and provide detailed analysis results, shortening sequencing time and reducing reagent costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN2025143227_12112026_PF_FP_ABST
    Figure CN2025143227_12112026_PF_FP_ABST
Patent Text Reader

Abstract

Provided are a plasmid sequencing and sequence correction and screening method based on nanopore sequencing and a use thereof. The method comprises: (1) designing vector primers carrying barcodes, performing PCR amplification on target fragments of plasmids, mixing PCR amplification products of inserted fragments of different plasmids into one tube, subjecting mixed DNA to unified library construction, loading and performing sequencing, and obtaining a corresponding amount of raw output data on the basis of the number of sequencing samples and the sizes of the plasmids; (2) on the basis of a sequence correction and screening system based on high-throughput sequencing, performing quality control on raw sequencing data, establishing a many-to-one alignment equation between sequencing sequences and a template sequence, deriving conserved bases by means of position-specific base frequency statistics, and obtaining a consensus base sequence by means of sequence recombination; and (3) developing a clone screening decision model to determine clones by means of setting an alignment threshold between the consensus sequence and the template sequence, thereby implementing automated classification and screening of clones.
Need to check novelty before this filing date? Find Prior Art

Description

A plasmid sequencing and sequence deviation correction screening method based on nanopore sequencing and its application Technical Field

[0001] This application belongs to the field of high-throughput sequencing technology and synthetic biology, specifically involving the application technology of third-generation single-molecule real-time sequencing (TGS) in the verification of genetically engineered plasmid cloning, and particularly involving a plasmid sequencing and sequence correction screening method based on nanopore sequencing and its application. Background Technology

[0002] Gene synthesis is the process of synthesizing a complete DNA molecule from a DNA sequence using chemical methods. It allows for the design and synthesis of various functional elements, including protein-coding sequences, RNA interference molecules, and gene expression regulators, according to specific needs. This technology has wide applications in modern life sciences, medical research, and biopharmaceuticals, playing a crucial role in various biological research and industrial applications such as antibody development, dual-luciferase reporter gene experiments, gene editing experiments, protein labeling experiments, and cloning expression experiments. It provides strong support for genetic engineering and genomics research. Verifying the correctness of the synthesized gene is essential for downstream use, and this can be achieved through DNA sequencing.

[0003] First-generation sequencing (Sanger sequencing) has the advantages of high accuracy, long read length and suitability for complex sequences. Currently, most plasmid validation sequencing work is concentrated on first-generation sequencing platforms. However, its low sequencing throughput, high sample preparation and sequencing costs, and long sequencing time limit its application in large-scale sequencing.

[0004] Next-generation sequencing (NGS high-throughput sequencing) has the advantages of high throughput and low cost, but it has disadvantages such as short read length (not exceeding 600 bp) requiring sequencing to be interrupted during plasmid sequencing, complex bioinformatics analysis for data splicing and analysis in the later stages of sequencing data, and long sequencing time and inflexibility.

[0005] Third-generation sequencing is suitable for long-fragment plasmid sequencing. Although it has significant technological advantages, existing sequencing platforms and commercial kits can only be used for simultaneous sequencing of a small number of samples. They suffer from low accuracy, low throughput, and high cost per sequencing run, and there is no corresponding mature data analysis algorithm model.

[0006] However, first-generation sequencing has low throughput, high time and reagent costs, second-generation NGS sequencing cannot achieve long-fragment sequencing, has long sequencing time, high cost and complex subsequent data analysis, and third-generation sequencing has low throughput, random errors and strand specificity bias in the raw data of a single sequencing run, and systematic bias in the data. To address the above key technical obstacles, developing a high-accuracy and high-throughput sequencing scheme has important application value. Summary of the Invention

[0007] This application provides a nanopore sequencing-based plasmid sequencing and sequence correction screening method and its application. This application constructs a plasmid sequencing validation method based on a nanopore sequencing platform for plasmid samples with known sequences. High-throughput sequencing of plasmids is achieved by adding tag sequences. By integrating multiple sequence alignment algorithms and cluster analysis models, the ability to resolve low-frequency mutation signals in the original data is significantly improved, achieving a screening throughput of over 10,000 plasmids per sequencing run. Compared to traditional Sanger sequencing, this method greatly improves screening efficiency and throughput, significantly reduces screening costs, and achieves high-throughput direct sequencing of long plasmid fragments compared to next-generation sequencing. It achieves efficient sequencing at low cost and is suitable for large-scale industrial plasmid screening and sequence discovery.

[0008] In a first aspect, this application provides a sequence correction and screening system based on high-throughput sequencing, the system comprising:

[0009] Data preprocessing module: used for quality control and filtering of raw sequencing data, removing low-quality DNA sequences;

[0010] Barcode matching module: (A) For DNA sequences without linker sequences, first perform multi-pattern string matching; then perform BLAST verification on the matching results; (B) For DNA sequences containing linker sequences, first... The function identifies the location of the linker sequence in the DNA sequence; then the DNA sequence is divided into multiple fragments according to the location of the linker sequence, and each fragment represents a functional region;

[0011] Sequence alignment analysis module: Performs BLAST alignment between the DNA sequence processed by the barcode matching module and the reference sequence; extracts alignment information, including alignment position and similarity; calculates the base frequency at each position based on the alignment information, merges the conserved bases at each position, and generates a consistent sequence.

[0012] The post-processing correction system of this application adopts a modular design, forming a complete sequencing data processing pipeline through the organic combination of various functional modules. The system mainly includes a data preprocessing module, a barcode matching module, and a sequence alignment analysis module.

[0013] In this application, the quality control and filtration use The function processes the input raw sequencing data.

[0014] Preferably, the quality control and filtering steps include: setting a quality fraction threshold of 10, filtering the DNA sequences by quality fraction, and removing low-quality DNA sequences; setting the length range of the DNA sequences to ±15% of the standard sequence length, screening the DNA sequence length, setting the minimum length to 400 and the maximum length to 3500, and removing DNA sequences that are too short or too long; and then converting the original DNA sequence data in FASTQ format to FASTA format data.

[0015] Preferably, the barcode matching module also employs a sequence front search optimization strategy during processing, searching for the barcode within the base range of the first 1-150 bp of the sequence.

[0016] In this application, since the barcode is designed to be at the very beginning of the sequence, the program only searches for the barcode in the range of 1-150, which greatly reduces the search space and improves the processing speed.

[0017] Preferably, in the barcode matching module, the multi-pattern string matching is performed based on the Aho-Corasick matching algorithm; the Aho-Corasick matching algorithm uses... Function implementation.

[0018] Preferably, in the barcode matching module, the BLAST verification passes. Function implementation.

[0019] Preferably, in the barcode matching module, the... The function employs a string search algorithm to locate multiple connection sequences; The function includes the following steps: (1) accurately locate the segmentation position according to the provided connectors; (2) the entire template sample will be divided into n+1 shorter subsequences according to the number of connectors n; (3) the obtained sub-segments are used for the next matching step.

[0020] Preferably, in the sequence alignment analysis module, the source of the reference sequence includes: the system constructs a complete template sequence based on the primer sequence and target DNA sequence in the sample information table, which serves as the reference sequence for alignment.

[0021] Preferably, in the sequence alignment analysis module, the parameters set for BLAST alignment include: setting a gap open penalty. Extended penalty points Word length Valid match score Minimum similarity .

[0022] Preferably, in the sequence alignment analysis module, the method for calculating the base frequency includes: using a hidden Markov model to statistically analyze the base frequency of each site, setting a threshold of ≥4:1 for determining a conserved site; and introducing a Bayesian correction factor. ,in, From platform-specific error distribution database, This represents the probability that the prediction is correct. This represents the probability of actually being observed.

[0023] In this application, the data processed by the above sequence correction and screening system can be further analyzed, such as performing sequence clustering analysis to identify subpopulations and variations in the sequence; or performing result evaluation and visualization.

[0024] In this application, sequence clustering analysis is performed based on MMseqs2. First, the input sequence is preprocessed and its format converted. Then, MMseqs2 clustering commands are executed according to set parameters (including any one or at least a combination of two of the following: minimum sequence similarity, coverage threshold, clustering mode, or sensitivity). Next, the clustering results are parsed and statistical information is generated. The function identifies subgroups in a sequence.

[0025] In this application, the result evaluation includes:

[0026] (1) The system passes The function calculates the degree of matching between the consistency sequence and the template sequence, accurately locates and records the location and type of mismatch, and provides quantitative indicators for sequence quality assessment; and / or (2) detection of special sequence features, including: bimodal detection and / or polymer detection.

[0027] The bimodal detection is used to identify positions where the second high-frequency base frequency is >8%.

[0028] The polymer detection is to identify sequence regions with more than 5 consecutive C / G.

[0029] The visualization includes: using the plotly library to generate an interactive base frequency chart, which visually displays the base distribution at each position and marks special sites and regions.

[0030] The data preprocessing module of this application is responsible for quality control and filtering of the raw sequencing data, removing low-quality sequences to provide a high-quality data foundation for subsequent analysis; the barcode matching module identifies and classifies sequences of different samples based on a specific algorithm, achieving accurate sample differentiation; the sequence alignment analysis module performs sequence alignment and consistency analysis using the BLAST algorithm, and is the core processing unit of the system; the sequence clustering module performs sequence clustering analysis on specific samples to identify subpopulations within the sequences; and the result evaluation module is responsible for calculating and displaying various indicators of the analysis results, providing users with an intuitive presentation of the results. This modular design not only improves the maintainability and scalability of the system but also allows each processing step to be optimized independently, thereby improving overall performance.

[0031] The error correction system in this application employs a multi-level data preprocessing technique to ensure the data quality for subsequent analysis. First, sequences are filtered for quality scores based on a set quality threshold (default 10), eliminating low-quality reads. Second, sequences of appropriate length are selected according to a preset length range (default ±15% of the standard sequence length) to avoid interference from excessively short or long sequences. Finally, the original FASTQ format data is converted to FASTA format data, which is more convenient for subsequent analysis, simplifying the data structure and improving processing efficiency. Specifically, this application uses the `seq_qc_filter` function to process the input sequencing data. This function calls the external tool `seqkit` (a sequence processing tool) for quality control and format conversion. This multi-level preprocessing strategy not only improves data quality but also reduces the computational burden of subsequent analysis, laying the foundation for efficient and accurate sequence analysis. Furthermore, this system allows users to adjust the quality threshold and length range parameters according to specific needs, providing flexible preprocessing options.

[0032] This application proposes a highly efficient barcode matching method for its post-processing correction system, enabling rapid and accurate identification of sequences from different samples within large datasets. This method integrates several advanced algorithms, including multi-pattern string matching based on the Aho-Corasick matching algorithm (through...). Function implementation) and BLAST-based sequence similarity search (through (Function implementation). To further improve matching efficiency, this application also adopts a sequence front-end search optimization strategy, that is, searching for the barcode only in the first N bases of the sequence (default 1-150 bp), which greatly reduces the search space and improves processing speed. In practical applications, the system first uses the Aho-Corasick matching algorithm for rapid initial screening, and then performs BLAST verification on the matching results to ensure the accuracy of the matching. This dual verification mechanism not only improves the accuracy of matching, but also effectively handles possible errors and variations in the barcode sequence. In addition, this system also supports custom settings for various barcode formats and matching parameters to adapt to the needs of different experimental designs.

[0033] The core of the post-processing error correction system in this application lies in its sequence alignment and consistency analysis technology, which is a key link in achieving high-accuracy sequence error correction.

[0034] First, the system constructs a complete template sequence based on the primer sequences and target DNA sequences in the sample information table as a reference for alignment. Then, a local BLAST database is created for the sequence set of each barcode to improve alignment efficiency. Next, optimized BLAST parameters (such as evalue (effective match score)) are used. (Minimum similarity) Efficient alignment is performed on (word length, etc.), balancing speed and accuracy; then, effective alignment information, including alignment position, similarity and other key data, is extracted from the BLAST output; finally, based on the alignment results, the base frequency at each position is calculated to generate a consistent sequence.

[0035] In particular, this application is in and The function implements two different sequence processing strategies, applicable to ordinary sequences and complex samples containing linker sequences, respectively, greatly expanding the system's applicability. Furthermore, this system employs various optimization strategies, such as adaptive parameter adjustment and result caching, further improving the efficiency and accuracy of alignment analysis.

[0036] The post-processing correction system of this application proposes a specialized processing method for complex samples containing linker sequences, a field rarely addressed in existing technologies. This method firstly... This function identifies the locations of linker sequences in a DNA sequence using an efficient string search algorithm that accurately locates multiple linker sequences. Then, the DNA sequence is segmented into multiple fragments based on the linker sequence locations, with each fragment representing a functional region. Next, each fragment is compared and analyzed separately to obtain fragment-level concordance sequences. Finally, the analysis results of each fragment are integrated in their original order to generate a complete concordance sequence. This segmentation strategy not only improves the accuracy of alignment but also preserves the structural information of the linker sequences, which is of great significance for understanding complex DNA structures. Furthermore, this system implements an automatic detection and verification mechanism for linker sequences, capable of handling different types and numbers of linker sequences to meet the needs of various complex samples. This technological innovation enables the system to handle structurally complex DNA sequences, significantly expanding its application scope.

[0037] This application's post-processing correction system provides sequence clustering analysis for specific samples (such as samples labeled "DB") to identify subpopulations and variations within sequences. This function is based on MMseqs2 for efficient sequence clustering and supports various adjustable clustering strategies, including minimum sequence similarity, coverage threshold, clustering pattern, and sensitivity. In practical applications, the system first preprocesses and converts the input sequence, then executes MMseqs2 clustering commands according to user-defined parameters, parses the clustering results, generates statistical information, and finally integrates the clustering results into the final analysis report. This system can identify subpopulations within sequences, providing deeper information for the analysis of complex samples. This clustering analysis not only helps discover heterogeneity in samples but also identifies potential mutations and variations, which is of great value for disease diagnosis and biomarker discovery. Furthermore, the system provides a visualization of the clustering results, enabling researchers to intuitively understand the distribution and relationships of sequences.

[0038] The post-processing correction system of this application also includes comprehensive result evaluation and visualization technology, providing users with intuitive and detailed analysis results. Regarding result evaluation, the system uses… The system calculates the degree of matching between the consistent sequence and the template sequence, accurately locating and recording the position and type of mismatches, providing quantitative indicators for sequence quality assessment. Furthermore, the system implements detection functions for various special sequence features, including bimodal detection (identifying positions where the second highest base frequency is >8%, potentially indicating double cloning) and polymer detection (identifying sequence regions with more than 5 consecutive C / G ratios, which may lead to sequencing errors). For visualization, the system uses the plotly library to generate interactive base frequency charts, visually displaying the base distribution at each position and annotating special sites and regions. These visualization tools not only provide a global view of the sequence data but also support local zoom-in and detailed viewing, facilitating researchers to discover and analyze special patterns and variations in the sequence. Through these assessment and visualization techniques, this system can provide comprehensive and intuitive analytical results, greatly improving the efficiency and accuracy of data interpretation.

[0039] To improve processing efficiency, this bioinformatics-based post-processing correction workflow employs various parallel processing and performance optimization techniques, enabling the entire analysis process to efficiently handle high-throughput sequencing data. In terms of parallel processing, the system can perform multi-process parallel processing of different barcodes and parallel execution of BLAST alignment tasks, fully utilizing the computing power of modern multi-core processors. Regarding performance optimization, the system adopts a streaming strategy for large files, avoiding loading large amounts of data into memory at once and reducing memory usage; it implements an intelligent temporary file management mechanism, automatically creating and cleaning temporary files to optimize disk usage; furthermore, the system employs various algorithm optimization strategies, such as indexing acceleration and caching mechanisms, to further improve processing speed. The comprehensive application of these techniques allows the system to complete the analysis of large-scale data within a reasonable timeframe, significantly improving work efficiency. Actual testing shows that for datasets containing hundreds of samples and millions of sequences, this system can complete comprehensive analysis within hours, while traditional methods may require several days.

[0040] This application's bioinformatics-based post-processing correction workflow also incorporates a robust error handling and fault tolerance mechanism to ensure stable operation and reliable results under various conditions. Regarding input validation, the system comprehensively validates input files and parameters, including file format checks and parameter range verification, promptly identifying and reporting potential problems. For anomaly detection, the system implements a multi-layered anomaly detection and handling mechanism in key processing steps, capable of identifying and recording various anomalies and providing detailed error information. For result validation, the system performs consistency checks on the generated results to ensure data integrity and accuracy. Regarding fault tolerance, the system implements various fault-tolerance strategies, such as allowing a certain degree of mismatch at the sequence ends (e.g., the last 5 bp), which is highly effective in handling the common end-quality degradation problem in sequencing data. Furthermore, the system provides detailed logging functionality to record key steps and potential problems during processing, facilitating subsequent analysis and debugging. The comprehensive application of these mechanisms ensures the system's stability and the reliability of the results, providing valuable analytical results even when processing data of poor quality or complex structure.

[0041] The post-processing correction system of this application provides a flexible and scalable command-line interface, enabling users to configure system parameters according to specific needs. This interface includes required parameters (sequence file path, sample information file path, and output file name) and various optional parameters (including parameters for various modules such as quality control, barcode matching, BLAST alignment, and sequence clustering). To improve user experience, the system organizes related parameters into logical groups, such as a quality control parameter group, a BLAST parameter group, and a clustering parameter group, making the parameter structure clearer. The system can parse and validate user-input parameters and configure corresponding processing flows based on these parameters. Furthermore, the system provides default parameter values ​​and parameter range checks to reduce the possibility of user configuration errors. This flexible parameter configuration mechanism allows the system to adapt to different research needs and data characteristics, from simple sequence alignment to complex multi-sample cluster analysis, all achievable through appropriate parameter settings.

[0042] Secondly, this application provides a plasmid sequencing and sequence deviation correction screening method based on nanopore sequencing, the method comprising:

[0043] (1) Design vector primers with barcodes to amplify the target fragment of plasmid by PCR. Mix the PCR amplification products of different plasmid insert fragments into one tube, construct a library and sequence the mixed DNA in a unified manner, and obtain the corresponding amount of raw data according to the number of sequencing samples and the size of plasmid.

[0044] (2) Based on the high-throughput sequencing sequence correction and screening system described in the first aspect, data quality control is performed on the original sequencing data, a many-to-one alignment equation between the sequencing sequence and the template sequence is established, conserved bases are derived through position-specific base frequency statistics, and a consistent base sequence is obtained through sequence recombination.

[0045] (3) Develop a clone screening decision model. By setting the threshold for the comparison degree between the consistency sequence and the template sequence, clones are judged and the automatic hierarchical screening of clones is realized.

[0046] Preferably, in step (1), the barcode-bearing vector primer is designed based on the vector backbone and is used to amplify the target fragment. The primer is attached with a tag sequence to distinguish different plasmids.

[0047] Preferably, in step (3), the method for determining the clone includes: using the Needleman-Wunsch global alignment algorithm to compare the consistent sequence with the real template sequence, setting a mismatch penalty of -5, a gap penalty of -2, a matching score of +1, and determining a valid clone when the similarity is 100%.

[0048] Thirdly, this application provides the application of the nanopore sequencing-based plasmid sequencing and sequence correction screening method described in the second aspect in high-throughput sequencing.

[0049] Fourthly, this application provides a computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the steps in the sequence correction and screening system based on high-throughput sequencing described in the first aspect.

[0050] Fifthly, this application provides a computer device including a memory and a processor, wherein a computer program capable of running on the processor is stored in the memory, and the processor executes the computer program to implement the steps in the sequence correction and screening system based on high-throughput sequencing as described in the first aspect.

[0051] Explanation of technical terms used in this application:

[0052] BLAST, or Basic Local Alignment Search Tool, is a software application that uses either a query sequence and a target sequence, or a database of sequences containing multiple such sequences. BLAST searches the database for subsequences that are similar to the query sequence.

[0053] Function: The meaning of this function name is: to use the BLAST algorithm to implement barcode matching, with the purpose of extracting the sequence corresponding to the barcode from the entire offline database and storing them separately according to different barcodes. The process is as follows: (1) construct the barcode library to be matched; (2) construct the search library from the offline sequence; (3) match each barcode one by one, with the parameter being: word length Valid match score Minimum similarity (4) Store the results matched by each barcode.

[0054] Aho-Corasick matching algorithm: For a detailed explanation of the algorithm, please refer to the following link: In computer science, the Aho-Corasick matching algorithm is a string search algorithm, applied for by Alfred V. Aho and Margaret J. Corasick in 1975.

[0055] Function: The meaning of this function name is: to use the Aho-Corasick algorithm to implement barcode matching, with the purpose of extracting the sequence corresponding to the barcode from the entire offline database and storing them separately according to different barcodes. The process is as follows: (1) constructing the barcode library to be matched; (2) constructing the search library from the offline sequence; (3) matching each barcode one by one; (4) storing the matching result of each barcode.

[0056] The Needleman-Wunsch global alignment algorithm, also known as the Needleman-Wunsch algorithm, is an algorithm used in bioinformatics for aligning protein or nucleotide sequences. It was one of the first applications of dynamic programming to comparing biological sequences. Essentially, the algorithm decomposes a large problem (such as a complete sequence) into a series of smaller problems, and then uses the solutions to the smaller problems to find the optimal solution to the larger problem. It is sometimes also referred to as the optimal matching algorithm and the global alignment technique. The Needleman-Wunsch algorithm remains widely used for optimal global alignment, especially when the quality of the global alignment is crucial. The algorithm assigns a score to each possible alignment, and its goal is to find all possible alignments with the highest scores.

[0057] Function: The meaning of this function name is: to segment the sequence according to the linker sequence. The purpose is to divide the entire sequence into different sub-sequences according to the position of the linker sequence in the sequence, so as to facilitate subsequent comparison. This function is mainly adapted to the plasmid sequencing corresponding to the bispecific antibody and scFv antibody. The function includes the following steps: (1) accurately locate the segmentation position according to the provided linker; (2) the entire template sample will be divided into n+1 shorter sub-sequences according to the number of linkers n; (3) the obtained sub-fragments are used for the next step of matching.

[0058] Function: The meaning of this function name is: to perform quality control screening on the sequence. The main purpose of this function is to screen all reads from the machine according to strict quality standards, thereby removing noise, improving the robustness of the whole process, and enhancing the credibility of the final result. The function includes the following process: (1) use the linux cat command to merge all the fastq format files from the machine to obtain the all.fastq file; (2) use the seqtk software to convert all.fastq to FATA to obtain the all.fasta file; (3) use the seqkit software to perform length and quality screening on all.fasta, with parameters m (minimum length) of 400, M (maximum length) of 3500, and Q (quality) of 10, and finally obtain The file contains the address of the seqtk software. seqkit software address: .

[0059] Function: The name of this function means parsing input parameters. It is mainly used to parse command-line parameters. Its core function is to convert the parameters (such as options, flags or parameter values) entered by the user through the command line into structured data that the program can recognize. This function realizes the standardized processing of input parameters by defining parameter rules (such as parameter type, default value, help information, etc.), thereby supporting the program to dynamically adjust the running logic according to the parameters. The steps include: (1) Initialize the parser: Create a parameter parser object and configure global description information; (2) Define parameter rules: Implement parameter description by adding parameter definitions; (3) Parse parameters: Call the parsing method. During this process, the validity of the parameters is automatically verified (such as type matching, existence of required parameters); (4) Return and pass parameters: Return the parsed parameter object for other modules of the program to call, such as reading the input file path, setting the running mode or enabling the debugging flag.

[0060] MMseqs2, short for Many-against-Many sequence searching, is a software used to search and cluster large sets of protein and nucleotide sequences. MMseqs2 is a free and open-source software implemented in C++. The steps include (1) creating a database of plasmid sequences that may have bimodal distributions for clustering, named DB.fasta; (2) performing clustering using MMseqs2 with the parameters set to...

[0061] ;

[0062] (3) The clustering results are sent to professionals for evaluation.

[0063] The `run_clustering` function: This function runs the clustering steps. It uses the MMSeqs2 software to cluster plasmid sequences that may have bimodal interference. The resulting clustering can assist professionals in reviewing the results and further reducing errors. The steps include: (1) creating a database (DB.fasta) of plasmid sequences that may have bimodal interference; (2) performing clustering using mmseqs2 with the following parameters:

[0064] ;

[0065] (3) The clustering results are sent to professionals for evaluation.

[0066] Function: The Chinese name of this function is: Alignment with Template Sequence. The main purpose of this function is to perform a final alignment between the obtained consistent sequence and the real template sequence to ensure that the consistent sequence obtained in the process is real and usable, and to reduce errors. The steps of this function include: (1) preparing the consistent sequence; (2) preparing the template sequence; (3) performing multiple sequence alignment using the Needleman-Wunch algorithm on the two sequences; (4) judging based on the alignment score. If the matching rate of the two sequences is 100%, they are considered usable; if it is less than 100%, they are considered unusable.

[0067] The plotly library: Plotly is a Python graphics library that creates publication-quality interactive graphics. Software link: .

[0068] Compared with the prior art, this application has the following advantages:

[0069] (1) This application enables simultaneous sequencing of mixed library construction for up to 10,000 or more gene fragments inserted into different plasmids, overcoming the limitation of the number of samples that can be used in a single run of commercial kits, and enabling simultaneous sequencing of plasmid insert fragments of different lengths, thereby reducing library construction time and reagent costs.

[0070] (2) This application provides a unified library construction method for mixed DNA, which improves the efficiency of library construction, significantly reduces the workload, realizes automated operation, reduces operational errors, and reduces library construction costs.

[0071] (3) This application optimizes the algorithm for judging plasmid sequencing sequences, enabling the plasmid sequence to be judged at a sequencing depth of 100×, and the plasmid sequence to be accurately read with 1G of raw data from 4000 samples. The amount of sequencing data can be freely selected according to the number of samples, and sequencing can be performed at any time on the nanopore sequencer. The sequencing time required for 4000 samples is as short as 1.5 hours. The sequencing time is flexible, and the chip can be used multiple times, greatly reducing the sequencing time and reagent cost.

[0072] (4) The plasmid sample interpretation algorithm designed in this application realizes automatic alignment of sequencing data to sample sequence to synthetic target sequence, and automatic judgment of positive plasmid clones. The interpretation time of 4000 samples is as short as 15 minutes, which greatly improves the judgment efficiency of plasmid sequencing data.

[0073] (5) The post-processing correction system of this application adopts a modular design. Through the organic combination of various functional modules, a complete sequencing data processing pipeline is formed. This modular design not only improves the maintainability and scalability of the system, but also enables each processing link to be optimized independently, thereby improving the overall performance. Attached Figure Description

[0074] Figure 1 is a flowchart of the sequencing analysis in this application.

[0075] Figure 2 is a summary analysis of the results of application example 1. Detailed Implementation

[0076] The technical solution of this application will be further described below through specific embodiments. Those skilled in the art should understand that the embodiments described are merely to help understand this application and should not be regarded as specific limitations on this application.

[0077] Where specific techniques or conditions are not specified in the examples, they shall be performed in accordance with the techniques or conditions described in the literature in this field, or in accordance with the product instructions. Reagents or instruments whose manufacturers are not specified are all conventional products that can be purchased through legitimate channels.

[0078] Example 1

[0079] This embodiment provides a plasmid sequencing and sequence correction screening method based on nanopore sequencing, and the flowchart is shown in Figure 1.

[0080] 1. Document preprocessing

[0081] Universal primers for different samples were designed, and barcode sequence tags were added to the upstream primers to amplify different target gene fragments of plasmids. PCR products amplified with different tagged primers were initially quantified using Nanodrop (micro-spectrophotometer), and the concentration difference was adjusted to be within 2-fold. DNA was pipetted into equal volumes and mixed into one tube using a 96-well pipette. The mixed DNA library was purified using AMPure XP magnetic beads. 1 μL of the purified product was quantified using qubit (Qubit dsDNA HS Assay Kit, Invitrogen, cat # Q32851) to ensure the library contained at least 100-200 fmol for subsequent use. DNA fragment size analysis was performed using Qseq-1 plus.

[0082] 2. Library Construction

[0083] The amplified DNA fragments were used to construct libraries using the NANOPORE SQK-LSK114 library preparation kit and T4 ligase.

[0084] 1) End-of-pipe repair

[0085] Thaw the DNA quality control sample (DCS) at room temperature, prepare the end repair system according to the NEBNext Ultra II End Repair / dA-tailing Module reagent instructions, perform end repair on 100-200 fmol of amplicon DNA and purify it using AMPure XP magnetic beads.

[0086] 2) Connector connection and purification

[0087] The adapter ligation system was prepared using T4 ligase (NEBNext Quick Ligation Module NEB, cat # E6056), and the DNA fragments were purified using AMPure XP magnetic beads and SFB washing buffer. The eluted DNA was quantified using a qubit (quantitative quantification system) to ensure sufficient DNA for the library.

[0088] 3) Chip preprocessing and sample loading

[0089] Remove the sequencing chip (FLO-MIN114) beforehand and allow it to equilibrate at room temperature for 5 minutes before use. Before loading the library, perform flow cell quality control on the GridION sequencer to ensure that the number of available wells meets the sequencing requirements. Prepare and load the sequencing chip pretreatment solution mix, let it stand, and then prepare and load the library solution mix, operating slowly to ensure that no air bubbles are introduced into the chip during the entire process.

[0090] 3. On-machine sequencing

[0091] Set the on-machine information of sequencing samples, calculate the amount of sequencing data according to the number of samples, terminate the sequencing when the data volume is sufficient, save the data and complete the sequencing run.

[0092] 4. Unload samples and clean the chip for later use

[0093] Clean the unloaded chip in time according to the instructions of the NANOPORE washing kit (EXP-WSH004). Prepare the chip washing buffer mix, stop or pause the sequencing experiment in MinKNOW, keep the sequencing chip in the sequencer, add the chip washing buffer and wait for 60 minutes before aspirating the waste liquid. After adding the storage buffer, store the sequencing chip in a 4-8°C refrigerator for next use.

[0094] 5. Quality control of off-machine data

[0095] Use the sliding window method (window length 50 bp, step size 10 bp) to calculate the local Phred quality score, and filter out reads with a quality value < Q10 and a length deviating from the template by ±15%.

[0096] Detect chimeras based on k-mer spectrum analysis (k=7) to ensure data quality.

[0097] 6. Barcode matching

[0098] Use the barcode corresponding to each clone to search and match the entire off-machine database by means of regular expressions, find the reads corresponding to each clone barcode and split them, waiting for the next step of analysis.

[0099] 7. Sequence alignment

[0100] The BLAST algorithm is used for multiple sequence alignment, and the gap opening penalty is set , the extension penalty , the word length , the effective match score , the minimum similarity .

[0101] Perform many-to-one alignment between high-quality sequencing sequences and template sequences to generate a multiple sequence alignment (MSA) matrix.

[0102] 8. Base frequency statistics and conserved base deduction

[0103] Use the Hidden Markov Model (HMM) to count the base frequency of each site. Since the background noise of third-generation sequencing is too strong, the judgment threshold is set such that a site is determined as a conserved site when the frequency ratio of the main peak to the secondary peak is ≥4:1.

[0104] Introducing a Bayesian correction factor: ,in, From platform-specific error distribution database, This represents the probability that the prediction is correct. This represents the probability of actually being observed.

[0105] 9. Consistent sequence generation

[0106] By merging the conserved bases at each position, a consistent base sequence is generated.

[0107] 10. Clone screening

[0108] The Needleman-Wunsch global alignment algorithm is used to compare consistent sequences with the true template sequences. A penalty of -5 is applied for mismatches, -2 for gaps, and +1 for matches. A sequence with 100% similarity is considered a valid clone. Valid clones are then transferred to the production process, while invalid clones are deleted.

[0109] Example 2

[0110] This embodiment provides a plasmid sequencing and sequence correction screening method based on nanopore sequencing. This embodiment also provides a method for processing complex samples containing linker sequences.

[0111] 1. Document preprocessing

[0112] The steps for library preprocessing are as described in Example 1.

[0113] 2. Library Construction

[0114] The steps for building the library are as described in Example 1.

[0115] 3. Sequencing

[0116] The sequencing procedure is as described in Example 1.

[0117] 4. Remove the sample from the machine and clean the chip for later use.

[0118] This step is based on Example 1.

[0119] 5. Data quality control after shutdown

[0120] This step is based on Example 1.

[0121] 6. Barcode matching

[0122] This step refers to Example 1.

[0123] 7. Sequence alignment

[0124] First, for a given connection sequence, use... This function identifies the locations of linker sequences in a DNA sequence. It employs an efficient string search algorithm to accurately locate multiple linker sequences. After this step, the entire template sample is divided into n+1 shorter subsequences based on the number of linker sequences (n). For each shorter subsequence, a BLAST algorithm is used for multiple sequence alignment, with an open penalty for gaps. Extended penalty points Word length Valid match score Minimum similarity .

[0125] The high-quality sequencing sequence is compared with the subsequence of the template sequence in a many-to-one manner to generate n+1 multiple sequence alignment matrices (MSA).

[0126] 8. Base frequency statistics and conserved base derivation

[0127] This step refers to Example 1.

[0128] 9. Consistent sequence generation

[0129] Similar to step 7 above, the conserved bases of the n+1 subsequences are merged, and then the previously separated linker sequence is inserted into the middle of the subsequences to restore the length of the original template sequence, thereby generating a consistent base sequence similar to that in Example 1 above.

[0130] 10. Clone screening

[0131] The Needleman-Wunsch global alignment algorithm is used to compare consistent sequences with the true template sequences. A penalty of -5 is applied for mismatches, -2 for gaps, and +1 for matches. A sequence with 100% similarity is considered a valid clone. Valid clones are then transferred to the production process, while invalid clones are deleted.

[0132] Application Example 1

[0133] This application example provides a plasmid sequencing and sequence deviation correction screening method based on nanopore sequencing, the flowchart of which is shown in Figure 1. Data numbered 20250116-02 (data sourced from Shanghai Baiying Biotechnology Co., Ltd.) is used to demonstrate and supplement the technical effects of this application; this batch of data contains 4464 clones.

[0134] This method achieves efficient and accurate plasmid sequencing and clone screening through optimized experimental procedures and computational analysis strategies. Specific technical indicators are as follows:

[0135] 1. Document preprocessing

[0136] Universal primers for different samples were designed, and barcode sequence tags were added to the upstream primers to amplify different target gene fragments of plasmids. The PCR products amplified with different tagged primers were initially quantified using Nanodrop (micro-spectrophotometer), and the concentration difference was adjusted to be within 2-fold. DNA was pipetted into equal volumes using a 96-well pipette and mixed into one tube. The mixed DNA library was purified using AMPure XP magnetic beads. 1 μL of the purified product was quantified using Qubit (Qubit dsDNA HS Assay Kit, Invitrogen, cat # Q32851) to ensure the library contained at least 100-200 fmol for subsequent use. DNA fragment size analysis was performed using Qseq-1 plus. The library pretreatment process took only 0.5 h for this batch of 4464 clones.

[0137] 2. Library Construction

[0138] The amplified DNA fragments were used to construct a library using the NANOPORE SQK-LSK114 library preparation kit and T4 ligase. The total time for the library preparation stage was 2 hours.

[0139] 1) End-of-pipe repair

[0140] Thaw the DNA quality control sample (DCS) at room temperature, prepare the end repair system according to the NEBNext Ultra II End Repair / dA-tailing Module reagent instructions, perform end repair on 200 fmol of amplicon DNA and purify it using AMPure XP magnetic beads.

[0141] 2) Connector connection and purification

[0142] The adapter ligation system was prepared using T4 ligase (NEBNext Quick Ligation Module NEB, cat # E6056), and the DNA fragments were purified using AMPure XP magnetic beads and SFB washing buffer. The eluted DNA was quantified using a qubit (quantitative quantification system) to ensure sufficient DNA for the library.

[0143] 3) Chip preprocessing and sample loading

[0144] Take out the sequencing chip (FLO-MIN114) in advance, equilibrate it at room temperature for 5 min before use, and complete the quality inspection of the flow cell on the GridION sequencer before loading the library to ensure that the number of available pores meets the sequencing requirements. Prepare and load the sequencing chip pretreatment mixture (mix), let it stand, prepare and load the library mixture (mix), and operate slowly to ensure that no air bubbles are introduced into the chip during the whole operation process.

[0145] 3. On-machine sequencing

[0146] Set the on-machine information of sequencing samples, calculate the sequencing data volume according to the number of samples, terminate the sequencing when the data volume is sufficient, save the data and unload the samples from the sequencer. The entire sequencing stage only takes 1 hour and 40 minutes, the raw data volume obtained by sequencing is 1 Gb, and the sequencing depth is 300X.

[0147] 4. Unload the sample and clean the chip for later use

[0148] Clean the unloaded chip in time according to the instructions of the NANOPORE cleaning kit (EXP-WSH004). Prepare the chip cleaning buffer mixture (mix), stop or pause the sequencing experiment in MinKNOW (a software used in sequencing), keep the sequencing chip in the sequencer, add the chip cleaning buffer and wait for 60 minutes before aspirating the waste liquid. After adding the storage buffer, store the sequencing chip in a 4-8°C refrigerator for the next use.

[0149] 5. Quality control of off-machine data

[0150] Use the sliding window method (window length 50 bp, step size 10 bp) to calculate the local Phred quality score, and filter out reads with a quality value < Q10 and a length deviating from the template by ±15%.

[0151] Detect chimeras based on k-mer spectrum analysis (k=7) to ensure data quality, and the quality control step can achieve an accuracy of 99.9%.

[0152] 6. Barcode matching

[0153] Barcode matching based on regular expressions (recall rate 99.9%), the average matching time for a single sample is < 0.1 s (32 cores, 64 threads for parallel processing).

[0154] 7. Sequence alignment

[0155] Adopt the parallelized BLAST algorithm (32 cores, 64 threads, 128 GB of memory), and the parameter optimization is as follows:

[0156] Set the gap open penalty , the extension penalty , the word length Valid match score Minimum similarity .

[0157] The alignment speed reaches 2000 sequences / second, the linker sequence identification accuracy is 96.25%, and the MSA generation time is < 10 minutes (for tens of thousands of reads).

[0158] 8. Base frequency statistics and conserved base derivation

[0159] The HMM model (training set accuracy > 99%) combined with Bayesian correction (error rate reduced by 40%) achieves a conservative site determination accuracy of 97.3% when the main peak / secondary peak threshold is ≥ 4:1.

[0160] 9. Consistent sequence generation

[0161] Needleman-Wunsch global alignment (parameters set to: mismatch -5, gap -2, match +1), achieving a 100% similarity clone screening accuracy of 96.8%.

[0162] 10. Clone screening

[0163] The Needleman-Wunsch global alignment algorithm is used to compare the consistent sequence with the real template sequence. A penalty of -5 is set for mismatch, -2 for gap, and +1 for matching. A similarity of 100% is considered a valid clone.

[0164] The invalid clone rejection rate was >99.5%, and the success rate of effective clone production conversion increased to 96.2%.

[0165] The results of this application example are summarized and analyzed in Figure 2.

[0166] This application example comprehensively validates the effectiveness of the third-generation sequencing bias correction method. Through four core indicators—accuracy, barcode matching recall, precision, and invalid clone rejection rate—it systematically demonstrates the superior performance of this technology. Specifically, the application example results show that the bias correction method achieves an accuracy of 96.24%, indicating that its bias correction processing of sequencing data almost completely eliminates errors, achieving near-perfect restoration of true sequence information and providing extremely high data reliability for subsequent analysis. Regarding barcode matching recall, this method reaches 99.99%, fully demonstrating its ability to effectively identify and match barcode sequences in samples, significantly reducing the risk of data loss due to barcode mismatches, and providing strong support for sample tracking and genotyping in high-throughput sequencing. The precision rate of 97.40% further illustrates the method's outstanding ability to suppress false positive results during the bias correction process, ensuring the authenticity of the output results and avoiding interference from misjudgments in the analytical conclusions. Furthermore, the invalid clone rejection rate is as high as 99.50%, highlighting the superior efficiency of this technology in filtering low-quality or invalid sequencing data, effectively improving data utilization and reducing the burden of invalid data on subsequent analysis.

[0167] In summary, the third-generation sequencing correction method proposed in this application demonstrates significant advantages in data accuracy, completeness, and reliability, far exceeding the performance of existing conventional methods. In particular, the synergistic effect of high accuracy and high invalid clone rejection rate not only significantly improves the quality of sequencing data but also optimizes the efficiency of data processing workflows, providing more precise and efficient technical support for applications such as genomics research and clinical diagnosis. Furthermore, the overall performance of the four indicators also verifies the stability and adaptability of this method in complex sequencing environments, laying a solid foundation for its industrial application.

[0168] The technical advantages of this application's technical solution compared to traditional methods are summarized in Table 1.

[0169]

[0170] Note: All data is based on a standard test environment (GridION X5, Intel Xeon 6248R, 128 GB DDR4, Linux).

[0171] In summary, this application provides a plasmid sequencing and sequence correction screening method based on nanopore sequencing and its application. The method includes: (1) designing barcode-enabled vector primers to perform PCR amplification on the target fragment of the plasmid, and constructing a sequencing library suitable for the nanopore sequencing platform from the amplified fragment; (2) sequencing on the sequencing machine, and harvesting the corresponding amount of raw sequencing data according to the number of sequencing samples and the size of the plasmid; (3) constructing a data analysis algorithm for the sequencing data, and establishing a deep correction algorithm based on multiple sequence alignment: implementing data quality control on the raw sequencing data, establishing a multiple-to-one alignment equation between the sequencing sequence and the template sequence, deriving conserved bases through position-specific base frequency statistics, and obtaining a consistent base sequence through sequence recombination; (4) developing a clone screening decision model, and realizing automated hierarchical screening of clones by setting the alignment degree threshold between the consistent sequence and the template sequence. The innovation of this application lies in constructing a plasmid sequencing validation method based on a nanopore sequencing platform for plasmid samples synthesized from known sequences. High-throughput sequencing of plasmids is achieved by adding tag sequences. By integrating multiple sequence alignment algorithms and cluster analysis models, the ability to resolve low-frequency mutation signals in the original data is significantly improved, enabling a screening throughput of more than 10,000 plasmids in a single sequencing run. Compared with traditional Sanger sequencing, this method greatly improves screening efficiency and throughput, significantly reduces screening costs, and achieves high-throughput direct sequencing of long-fragment plasmids compared with second-generation sequencing. It achieves efficient sequencing at low cost and is suitable for large-scale industrial plasmid screening and sequence discovery.

[0172] The applicant declares that the above description is only a specific implementation of this application, but the protection scope of this application is not limited thereto. Those skilled in the art should understand that any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application fall within the protection and disclosure scope of this application.

Claims

1. A high-throughput sequencing based sequence correction screening system, comprising: a data preprocessing module for quality control and filtering of raw sequencing data, removing low-quality DNA sequences; Barcode matching module: (A) for DNA sequences without connection sequences, first multi-mode string matching is performed; then the matching results are verified by BLAST; (B) for DNA sequences with connection sequences, first the connection sequences are removed, and then multi-mode string matching is performed; finally, the matching results are verified by BLAST. a function identification module for identifying the junction sequence position in the DNA sequence, and dividing the DNA sequence into multiple fragments according to the junction sequence position, each fragment representing a functional region; a sequence alignment analysis module for BLAST alignment of the DNA sequence processed by the barcode matching module with a reference sequence; extracting alignment information, including alignment position and similarity; based on the alignment information, calculating the base frequency of each position, and combining the conserved bases of each position to generate a consensus sequence.

2. The high-throughput sequencing based sequence correction and screening system of claim 1, wherein, The quality control and filtering step includes setting the quality score threshold to 10, filtering the DNA sequence by quality score, and removing low-quality DNA sequences; setting the length range of the DNA sequence to ± 15% of the standard sequence length, screening the length of the DNA sequence, setting the minimum length to 400 and the maximum length to 3500, and removing DNA sequences that are too short or too long; and converting the FASTQ format data of the original DNA sequence to FASTA format data.

3. The high-throughput sequencing based sequence correction and screening system of claim 1, wherein, The barcode matching module also uses a sequence front-end search optimization strategy to search for barcodes in the first 1-150 bp of the sequence.

4. The high-throughput sequencing based sequence correction and screening system of claim 1, wherein, In the barcode matching module, the multi-mode string matching is based on Aho-Corasick matching algorithm, and the Aho-Corasick matching algorithm is matched through function implementation; The BLAST verification passes Function implementation; The The function adopts a string search algorithm to realize positioning of multiple connection sequences. The function includes the following steps: (1) accurately positioning the division position according to the provided linker; (2) the entire template sample will be divided into n+1 shorter sub-sequences according to the number of linkers n, (3) the obtained sub-fragments are used for the next matching; In the sequence alignment analysis module, the source of the reference sequence includes: the system constructs the complete template sequence based on the primer sequence and the target DNA sequence in the sample information table, and uses it as the reference sequence for alignment.

5. The high-throughput sequencing based sequence correction and screening system of claim 1, wherein, In the sequence alignment analysis module, the parameters of the BLAST alignment setting include: setting gap opening penalty , extension penalty , word length , effective matching score , and minimum similarity . In the sequence alignment analysis module, the base frequency calculation method comprises: using a hidden Markov model to calculate the base frequency of each site, setting a threshold value, and determining a conservative site when the ratio of the main peak / second highest peak frequency is greater than or equal to 4:1; and introducing a Bayesian correction factor: wherein, from a database of platform-specific error distributions, denotes the probability of being predicted as correct, Indicates the probability of a true observation. 6.A plasmid sequencing and sequence correction screening method based on nanopore sequencing, comprising: (1) designing a barcode-containing vector primer, PCR amplifying the target fragment of the plasmid, mixing the PCR amplification products of different plasmid insert fragments into one tube, uniformly building a library and sequencing the mixed DNA, and harvesting raw data of the corresponding data amount according to the number of sequencing samples and the size of the plasmid; (2) using the high-throughput sequencing based sequence correction screening system according to any one of claims 1-5 to perform data quality control on the raw sequencing data, establishing a many-to-one comparison equation between the sequencing sequence and the template sequence, deducing the conserved bases through position-specific base frequency statistics, and obtaining a consensus base sequence through sequence recombination; (3) developing a clone screening decision model, determining the clones by setting the alignment degree threshold of the consensus sequence and the template sequence, and realizing the automatic classification screening of the clones.

7. The nanopore sequencing based plasmid sequencing and sequence correction screening method of claim 6, wherein, In step (1), the barcode-containing vector primer is designed based on the vector backbone and is used to amplify the target fragment, and a label sequence is connected to the primer to distinguish different plasmids. In step (3), the determination method of the clone comprises: using Needleman-Wunsch global alignment algorithm to align the consistent sequence with the real template sequence, setting the mismatch penalty-5, the gap penalty-2, the matching score+1, and the similarity 100% to determine the effective clone.

8. Use of the nanopore sequencing based plasmid sequencing and sequence correction screening method of claim 6 or 7 in high-throughput sequencing.

9. A computer readable storage medium, wherein, The storage medium has stored thereon a computer program, wherein the computer program is executed by the processor to implement the steps in the high-throughput sequencing based sequence correction screening system of any one of claims 1-5.

10. A computer device comprising a memory and a processor, having stored on said memory a computer program capable of running on the processor, wherein, The processor executes the computer program to implement the steps in the high-throughput sequencing based sequence correction screening system of any one of claims 1-5.