A high-throughput microbial identification method based on 16S rRNA gene full-length nanopore sequencing

Abstract

The application discloses a high-throughput microbial identification method based on 16S rRNA gene full-length nanopore sequencing, and belongs to the technical field of biology. The method establishes a complete process of multi-strain tagging library construction, nanopore sequencing and bioinformatics analysis in response to the demand of large-scale strain deduplication and identification in high-throughput microbial isolation and cultivation omics. The multi-strain tagging library construction adopts a double-end barcode system, and connects a well tag and a plate tag to the 5' ends of universal amplification primers 27F and 1492R of 16S rRNA gene respectively, so as to realize independent marking of each well strain in a multi-well plate culture. The method can simultaneously complete the identification of 9,216-884,736 strains in a single nanopore sequencing operation, and is matched with a special bioinformatics analysis process to automatically complete quality control, tag splitting, consistent sequence acquisition and species annotation. Compared with a conventional Sanger sequencing method, the application improves the strain identification throughput by more than two orders of magnitude, reduces the manual operation time by more than 90%, realizes the whole-process standardization from samples to reports, significantly reduces the cost of single-strain identification, and provides key technical support for high-throughput microbial isolation and cultivation omics research.

Description

Technical Field

[0001] This invention relates to the field of biotechnology, specifically to a high-throughput microbial identification method based on full-length nanopore sequencing of 16S rRNA genes, which is suitable for rapid and standardized identification of large-scale bacterial strains in high-throughput microbial isolation and culture omics. Background Technology

[0002] While culture-free technologies based on high-throughput sequencing have greatly expanded our understanding of the diversity and function of environmental microorganisms, culture-based research methods remain crucial for deeply elucidating the mechanisms of microbial action and realizing their functional applications. In recent years, the rise of high-throughput microbial isolation and culture and culturomics has placed even higher demands on the ability to identify high-throughput strains.

[0003] Currently, researchers commonly use Sanger sequencing to obtain bacterial 16S rRNA gene sequences for strain identification. However, when using traditional methods such as Sanger sequencing to identify large numbers of bacteria in environmental samples, researchers often face problems such as numerous manual steps, cumbersome data processing, lack of standardized procedures, and difficulty in comparing and integrating results. In practice, this has become one of the main bottlenecks restricting the large-scale acquisition of culturable microorganisms from environmental samples. When identifying large batches of bacterial strains, these problems specifically manifest as follows:

[0004] (1) Each strain’s 16S rRNA gene PCR product needs to be labeled and sequenced separately, and labeling and recording alone takes a lot of time.

[0005] (2) The PCR product of the 16S rRNA gene of each strain requires two Sanger sequencing results. Each sequencing reaction returns a peak plot file (.ab1) and a sequence file (.seq). Researchers need to open and check the peak plot quality one by one and determine whether there are double peaks, background signals, premature termination, etc.

[0006] (3) The Sanger sequencing sequence is of poor quality at the front and back ends, requiring manual extraction of high-quality regions, and then splicing the two reaction sequences into a full-length sequence based on their overlap.

[0007] (4) The full-length sequence is then manually compared with BLAST and homology is determined to obtain species annotation results. Then the annotation results of each strain need to be manually sorted.

[0008] (5) There may be differences in background noise judgment, quality extraction, database comparison, similarity threshold, etc. between different laboratories and researchers, making it difficult to compare or integrate the results horizontally.

[0009] Applying next-generation sequencing technologies such as Illumina to the identification of high-throughput microbial isolation and culture (Nature Protocols, 2021, 16: 988-1012) can avoid most of the above problems. However, due to the short read length of next-generation sequencing sequences (usually 200-500 bp), it is difficult to achieve species-level annotation accuracy, especially for closely related species. In practice, for deduplicated strains, it is still necessary to rely on the Sanger method to determine the full-length 16S rRNA gene sequence for reliable identification.

[0010] In recent years, nanopore sequencing technology has identified base sequences by capturing the changes in current generated when DNA molecules pass through motor proteins and an applied electric field. It boasts significant advantages such as long read lengths, real-time sequencing, and portable equipment, making it highly suitable for bacterial species identification based on the full-length 16S rRNA gene sequence. However, applying nanopore sequencing technology to large-scale strain identification in cultinomy still presents the following challenges:

[0011] (1) The accuracy of a single sequencing is relatively low (about 92-97%), but since 16S rRNA gene sequencing can obtain consistent sequences through multiple sequence alignment, increasing the sequencing depth can effectively improve the accuracy, but increasing the sequencing depth directly increases the cost of single-bacterial sequencing.

[0012] (2) Existing commercial nanopore sequencing kits have a limited number of barcodes (usually 12-96), and the number of strains that can be distinguished in a single sequencing run is limited, which cannot meet the throughput requirements for the simultaneous identification of hundreds, thousands or even more strains in cultinomy.

[0013] (3) There is a lack of standardized library construction procedures and bioinformatics analysis procedures specifically for large-scale strain identification.

[0014] To address the aforementioned issues, existing technologies have explored some approaches, but none have provided a complete solution. For example, an existing patent application (application publication number: CN113667729A) attempts to use nanopore sequencing for the direct detection of pathogenic microorganisms in clinical samples, but it uses single-layer commercial barcodes (≤96 layers) and is mainly aimed at detecting low-abundance pathogens in mixed clinical samples, and still cannot achieve large-scale parallel identification of pure cultures.

[0015] In summary, current technologies do not provide a complete high-throughput culture omics identification solution, encompassing well plate layout design, structured barcode amplification, mixed library construction, consistency sequence correction, and automated report generation. Therefore, there is an urgent need to develop a high-throughput microbial identification method that can significantly increase the throughput of a single sequencing run, reduce the cost of single-bacterial identification, and achieve full-process standardization. Summary of the Invention

[0016] The purpose of this invention is to provide a high-throughput microbial identification method based on full-length 16S rRNA gene nanopore sequencing, in order to solve the problems of cumbersome operation, low throughput, and poor standardization in the existing Sanger sequencing method for large-scale strain identification, as well as the problems of limited throughput, high cost, and lack of supporting procedures in the application of existing nanopore sequencing schemes in cultinomy.

[0017] To achieve the above objectives, the present invention provides the following technical solution:

[0018] A high-throughput microbial identification method based on full-length nanopore sequencing of 16S rRNA genes, characterized by comprising the following steps:

[0019] (1) Multi-strain barcode labeling library construction: The full-length 16S rRNA gene of multiple strains cultured in a multi-well plate was amplified by PCR using a paired-end barcode primer system. The paired-end barcode primer system includes a well barcode forward primer and a plate barcode reverse primer. The well barcode forward primer is composed of the 5' end of the universal 16S rRNA gene forward primer 27F connected to the well barcode sequence, and the plate barcode reverse primer is composed of the 5' end of the universal 16S rRNA gene reverse primer 1492R connected to the plate barcode sequence. By assigning different well barcode forward primers to different wells of the same multi-well plate and different plate barcode reverse primers to different multi-well plates, independent labeling of strains in each well is achieved. The amplification products of each well are mixed to construct a nanopore sequencing library.

[0020] (2) Nanopore sequencing: Perform nanopore sequencing on the sequencing library constructed in step (1) to obtain sequencing data;

[0021] (3) Bioinformatics analysis: The sequencing data obtained in step (2) are subjected to quality control, barcode splitting, single strain consistency sequence acquisition and species annotation to obtain the identification results of each strain in batches.

[0022] Preferably, the length of the perforated barcode sequence and the plate barcode sequence in step (1) is 12 to 24 bp, the GC content is 40% to 60%, and the Hamming distance between any two perforated barcode sequences and between any two plate barcode sequences is not less than 6.

[0023] Preferably, there are 96 well barcode sequences and 96 plate barcode sequences. The well barcode sequences are selected from the nucleotide sequences shown in SEQ ID NO.3 to SEQ ID NO.98, and the plate barcode sequences are selected from the nucleotide sequences shown in SEQ ID NO.99 to SEQ ID NO.194. The nucleotide sequence of the forward primer 27F is shown in SEQ ID NO.1, and the nucleotide sequence of the reverse primer 1492R is shown in SEQ ID NO.2.

[0024] Preferably, the multi-well plate is a 96-well plate, with each well corresponding to an independent combination of well barcode and plate barcode. Different wells within the same 96-well plate are distinguished by using the same plate barcode reverse primer and different well barcode forward primers, or by using the same well barcode forward primer and different plate barcode reverse primers, and the barcode combination corresponding to each well remains unique throughout the entire experimental system.

[0025] Preferably, the construction of the nanopore sequencing library in step (1) includes: mixing equal volumes of PCR amplification products from each well, purifying them with magnetic beads, repairing the ends, and then connecting them with sequencing adapters to obtain the final sequencing library; when it is necessary to perform batch sequencing on amplification products from multiple multi-well plates from different environmental samples, a secondary library barcode provided by a commercial nanopore sequencing kit is introduced for marking and differentiation when or before connecting the sequencing adapters.

[0026] Preferably, the PCR reaction system for full-length amplification of the 16S rRNA gene in step (1) is as follows: 12.5 μL of 2× PCR premix, 1 μL of 10 μM barcode forward primer, 1 μL of 10 μM plate barcode reverse primer, 2 μL of template DNA or colony suspension, and ddH2O to a final volume of 25 μL; the PCR amplification program is as follows: 95℃ pre-denaturation for 3 min; 95℃ denaturation for 30 s, 58℃ annealing for 30 s, 72℃ extension for 90 s, for a total of 30 cycles; and 72℃ final extension for 5 min.

[0027] Preferably, in step (2), the nanopore sequencing performs real-time preliminary quality control during base recognition, filtering out reads with an average Phred quality value <10 or a sequence length <1000 bp.

[0028] Preferably, the further filtering parameters for quality control in step (3) are: retaining reads with a sequence length of 1300 bp to 1800 bp, an average Phred quality value of ≥10, and containing no more than 5 unidentified base Ns.

[0029] Preferably, the barcode splitting in step (3) includes: comparing the 5' end sequence of each read segment with the well barcode sequence library, and comparing the 3' end sequence with the plate barcode sequence library; when both ends of the barcode can uniquely match a pair of barcodes preset in the experiment, the read segment is assigned to the strain dataset of the corresponding well position; the comparison adopts an exact match or a fuzzy match strategy that allows no more than 2 base mismatches.

[0030] Preferably, the single-strain consistency sequence acquisition in step (3) includes: performing de novo clustering on all reads belonging to the same well position, and generating a full-length consistency sequence of the 16S rRNA gene for each cluster using the spoa and medaka algorithms; wherein, the number of effective reads for each well position used to generate the consistency sequence should be no less than 50, otherwise it is determined that the sequencing depth is insufficient and a consistency sequence is not generated for that well position.

[0031] Preferably, the species annotation in step (3) uses the BLAST algorithm to compare the obtained consistent sequences with the reference database and obtain species classification information based on the best comparison result. The comparison and judgment criteria are as follows: when the sequence consistency is ≥98.65%, it is identified to the species level; when the sequence consistency is between 95.0% and 98.65%, it is judged as a suspected new species within the genus; when the sequence consistency is <95%, it is judged as a suspected new species at the genus or higher classification level; the latter two cases are marked as requiring further verification.

[0032] Preferably, the reference database is a self-built database containing the 16S rRNA gene sequences of type strains of all known prokaryotic microbial species, constructed by integrating and filtering high-quality full-length 16S rRNA sequences from the NCBI database.

[0033] Preferably, the step (3) of obtaining the identification results of each strain in batch includes: automatically summarizing the consistency sequence of all well sites and their species annotation results according to the well site layout of the multi-well plate, generating a standardized identification report, which includes strain number, well site coordinates, consistency sequence, best comparison species name, sequence consistency, comparison coverage and expected value level information.

[0034] The present invention also provides a multi-well plate strain barcode labeling amplification kit for the method, comprising: a well barcode forward primer mixture and a plate barcode reverse primer mixture; the well barcode forward primer is formed by linking the well barcode sequence with a 27F primer sequence, and the plate barcode reverse primer is formed by linking the plate barcode sequence with a 1492R primer sequence.

[0035] Preferably, there are 96 well barcode sequences and 96 plate barcode sequences. The well barcode sequences are selected from the nucleotide sequences shown in SEQ ID NO.3 to SEQ ID NO.98, and the plate barcode sequences are selected from the nucleotide sequences shown in SEQ ID NO.99 to SEQ ID NO.194. The 27F primer sequence is shown in SEQ ID NO.1, and the 1492R primer sequence is shown in SEQ ID NO.2.

[0036] Preferably, the kit further comprises 2× PCR premix, magnetic bead purification reagent, and sequencing adapter ligation reagent.

[0037] The present invention also provides the application of the above-mentioned multi-strain barcode marker primer combination in the preparation of a high-throughput microbial identification kit, characterized in that the primer combination includes multiple well barcode forward primers and multiple plate barcode reverse primers. The combination of the well barcode forward primers and the plate barcode reverse primers is used to differentially label strains in different wells of a multi-well plate, so as to achieve simultaneous identification of strains in all wells of the multi-well plate in a single nanopore sequencing run.

[0038] Preferably, the well barcode forward primer is formed by linking the well barcode sequence with the 27F primer sequence, and the plate barcode reverse primer is formed by linking the plate barcode sequence with the 1492R primer sequence; there are 96 well barcode sequences and 96 plate barcode sequences, the well barcode sequences are selected from the nucleotide sequences shown in SEQ ID NO.3 to SEQ ID NO.98, and the plate barcode sequences are selected from the nucleotide sequences shown in SEQ ID NO.99 to SEQ ID NO.194.

[0039] This invention also provides a batch bacterial strain identification system based on nanopore sequencing, comprising:

[0040] (1) Barcode labeling amplification module: used for differential labeling and full-length amplification of the 16S rRNA gene of each strain in a multi-well plate, wherein the differential labeling adopts the paired-end barcode primer system as described in claim 1;

[0041] (2) Library construction module: used to mix multi-well plate amplification products and construct nanopore sequencing libraries;

[0042] (3) Sequencing module: used to perform nanopore sequencing on the library and perform preliminary quality control and data output in real time;

[0043] (4) Data analysis module: used to automatically complete further quality control of sequencing data, barcode splitting, consistency sequence generation and species annotation, and output a standardized report containing the identification results of strains in each well.

[0044] Preferably, the data analysis module employs an automated analysis process, specifically including:

[0045] (1) Quality control submodule: used to filter sequencing reads by length and quality;

[0046] (2) Barcode splitting submodule: used to determine the hole location of each reading segment based on the dual-end barcode sequence;

[0047] (3) Consistency sequence generation submodule: used to generate the full-length consistency sequence of the 16S rRNA gene from the set of reads belonging to each well site;

[0048] (4) Species annotation submodule: used to compare consistent sequences with reference databases to obtain species classification information;

[0049] (5) Report generation submodule: used to summarize the identification results of all holes and generate a standardized identification report.

[0050] The present invention provides a high-throughput microbial identification method based on full-length nanopore sequencing of 16S rRNA genes, comprising three main steps: multi-strain tagging library construction, nanopore sequencing, and bioinformatics analysis.

[0051] Step 1: Multi-strain tagging library construction

[0052] The multi-strain labeled library construction includes the following sub-steps:

[0053] (1.1) Design of a dual-end tag primer system: A set of well tag sequences (Forward Barcode, abbreviated as FB) were attached to the 5' end of the universal amplification primer 27F for bacterial 16S rRNA gene (SEQ ID NO.1, sequence AGAGTTTGATCCTGGCTCAG), and a set of plate tag sequences (Reverse Barcode, abbreviated as RB) were attached to the 5' end of the reverse primer 1492R (SEQ ID NO.2, sequence GGTTACCTTGTTACGACTT), to form well tag forward primers and plate tag reverse primers, respectively.

[0054] Preferably, the hole tag sequence is any one of the sequences shown in SEQ ID NO.3 to SEQ ID NO.98; the plate tag sequence is any one of the sequences shown in SEQ ID NO.99 to SEQ ID NO.194. All the above tag sequences are 15 bp in length, with a GC content of 40-60%, and are optimized to ensure sufficient sequence differences between them to avoid cross-identification.

[0055] (1.2) Multi-well plate culture and amplification: Single colonies to be identified were inoculated into 96-well plates for culture. The culture was used directly as a PCR template, or genomic DNA was extracted as a PCR template. One positive control well (using E. coli culture or DNA as a template) and one negative control well (using sterile deionized water as a template) were selected for each plate. In each well, the full-length 16S rRNA gene was amplified using the forward primer corresponding to the well label and the reverse primer corresponding to the plate label.

[0056] The PCR reaction system consisted of: 12.5 μL of 2× PCR Mix, 1 μL of forward primer (10 μM) for well label, 1 μL of reverse primer (10 μM) for plate label, 2 μL of template DNA (or colony suspension), and ddH2O to a final volume of 25 μL.

[0057] The PCR amplification program was as follows: 95℃ pre-denaturation for 3 min; 95℃ denaturation for 30 s, 58℃ annealing for 30 s, 72℃ extension for 90 s, for a total of 30 cycles; 72℃ final extension for 5 min; and storage at 4℃.

[0058] (1.3) Mixing of amplification products from multiple strains: Mix the PCR amplification products from all wells (or selected wells) of the same 96-well plate in equal volumes to form a plate-level mixed library. Further, plate-level mixed libraries from multiple plates can be mixed in equal volumes to form the final sequencing library.

[0059] (1.4) Library purification and quantification: The mixed library was purified by magnetic bead method, and the concentration and fragment size of the library were checked by Qubit fluorometer and agarose gel electrophoresis.

[0060] (1.5) Sequencing adapter ligation: The purified mixed library is repaired at the ends and ligated with a sequencing adapter compatible with the nanopore sequencing platform (e.g., the AMX adapter for the ONT platform or the sequencing adapter for the Puyi Bio platform), thus completing the sequencing library construction. If multiple plates of libraries from different environmental sample sources need to be sequenced in the same batch, the secondary barcode provided with the commercial sequencing kit can be added before or during adapter ligation.

[0061] Step 2: Nanopore sequencing

[0062] The nanopore sequencing includes the following sub-steps:

[0063] (2.1) Load the sequencing library constructed in step one into a nanopore sequencing chip (e.g., ONT MinION / GridION chip or PY-NFC003 sequencing chip from Puyi Biotechnology) and run sequencing on the nanopore sequencing platform.

[0064] (2.2) During the sequencing process, current signal data is collected in real time and converted into base sequence (fastq format) by base recognition software (such as ONT Dorado or PolyseqCtrl software).

[0065] (2.3) Set preliminary quality control parameters during base recognition: average Phred quality value Q≥10, sequence length≥1000bp. Sequences that do not meet the quality control conditions are filtered out during output.

[0066] (2.4) Perform library-level splitting of the sequencing data based on the secondary barcodes provided with the sequencing kit. During splitting, retain the secondary barcode sequences of all reads.

[0067] Step 3: Bioinformatics Analysis

[0068] The bioinformatics analysis uses nanopore sequencing and the aforementioned quality control and secondary barcode splitting output data as input, and includes the following sub-steps:

[0069] (3.1) Further quality control filtering: Perform secondary quality control on the sequences after the initial quality control, retain sequences with a length between 1300 bp and 1800 bp, and remove sequences with an average quality value Q < 10 or containing too many N bases.

[0070] (3.2) Plate label splitting: The sample attribution of each read is determined based on the paired label sequence. Specifically, the amplicon search function of seqkit is used to find regions that are highly similar to the plate label, and then the reads are assigned to the strain dataset of the corresponding well position by matching the well label and plate label for each well.

[0071] Preferably, the tag alignment employs an exact match or a fuzzy match strategy that allows ≤2 mismatches. For reads with a small number of misread bases in the tag sequence due to sequencing errors, fuzzy matching can still correctly assign them.

[0072] (3.3) Single-strain homogeneity sequence acquisition: All reads belonging to the same well are used to generate the full-length homogeneous sequence of the 16S rRNA gene of the strain using the NGSpeciesID tool. That is, each well's reads are first clustered into de novo clusters, and then the SPAA and medaka algorithms are used for each cluster to obtain the homogeneous sequence. For each well, the number of valid reads must be no less than 15; otherwise, it is considered that the sequencing depth is insufficient and no homogeneous sequence is generated.

[0073] (3.4) Species Annotation: The identical sequences obtained in step (3.3) are compared with reference databases (such as NCBI 16S rRNA database) using BLAST. Species annotation results are obtained based on the species classification information (kingdom, phylum, class, order, family, genus, species) of the best alignment result. Annotation parameters are as follows: when sequence identity is ≥98.65%, it is identified to the species level; when sequence identity is between 95.0% and 98.65%, it is judged as a suspected new species within the genus; when sequence identity is <95%, it is judged as a suspected new species at the genus or higher classification level; the latter two cases are marked as requiring further verification.

[0074] (3.5) Results summary: The consistency sequence and species annotation results of all well sites are automatically summarized by well site to generate a standardized identification report containing information such as strain number, well site number, consistency sequence, best comparison species, similarity, and coverage.

[0075] (III) Optimization of Technical Solutions

[0076] Furthermore, the length of the hole tag sequence and the plate tag sequence is 15 bp, the GC content is 40% to 60%, and the Hamming distance between any two tag sequences is not less than 6.

[0077] Furthermore, in the dual-ended labeled primer system, the well label and the plate label are interchangeable, meaning that the forward primer connecting to the plate label and the reverse primer connecting to the well label are equally applicable to this invention.

[0078] Furthermore, when the number of strains to be identified exceeds the combination capacity of a single tag set, a second level of sample labeling dimension can be added on top of the tag system by introducing secondary barcodes from commercial sequencing kits during library construction. Specifically, amplification products from multiple 96-well plates are ligated with different secondary barcodes and then mixed for sequencing, thereby increasing the sample throughput of a single sequencing run by tens of times.

[0079] Preferably, the secondary barcode is the Native Barcode (1-96) of the ONT platform or the Barcode (1-96) of the Puyi Bio platform.

[0080] Furthermore, the reference database mentioned in step (3.4) is a self-built database, which contains high-quality full-length 16S rRNA sequences of model strains obtained from public databases such as NCBI, after quality filtering.

[0081] Furthermore, the threshold requirements for obtaining the single strain homogeneous sequence in step (3.3) are: the number of valid reads is not less than 15, and the length of the homogeneous sequence is not less than 1400 bp.

[0082] Furthermore, the method further includes a manual verification step after step (3.4): strains identified at the genus level but not reaching the species level identification threshold are judged as suspected new species and further confirmed by phylogenetic tree construction or comparative genomics analysis and other polymorphic classification methods.

[0083] (iv) Beneficial effects

[0084] Compared with the prior art, the present invention has the following advantages:

[0085] (1) High throughput: This invention utilizes a combination of dual-end tag systems, employing N forward tags and M reverse tags to achieve N×M unique tag combinations. Taking the combination of 96 well tags and 96 plate tags as an example, a single sequencing run can simultaneously distinguish 9,216 strains from different wells, increasing throughput by more than two orders of magnitude compared to traditional Sanger sequencing, and by approximately 50-400 times compared to commercial nanopore barcoding schemes (12-96 layers). If commercial secondary barcoding is further used, the theoretical throughput can be increased by up to 96 times, i.e., 884,736 strains.

[0086] (2) Simple operation and high degree of automation: This invention uses a 96-well plate as the basic operating unit. All PCR amplifications are completed in parallel within the plate. After mixing, only one library construction and one sequencing run are needed to identify thousands to hundreds of thousands of strains, which greatly reduces the number of manual operation steps. Bioinformatics analysis adopts a standardized process and generates identification reports with one click, without the need for manual processing of peak diagrams and sequences for each strain.

[0087] (3) Low cost: By significantly increasing the sample throughput of a single sequencing run, the sequencing cost per strain is significantly reduced. Based on a single nanopore sequencing chip running 4,761 samples and obtaining 50 valid reads per sample, the total data volume is approximately 238,000 reads, which is far lower than the chip capacity. The cost of single-strain sequencing is reduced by more than 90% compared to Sanger sequencing.

[0088] (4) High degree of standardization: This invention provides a standardized solution for the entire process from sample processing, PCR amplification, library construction to sequencing and data analysis, eliminating operational differences between different operators and laboratories, and making the identification results have good reproducibility and comparability.

[0089] (5) High species resolution: This invention uses full-length 16S rRNA gene sequencing (about 1,500 bp), which contains more comprehensive species differentiation information than the short fragment amplicon of second-generation sequencing (usually 200-500 bp). Combined with the consistency sequence correction strategy, it can effectively overcome the defect of low single-time accuracy of nanopore sequencing and achieve high-accuracy species identification.

[0090] (6) Good compatibility: This invention is based on the universal 27F / 1492R amplification system, which is suitable for the amplification of 16S rRNA genes in most bacteria; the library construction and sequencing process is compatible with mainstream nanopore sequencing platforms such as ONT and Puyi Bio, and has good platform compatibility. Attached Figure Description

[0091] Figure 1 This is a schematic diagram of the overall process of the high-throughput microbial identification method of the present invention.

[0092] Figure 2 This is a schematic diagram of the dual-end label system of the present invention.

[0093] Figure 3 This is a graph illustrating the accuracy assessment of single-strain homogeneity sequences in an embodiment of the present invention.

[0094] Figure 4 This is a statistical graph showing the accuracy assessment of the homogeneity sequence of a single strain in an embodiment of the present invention.

[0095] Figure 5 This diagram illustrates the impact of different sequencing data volumes on the accuracy of the obtained consistent sequences, as presented in this invention.

[0096] Figure 6 The number of effective reads required to obtain an accurate and consistent sequence for a single strain, as calculated by this invention. Detailed Implementation

[0097] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0098] Example 1: Design and Synthesis of Two-Ended Tag Primers

[0099] This embodiment provides a set of paired-end tagged primer systems for high-throughput identification ( Figure 2 ).

[0100] Design principles:

[0101] • The tag sequence length is 15 bp, and the GC content is controlled between 40% and 60%;

[0102] • There is sufficient sequence dissimilarity between any two label sequences, with a Hamming distance of not less than 6;

[0103] • The 3' end of the tag sequence does not form a complementary or hairpin structure with the amplification primer sequence;

[0104] • The tag sequences themselves and with each other do not tend to form dimers.

[0105] Based on the above principles, this embodiment designs 96 forward tag sequences (hole tags, FB) (SEQ ID NO.3~SEQ ID NO.98) and 96 reverse tag sequences (plate tags, RB) (SEQ ID NO.99~SEQ ID NO.194). The specific composition of each tag sequence is as follows:

[0106] FB001: ACAGTCGATGCTACG (SEQ ID NO.3)

[0107] FB002: AGTCGATGCAGCTAC (SEQ ID NO.4)

[0108] FB003: ATCGACGTAGCTGAC (SEQ ID NO.5)

[0109] ... (To save space, the intermediate sequence is omitted here; the complete sequence is shown in the sequence listing) FB096:ACGTAGCTGATCGCT (SEQ ID NO.98)

[0110] RB001: GTACGATCGTCTGAC (SEQ ID NO.99)

[0111] RB002: CAGTGCTAGATCGTC (SEQ ID NO.100)

[0112] RB003: AGCTGATCGTACGTC (SEQ ID NO.101)

[0113] ... (To save space, the intermediate sequence is omitted here; see the sequence list for the complete sequence.)

[0114] RB096: GATCGTCTGACGTCT (SEQ ID NO.194)

[0115] Taking the ligation of FB001 and 27F primers as an example, the full sequence of the forward primer is: 5'-ACAGTCGATGCTACGAGAGTTTGATCCTGGCTCAG-3' (SEQ ID NO.141). The full sequence of the reverse primer is similarly the ligation product of the RB sequence and the 1492R sequence.

[0116] Example 2: Tagging and amplification of Strains in 96-well Plates and Library Construction

[0117] This embodiment uses the identification of environmentally isolated strains from 10 96-well plates (a total of 960 wells) as an example to illustrate the process of building a multi-strain labeled library.

[0118] 2.1 Strain culture and template preparation

[0119] The isolated and purified single colonies were inoculated into 96-well deep-well plates containing 200 μL of liquid culture medium (one colony per well) and cultured overnight with shaking under suitable conditions. One μL of the culture was used directly as a PCR template; genomic DNA extraction was not required.

[0120] For strains that are difficult to amplify directly, crude DNA can be prepared by thermal lysis or freeze-thaw method: take 100 μL of culture, heat at 95℃ for 10 min, centrifuge and take the supernatant as template.

[0121] 2.2 Full-length amplification of the 16S rRNA gene

[0122] Using a 96-well plate layout, each well was used with the corresponding well-labeled forward primer (FB001-FB096, assigned by column) and the plate-labeled reverse primer (RB001-RB010, assigned by plate) for PCR amplification.

[0123] PCR reaction system (25 μL):

[0124] ·2× PCR Mix (Vazyme): 12.5 μL

[0125] • Forward primer for well label (10 μM): 1 μL

[0126] • Plate-labeled reverse primer (10 μM): 1 μL

[0127] • Bacterial culture template: 2 μL

[0128] ddH2O: 8.5 μL

[0129] PCR amplification procedure:

[0130] • Pre-denaturation at 95℃: 3 min

[0131] Denaturation at 95℃: 30 s

[0132] Annealing at 58℃: 30 s × 30 cycles

[0133] • 72℃ extension: 90 s

[0134] · Final extension at 72℃: 5 min

[0135] ·Stored at 4℃

[0136] 2.3 Mixing and purification of amplification products

[0137] Take 5 μL of PCR product from each well and combine it into a 50 mL centrifuge tube. After mixing, take 200 μL of the mixed product. Purify the product using magnetic beads (AMPure XP magnetic beads, 1:1 volume ratio) and elute with 50 μL of sterile water to obtain a plate-grade mixed library.

[0138] Library concentration was determined using Qubit 4.0, and library fragment size was detected by 1% agarose gel electrophoresis (target band approximately 1,500 bp).

[0139] 2.4 Sequencing adapter ligation

[0140] Take 1 μg of the purified plate-grade mixed library and perform end repair using the PY-DLB103 kit from Puyi Biotechnology (20℃ for 5 min, 65℃ for 5 min). After purification with magnetic beads, ligate the ONT sequencing adapter using the PY-BLP103 kit from Puyi Biotechnology and react at 20℃ for 15 min. Elute the purified magnetic beads to 20 μL to obtain the final sequencing library.

[0141] Example 3: Nanopore Sequencing and Preliminary Data Processing

[0142] 3.1 Chip preparation and sequencing

[0143] Take an ONT R9.4.1 sequencing chip or a Puyi Bio PY-NFC003 sequencing chip and perform chip quality testing according to the manufacturer's instructions. After confirming that the number of usable wells on the chip is acceptable, add storage buffer to the chip and let it stand for 5 minutes.

[0144] The sequencing library prepared in Example 2 was added to the chip through the sample well. The sequencing program was started, the sequencing duration was set to 24 hours, and the base recognition mode was set to high precision mode.

[0145] 3.2 Real-time base identification and preliminary quality control

[0146] During sequencing, the current signal is converted into a base sequence using Dorado software (ONT platform) or the sequencer's associated base recognition software. The base recognition parameters are set as follows:

[0147] • Minimum average quality score (Q-score): ≥10

[0148] Minimum sequence length: 1,000 bp

[0149] Sequences that meet the quality control criteria are output in real time in FastQ format. In this example, the mixed library of 10 plates (960 strains) was sequenced for 24 hours, and a total of about 663,527 sequences that passed the quality control were obtained, with an average of about 691 reads per strain, which is far higher than the minimum threshold (15 reads) required for the construction of consistent sequences.

[0150] Example 4: Bioinformatics Analysis

[0151] 4.1 Further quality control filtration

[0152] Secondary quality control was performed on the fastq files output by Dorado, using NanoFilt software for filtering:

[0153] • Retain reads with sequence lengths between 1,300 bp and 1,800 bp;

[0154] • Retain reads with an average quality value Q ≥ 10;

[0155] • Remove reads containing ≥5 unidentified bases (N).

[0156] 4.2 Label Splitting and Sample Attribution

[0157] Construct a tag sequence database containing complete sequences of all well tags (FB001-FB096) and plate tags (RB001-RB010). For each read, use seqkit to search for matching tag sequences.

[0158] Matching rules: exact match or fuzzy match with ≤2 mismatches allowed; the 5' end and 3' end of a read must uniquely match a hole label and a board label respectively; the matched hole label and board label combination must correspond to the preset hole position (i.e., the combination has actually been used in this batch).

[0159] Reads meeting the above criteria were assigned to the corresponding plate-well strain dataset. In this example, approximately 92.3% of the post-quality control reads successfully underwent tag delimitation.

[0160] 4.3 Generation of Uniform Sequences from Single Strains

[0161] For all reads (at least 15) collected in each well, the full-length consensus sequence of the 16S rRNA gene of the strain is generated using the NGSpeciesID tool. That is, each well's reads are first clustered into de novo clusters, and then the consensus sequence is obtained for each cluster using the SPAA and medaka algorithms. The full-length consensus sequence of the 16S rRNA gene of the strain representing that well is generated, and the consensus sequence FASTA file is output. The number of reads involved in the consensus sequence acquisition, tags, and other information are recorded.

[0162] In this example, out of 960 wells, after removing invalid wells (142 wells) with no bacterial growth or no PCR product, 774 wells successfully generated high-quality homologous sequences (≥1,400 bp in length), accounting for 94.6% of the valid wells. 49 wells did not obtain accurate homologous sequences due to insufficient number of valid reads (<15) or low sequence quality.

[0163] 4.4 Species Annotation

[0164] The obtained consistent sequences were compared with the NCBI 16S rRNA reference database using BLAST: alignment software: BLAST+ 2.10.0; parameters: -evalue 1e-5 -max_target_seqs 10 -outfmt 6; similarity ≥ 98.65% was considered species-level identification; similarity ≤ 95% < 98.65% was considered a suspected new species of the genus; similarity < 95% was considered a new genus or higher-level new species, requiring further verification.

[0165] For each sample, the best alignment result is used to extract species classification information (kingdom, phylum, class, order, family, genus, species), and the similarity, alignment coverage, and alignment score are recorded.

[0166] 4.5 Results Summary and Report Generation

[0167] The system automatically generates a summary table of identification results containing the following fields: plate number, well number; strain number; number of valid reads; best comparison species; species taxonomic lineage; similarity; coverage; expected value (E value). The final results are output in both Excel and PDF report formats. The results report is shown in Table 1.

[0168] Table 1

[0169] Example 5: Effect Verification and Comparison

[0170] 5.1 Accuracy Verification

[0171] To verify the accuracy of the identification method of the present invention, 59 microbial strains (covering 16 genera and 45 species) isolated from litchi fruits and rhizosphere soil were selected and identified using the method of the present invention and Sanger sequencing (paired-end sequencing + manual splicing and alignment).

[0172] The results showed that the method of this invention can accurately identify all test strains. At an effective sequencing depth (≥15 reads), the species-level identification accuracy is 100%, which is completely consistent with Sanger sequencing. Figure 3(See Table 2). Furthermore, the sequences obtained by the method of this invention exhibit high consistency with those obtained by Sanger sequencing, with the consistency between the two methods ranging from 99.64% to 100%. Specifically, 43 strains achieved 100% sequence consistency between the two methods, 14 strains achieved 99.85%–99.95% sequence consistency, and two other strains achieved 99.64% and 99.78% sequence consistency, respectively, with the two methods yielding 99.78% and 99.64% sequence consistency, respectively. Figure 4 Average identification time: Sanger sequencing takes about 4 days, while the method of this invention takes about 36 hours.

[0173] Table 2. Consistent sequences obtained from nanopore sequencing and Sanger sequencing ; ;

[0174] 5.2 Determination of the minimum effective number of reads per hole in the method of the present invention

[0175] To determine the minimum number of effective reads per well required for accurate identification results, the sequencing libraries constructed from the aforementioned 59 strains were further sequenced until 8.95 G of high-depth data was obtained (the unsequential data is denoted as A.fastq). Different sequencing volumes were simulated by randomly selecting 80%, 50%, 20%, 10%, 5%, 1%, 0.5%, and 0.1% of the data from the A.fastq file. Simultaneously, each sequencing volume was independently sampled 10 times. The data obtained from these randomly selected sequencing volumes were analyzed using the method described in this invention to obtain the homologous sequences for each well, and the number of effective reads per well was recorded. The obtained homologous sequences were compared with the Sanger sequencing sequences, and the alignment values ​​were recorded. A scatter plot was plotted with the number of effective reads per well as the x-axis and the homologous sequence and Sanger sequencing sequence alignment values ​​as the y-axis. The scatter distribution was observed by scaling the data range on the x-axis.

[0176] The results showed that as the number of valid reads per well increased, the consistency value between the consistent sequence and the Sanger sequencing sequence increased ( Figure 5 By scaling the data range on the horizontal axis, it was found that a number of valid reads ≥ 15 per well ensured more than 98.65% sequence identity with the Sanger sequencing sequence, resulting in accurate species-level identification. However, when the number of valid reads per well was < 15, 91 wells (12.8%) showed partial sequences with less than 98.65% sequence identity with the Sanger sequencing sequence. Figure 6 Therefore, the minimum effective number of reads per hole determined by the method of the present invention is 15.

[0177] 5.3 Flux Comparison

[0178] Table 3 compares the identification of strains isolated from litchi fruits and rhizosphere soil in 960 wells using the traditional Sanger sequencing method and the method of this invention:

[0179] Table 3. Throughput comparison of large-scale bacterial isolates processed by Sanger sequencing method and the method of this invention.

[0180] As can be seen from the table above, the present invention is significantly superior to the traditional Sanger sequencing method in terms of throughput, efficiency, cost, and standardization.

[0181] Example 6: Secondary Barcode Extended Flux

[0182] When a single tagging system (96×96=9,216 combinations) is still insufficient to meet the needs of larger-scale identification, throughput can be further expanded by introducing secondary barcodes that come with commercial sequencing kits.

[0183] This embodiment uses 96-well plates as the unit, with each plate using a unique secondary barcode (Puyi Bio PY-DTB103, NB01-NB96). The amplification products from 10 96-well plates (960 strains) were ligated with different secondary barcodes and then merged for library construction. After sequencing data was processed, reads were first separated into individual plates based on the secondary barcodes, and then further separated into individual wells based on the paired-end tags (FB+RB).

[0184] With the adoption of secondary barcode expansion, the theoretical throughput of a single sequencing run can reach 9,216 × 96 = 884,736 samples (limited by the total output throughput of the sequencing chip), which fully meets the identification needs of ultra-large cultinomy projects.

[0185] Industrial applicability

[0186] This invention provides a high-throughput microbial identification method based on full-length nanopore sequencing of 16S rRNA genes. It has significant advantages such as high throughput, low cost, standardization, and high degree of automation. It can be widely used in fields such as culture omics, environmental microbial resource mining, clinical pathogen identification, food microbial detection, and agricultural microbial resource preservation. It has good industrial application value and broad industrialization prospects.

[0187] The following are sequences from SEQ ID NO.1 to SEQ ID NO.194: SEQ ID NO.1: AGAGTTTGATCCTGGCTCAG SEQ ID NO.2: GGTTACCTTGTTACGACTT SEQ ID NO.3: ACAGTCGATGCTACG SEQ ID NO.4: AGTCGATGCAGCTAC SEQ ID NO.5: ATCGACGTAGCTGAC SEQ ID NO.6: ACGTAGCTGATCGAC SEQ ID NO.7: AGCTGATCGACGTAG SEQ ID NO.8: ATCGACGTAGCTGAT SEQ ID NO.9: ACGTAGCTGATCGTA SEQ ID NO.10: AGCTGATCGTACGAT SEQ ID NO.11: ATCGATGCTAGCTGA SEQ ID NO.12: ACGATGCTAGCTGAC SEQ ID NO.13: AGCTACGTAGCTGAT SEQ ID NO.14: ATCGTACGATCGTAC SEQ ID NO.15: ACGATCGTACGATGC SEQ ID NO.16: AGTCGATCGTACGAT SEQ ID NO.17: ATGCTAGCTGATCGA SEQ ID NO.18: ACGTAGCGATCGTAC SEQ ID NO.19: AGCTAGCGATCGTAC SEQ ID NO.20: ATCGACGATGCTAGC SEQ ID NO.21: ACGATCGACGTAGCT SEQ ID NO.22: AGTCGACGTAGCTAG SEQ ID NO.23: ATGCTACGTAGCTGC SEQ ID NO.24: ACGTAGCTACGATGC SEQ ID NO.25: AGCTACGATCGTGCT SEQ ID NO.26: ATCGATGCTACGTGC SEQ ID NO.27: ACGATGCTACGTAGC SEQ ID NO.28: AGTCGATGCTACGTG SEQ ID NO.29: ATGCTAGCTACGATG SEQ ID NO.30: ACGTAGCTGATGCTG SEQ ID NO.31: AGCTAGCTGATCGTG SEQ ID NO.32: ATCGACGTAGCTGCT SEQ ID NO.33: ACGATCGTAGCTGCT SEQ ID NO.34: AGTCGATCGTAGCTG SEQ ID NO.35: ATGCTACGATCGTGC SEQ ID NO.36: ACGTAGCGATGCTAC SEQ ID NO.37: AGCTAGCGATGCTAC SEQ ID NO.38: ATCGACGATCGTGCT SEQ ID NO.39: ACGATCGACGTGCTA SEQ ID NO.40: AGTCGACGTAGCTGA SEQ ID NO.41: ATGCTAGCGACGTAG SEQ ID NO.42: ACGTAGCTGACGTAG SEQ ID NO.43: AGCTGATCGACGTCT SEQ ID NO.44: ATCGTACGATCGTGC SEQ ID NO.45: ACGATCGTACGTGCT SEQ ID NO.46: AGTCGATCGTACGCT SEQ ID NO.47: ATGCTAGCTGACGTC SEQ ID NO.48: ACGTAGCTGCTCGCT SEQ ID NO.49: AGCTAGCTGCTCGCT SEQ ID NO.50: AGTCGATGCTAGCTG SEQ ID NO.51: ATGCTACGTAGCGTC SEQ ID NO.52: ACGTAGCTACGTGCT SEQ ID NO.53: AGCTACGATGCTGCT SEQ ID NO.54: ATCGATGCTAGCGTC SEQ ID NO.55: ACGATGCTAGCGTCT SEQ ID NO.56: AGTCGATGCTAGCGT SEQ ID NO.57: ATGCTAGCTGATCGC SEQ ID NO.58: ACGTAGCGATCGTCT SEQ ID NO.59: AGCTAGCGATCGTCT SEQ ID NO.60: ATCGACGATGCTAGT SEQ ID NO.61: ACGATCGACGTAGTG SEQ ID NO.62: AGTCGACGTAGCTGT SEQ ID NO.63: ATGCTACGTAGCTGT SEQ ID NO.64: ACGTAGCTACGATGT SEQ ID NO.65: AGCTACGATCGTGTA SEQ ID NO.66: ATCGATGCTACGTGT SEQ ID NO.67: ACGATGCTACGTAGT SEQ ID NO.68: AGTCGATGCTACGTA SEQ ID NO.69: ATGCTAGCTACGATC SEQ ID NO.70: ACGTAGCTGATCGTC SEQ ID NO.71: AGCTAGCTGATCGAC SEQ ID NO.72: ATCGACGTAGCTGTA SEQ ID NO.73: ACGATCGTAGCTGTA SEQ ID NO.74: AGTCGATCGTAGCTA SEQ ID NO.75: ATGCTACGATCGTGT SEQ ID NO.76: ACGTAGCGATGCTGT SEQ ID NO.77: AGCTAGCGATGCTGT SEQ ID NO.78: ATCGACGATCGTGTA SEQ ID NO.79: ACGATCGACGTGCTG SEQ ID NO.80: AGTCGACGTAGCTGC SEQ ID NO.81: ATGCTAGCGACGTCT SEQ ID NO.82: ACGTAGCTGACGTCT SEQ ID NO.83: AGCTGATCGACGTGC SEQ ID NO.84: ATCGTACGATCGTGT SEQ ID NO.85: ACGATCGTACGTGTA SEQ ID NO.86: AGTCGATCGTACGTG SEQ ID NO.87: ATGCTAGCTGACGTG SEQ ID NO.88: ACGTAGCTGCTCGTG SEQ ID NO.89: AGCTAGCTGCTCGTG SEQ ID NO.90: AGTCGATGCTAGCTA SEQ ID NO.91: ATGCTACGTAGCGTG SEQ ID NO.92: ACGTAGCTACGTGTG SEQ ID NO.93: AGCTACGATGCTGTG SEQ ID NO.94: ATCGATGCTAGCGTG SEQ ID NO.95: ACGATGCTAGCTGTA SEQ ID NO.96: AGCTGATCGTCTGAC SEQ ID NO.97: ATCGACGTAGCGTAC SEQ ID NO.98: ACGTAGCTGATCGCT SEQ ID NO.99: GTACGATCGTCTGAC SEQ ID NO.100: CAGTGCTAGATCGTC SEQ ID NO.101: AGCTGATCGTACGTC SEQ ID NO.102: TGACGATCGTCTAGC SEQ ID NO.103: GTCAGCTAGATCGTC SEQ ID NO.104: CTGATCGTACGAGTC SEQ ID NO.105: ATCGAGTCGTAGCTC SEQ ID NO.106: GCTAGATCGTCGACT SEQ ID NO.107: CGTAGATCGTCTGAC SEQ ID NO.108: AGTCGATCGTCTAGC SEQ ID NO.109: TGCGATCTAGCTGAC SEQ ID NO.110: GATCGTCAGCTAGTC SEQ ID NO.111: CTAGAGTCGTCTGAC SEQ ID NO.112: ATCGTGCTAGCTGAC SEQ ID NO.113: GTCGATCTAGCGTAC SEQ ID NO.114: GATCGTCTGACGTCT SEQ ID NO.115: AGCTCGATCGTCTAG SEQ ID NO.116: TGCTAGAGTCGTCTC SEQ ID NO.117: GATCGACGTAGCTCT SEQ ID NO.118: CGTAGAGTCTCGATC SEQ ID NO.119: AGTCGACGTAGCTCT SEQ ID NO.120: TGCGTAGCTCTGATC SEQ ID NO.121: GTCGATCGTCTAGAC SEQ ID NO.122: CTAGCGATCTGACGT SEQ ID NO.123: ATCGAGCTCGTCTAG SEQ ID NO.124: GCTAGTCGATCTGAC SEQ ID NO.125: CAGATCGTCTGACGT SEQ ID NO.126: AGCTGACGTAGCTCT SEQ ID NO.127: TGATCGTCGAGCTCT SEQ ID NO.128: GATCGTCTGACGTAG SEQ ID NO.129: CGTAGCTCTGATCGT SEQ ID NO.130: AGTCGTCTAGACGTC SEQ ID NO.131: TGCTCGATCTAGACG SEQ ID NO.132: GTCAGAGTCTCGATC SEQ ID NO.133: CTGACGTAGCTCTCG SEQ ID NO.134: ATCGTCTAGACGTCG SEQ ID NO.135: GCTCGATCTGACGTA SEQ ID NO.136: CAGCTCTCGATCGTC SEQ ID NO.137: AGCGATCTGACGTAG SEQ ID NO.138: TGATCGAGTCTCGTC SEQ ID NO.139: GATCTGACGTAGCTC SEQ ID NO.140: CGTCTAGACGTCGAT SEQ ID NO.141: AGCTCTGATCGTCGA SEQ ID NO.142: TGCGAGTCTCGATCT SEQ ID NO.143: GTCGAGCTCTCGATC SEQ ID NO.144: CTAGACGTCGATCTG SEQ ID NO.145: ATCGTCGATCTGACG SEQ ID NO.146: GCTCTGATCGTCGAG SEQ ID NO.147: CAGTCTCGATCTGAC SEQ ID NO.148: AGCGTCGATCTGACT SEQ ID NO.149: TGACGTAGCTCTGAT SEQ ID NO.150: GATAGTCGTCTAGAC SEQ ID NO.151: CGTCGATCTGACTAG SEQ ID NO.152: AGCTGATCGTCGAGT SEQ ID NO.153: TGCTCTCGATCGTCT SEQ ID NO.154: GTCCTAGACGTCGAT SEQ ID NO.155: CTAGTCTCGATCGTC SEQ ID NO.156: ATCGATCTGACGTAG SEQ ID NO.157: GCTGATCGTCGAGTC SEQ ID NO.158: CAGATCTGACGTAGC SEQ ID NO.159: AGTCCTAGACGTCGT SEQ ID NO.160: TGATCTGACGTAGCT SEQ ID NO.161: GATCGAGTCTCGATC SEQ ID NO.162: CGTTAGCTCTGATCG SEQ ID NO.163: AGCTAGTCGTCTAGA SEQ ID NO.164: TGCGATCGTCTAGAC SEQ ID NO.165: GTCGATCTGACTAGC SEQ ID NO.166: CTGATCGTCGAGTCT SEQ ID NO.167: ATCGTAGCTCTGATC SEQ ID NO.168: GCTAGTCTCGATCGT SEQ ID NO.169: CAGCGTCGATCTGAT SEQ ID NO.170: AGTCGATCTGACTAG SEQ ID NO.171: TGCTAGACGTCGATC SEQ ID NO.172: GATCTCGATCGTCTG SEQ ID NO.173: CGTTAGACGTCGATC SEQ ID NO.174: AGCTCGTCTAGACGT SEQ ID NO.175: TGATCGTCTAGACGT SEQ ID NO.176: GTCCTGATCGTCGAG SEQ ID NO.177: CTAGATCGTCGAGTC SEQ ID NO.178: ATCGAGTCTCGATCT SEQ ID NO.179: GCTCGTCTAGACGTC SEQ ID NO.180: CAGATCGTCGAGTCT SEQ ID NO.181: AGCGATCGTCTAGAC SEQ ID NO.182: TGACGAGTCTCGATC SEQ ID NO.183: GATCGTCTAGACGTC SEQ ID NO.184: CGTTCTGATCGTCGA SEQ ID NO.185: AGCTGTCGATCTGAC SEQ ID NO.186: GTCGATCGTCTGACT SEQ ID NO.187: GTCGAGTCTCGATCT SEQ ID NO.188: CTGATCGTCTAGACG SEQ ID NO.189: TGCGATCTGACGTCT SEQ ID NO.190: GCTGTCGATCTGACG SEQ ID NO.191: CAGCTGATCGTCGAG SEQ ID NO.192: CGTTGATCGTCGAGT SEQ ID NO.193: TGCTATCGTCGAGTC SEQ ID NO.194: GATCGTCTGACGTCT。

Claims

1. A high-throughput microbial identification method based on full-length nanopore sequencing of 16S rRNA genes, characterized in that, Includes the following steps: (1) Multi-strain barcode labeling library construction: The full-length 16S rRNA gene of multiple strains cultured in a multi-well plate was amplified by PCR using a paired-end barcode primer system. The paired-end barcode primer system includes a well barcode forward primer and a plate barcode reverse primer. The well barcode forward primer is composed of the 5' end of the universal 16S rRNA gene forward primer 27F connected to the well barcode sequence, and the plate barcode reverse primer is composed of the 5' end of the universal 16S rRNA gene reverse primer 1492R connected to the plate barcode sequence. By assigning different well barcode forward primers to different wells of the same multi-well plate and different plate barcode reverse primers to different multi-well plates, independent labeling of strains in each well is achieved. The amplification products of each well are mixed to construct a nanopore sequencing library. (2) Nanopore sequencing: Perform nanopore sequencing on the sequencing library constructed in step (1) to obtain sequencing data; (3) Bioinformatics analysis: The sequencing data obtained in step (2) are subjected to quality control, barcode splitting, single strain consistency sequence acquisition and species annotation to obtain the identification results of each strain in batches.

2. The method according to claim 1, characterized in that, The lengths of the perforated barcode sequence and the plate barcode sequence in step (1) are both 12 to 24 bp, the GC content is 40% to 60%, and the Hamming distance between any two perforated barcode sequences and between any two plate barcode sequences is not less than 6.

3. The method according to claim 2, characterized in that, The well barcode sequence and plate barcode sequence each consist of 96 sequences. The well barcode sequence is selected from the nucleotide sequences shown in SEQ ID NO.3 to SEQ ID NO.98, and the plate barcode sequence is selected from the nucleotide sequences shown in SEQ ID NO.99 to SEQ ID NO.

194. The nucleotide sequence of the forward primer 27F is shown in SEQ ID NO.1, and the nucleotide sequence of the reverse primer 1492R is shown in SEQ ID NO.

2.

4. The method according to claim 1, characterized in that, The multi-well plate is a 96-well plate, with each well corresponding to an independent combination of well barcode and plate barcode. Different wells within the same 96-well plate are distinguished by using the same plate barcode reverse primer and different well barcode forward primers, or by using the same well barcode forward primer and different plate barcode reverse primers. The barcode combination corresponding to each well remains unique throughout the entire experimental system.

5. The method according to claim 1, characterized in that, The nanopore sequencing described in step (1) includes: mixing equal volumes of PCR amplification products from each well, purifying them with magnetic beads, repairing the ends, and then ligating sequencing adapters to obtain the final sequencing library; when it is necessary to perform batch sequencing on amplification products from multiple multi-well plates from different environmental samples, the secondary library barcodes provided by a commercial nanopore sequencing kit are introduced for marking and differentiation when or before ligating the sequencing adapters.

6. The method according to claim 1, characterized in that, The PCR reaction system for full-length PCR amplification of the 16S rRNA gene in step (1) is as follows: 12.5 μL of 2× PCR premix, 1 μL of 10 μM barcode forward primer, 1 μL of 10 μM plate barcode reverse primer, 2 μL of template DNA or colony suspension, and ddH2O to 25 μL; the PCR amplification program is as follows: 95℃ pre-denaturation for 3 min; 95℃ denaturation for 30 s, 58℃ annealing for 30 s, 72℃ extension for 90 s, for a total of 30 cycles; 72℃ final extension for 5 min; preferably, the nanopore sequencing in step (2) performs real-time preliminary quality control during base recognition, filtering out reads with an average Phred quality value <10 or a sequence length <1000 bp.

7. The method according to claim 1, characterized in that, The further filtering parameters for quality control in step (3) are: retaining reads with a sequence length of 1300 bp to 1800 bp, an average Phred quality value ≥10, and containing no more than 5 unidentified N bases; preferably, the barcode splitting in step (3) includes: comparing the 5' end sequence of each read with the well barcode sequence library, and comparing the 3' end sequence with the plate barcode sequence library; when both ends of the barcode can uniquely match a pair of barcodes preset in the experiment, the read is assigned to the strain dataset of the corresponding well position; the comparison adopts an exact match or a fuzzy match strategy that allows no more than 2 base mismatches; preferably, the single strain consistency sequence acquisition in step (3) includes: performing de novo clustering on all reads belonging to the same well position, and generating 16S sequences for each cluster using the SPOA and Medaka algorithms. The full-length rRNA gene sequence is consistent; wherein, the number of effective reads for each well used to generate the consistent sequence should not be less than 15, otherwise it is judged as insufficient sequencing depth and no consistent sequence is generated for that well. Preferably, the species annotation in step (3) adopts the BLAST algorithm, and the obtained consistent sequence is compared with the reference database. The species classification information is obtained according to the best comparison result. The comparison and judgment criteria are: when the sequence consistency is ≥98.65%, it is identified to the species level; when the sequence consistency is between 95.0% and 98.65%, it is judged as a suspected new species in the genus; when the sequence consistency is <95%, it is judged as a suspected new species at the genus or above classification level. The latter two cases are marked as needing further verification. Preferably, the reference database is a self-built database containing the 16S rRNA gene sequences of the type strains of all known prokaryotic microbial species, which are obtained from the high-quality full-length 16S rRNA gene sequences of the NCBI database. The rRNA sequence is constructed after integration and filtering; preferably, the batch acquisition of identification results of each strain in step (3) includes: automatically summarizing the consistency sequence of all wells and its species annotation results according to the well layout of the multi-well plate, generating a standardized identification report, which includes strain number, well coordinates, consistency sequence, best comparison species name, sequence consistency, comparison coverage and expected value level information.

8. A multi-well plate strain barcode labeling amplification kit for use in the method of any one of claims 1-6, characterized in that, The kit comprises: a mixture of forward primers for well barcodes and a mixture of reverse primers for plate barcodes; the forward primers for well barcodes are formed by linking well barcode sequences with 27F primer sequences, and the reverse primers for plate barcodes are formed by linking plate barcode sequences with 1492R primer sequences; preferably, there are 96 well barcode sequences and 96 plate barcode sequences, the well barcode sequences are selected from the nucleotide sequences shown in SEQ ID NO.3 to SEQ ID NO.98, and the plate barcode sequences are selected from the nucleotide sequences shown in SEQ ID NO.99 to SEQ ID NO.194; the 27F primer sequence is shown in SEQ ID NO.1, and the 1492R primer sequence is shown in SEQ ID NO.2; preferably, the kit also comprises 2× PCR premix, magnetic bead purification reagent, and sequencing adapter ligation reagent.

9. The application of a multi-strain barcode-labeled primer combination in the preparation of a high-throughput microbial identification kit, characterized in that, The primer combination comprises multiple well-barcode forward primers and multiple plate-barcode reverse primers. The combination of these primers is used to differentially label strains in different wells of a multi-well plate, enabling simultaneous identification of strains in all wells of the multi-well plate during a single nanopore sequencing run. Preferably, the well-barcode forward primers are formed by linking a well barcode sequence with a 27F primer sequence, and the plate-barcode reverse primers are formed by linking a plate barcode sequence with a 1492R primer sequence. There are 96 well-barcode sequences and 96 plate-barcode sequences. The well-barcode sequences are selected from the nucleotide sequences shown in SEQ ID NO. 3 to SEQ ID NO. 98, and the plate-barcode sequences are selected from the nucleotide sequences shown in SEQ ID NO. 99 to SEQ ID NO.

194.

10. A batch bacterial strain identification system based on nanopore sequencing, characterized in that, include: (1) Barcode labeling amplification module: used for differential labeling and full-length amplification of the 16S rRNA gene of each strain in a multi-well plate, wherein the differential labeling adopts the paired-end barcode primer system as described in claim 1; (2) Library construction module: used to mix multi-well plate amplification products and construct nanopore sequencing libraries; (3) Sequencing module: used to perform nanopore sequencing on the library and perform preliminary quality control and data output in real time; (4) Data analysis module: used to automatically complete further quality control of sequencing data, barcode splitting, consistency sequence generation and species annotation, and output a standardized report containing the identification results of strains in each well; Preferably, the data analysis module employs an automated analysis process, specifically including: (1) Quality control submodule: used to filter sequencing reads by length and quality; (2) Barcode splitting submodule: used to determine the hole location of each reading segment based on the dual-end barcode sequence; (3) Consistency sequence generation submodule: used to generate the full-length consistency sequence of the 16S rRNA gene from the set of reads belonging to each well site; (4) Species annotation submodule: used to compare consistent sequences with reference databases to obtain species classification information; (5) Report generation submodule: used to summarize the identification results of all holes and generate a standardized identification report.

Images