Pathogen identification and drug resistance prediction method and system based on metagenomic sequencing
By combining metagenomic sequencing with machine learning models, the identification of pathogens and the prediction of drug resistance have been automated and standardized. This has solved the detection limitations of traditional methods and the complexity of interpreting mNGS data, improved the timeliness and accuracy of infectious disease diagnosis, and provided precise medication recommendations for clinicians.
Patent Information
- Application Number
- CN202610791404.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-06-03
- Publication Date
- 2026-08-25
AI Technical Summary
Traditional pathogen diagnostic methods are time-consuming, have a narrow detection range, and low positive rates. In clinical applications, mNGS technology involves complex data analysis and relies on human experience for result interpretation, making it impossible to directly predict drug resistance phenotypes, which leads to serious problems with antibiotic resistance.
By combining metagenomic sequencing with machine learning models, we can automate and standardize pathogen identification, drug resistance phenotype prediction, and clinical treatment recommendations, and integrate multidimensional feature inputs to generate structured clinical reports.
It has achieved full automation and standardization from sample to clinical report, which has improved the timeliness and accuracy of infectious disease diagnosis, reduced host background interference, improved the sensitivity of low-abundance pathogen detection and the accuracy of drug resistance phenotype prediction, and provided precise clinical medication guidance.
Smart Images
Figure CN122637894A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of clinical infectious disease diagnosis technology, and more specifically to a method and system for pathogen identification and drug resistance prediction based on metagenomic sequencing. Background Technology
[0002] Infectious diseases are a leading cause of intensive care unit (ICU) hospitalizations and deaths worldwide. Clinically, for patients with fever of unknown origin, sepsis, severe pneumonia, etc., rapid and accurate identification of the causative pathogen and guidance of precise anti-infective therapy are crucial to prognosis. However, traditional pathogen diagnostic methods, such as microbial culture, serological testing, and PCR, have many limitations: they are time-consuming (culture can take several days), have a narrow detection range (only for pre-specified pathogens), and have low positive rates (reports indicate that pathogens cannot be found in up to 60% of pneumonia cases and more than 30% of sepsis cases).
[0003] Metagenomic next-generation sequencing (mNGS) technology can unbiasedly detect the nucleic acid sequences of all microorganisms (bacteria, viruses, fungi, and parasites) in a sample, providing a revolutionary tool for comprehensive and rapid pathogen identification. However, mNGS technology still faces challenges in clinical applications: 1) The data analysis process is complex and requires high levels of bioinformatics expertise; 2) The interpretation of results relies on human experience and is difficult to standardize; 3) It can only provide information on pathogen types and cannot directly predict their drug resistance phenotypes. Clinicians still need to combine drug sensitivity tests or experience to select antibiotics, leading to widespread empirical drug use and exacerbating the problem of antimicrobial resistance.
[0004] Therefore, there is an urgent need for a system that can deeply integrate mNGS wet testing with intelligent dry analysis to achieve full automation and standardization from sample to clinical report, and integrate drug resistance gene and phenotypic prediction functions, so as to improve the timeliness, accuracy and clinical operability of infectious disease diagnosis. Summary of the Invention
[0005] The purpose of this invention is to provide a method and system for pathogen identification and drug resistance prediction based on metagenomic sequencing. By extracting multidimensional features through metagenomic sequencing alignment and combining them with machine learning models, pathogen identification, drug resistance phenotype prediction, and clinical treatment recommendations can be generated.
[0006] To achieve the above objectives, the present invention provides the following technical solution: A method for pathogen identification and drug resistance prediction based on metagenomic sequencing includes the following steps: S1. Obtain clinical samples, perform metagenomic sequencing to obtain raw sequencing data, and preprocess the raw sequencing data; wherein the preprocessing includes: quality control and host nucleic acid removal; S2. The preprocessed sequencing data is compared with the pathogen reference database to identify the types and relative abundance of pathogens in clinical samples. S3. The preprocessed sequencing data is compared with the drug resistance gene database and the virulence factor database to identify the drug resistance genes and virulence factors carried by pathogens in clinical samples. S4. Using at least one of the pathogen type, relative abundance, drug resistance gene and patient clinical information as input features, a trained machine learning model is used to predict the drug resistance phenotype of the pathogen to a specific antibiotic. S5. Integrate the pathogen types and relative abundance, drug resistance gene spectrum, virulence factors, and predicted drug resistance phenotypes to generate and output a structured clinical report containing treatment recommendations.
[0007] Furthermore, in S1, the clinical sample includes at least one of blood, cerebrospinal fluid, bronchoalveolar lavage fluid, tissue, urine, or feces.
[0008] Furthermore, in S4, the patient's clinical information includes at least one of the following: site of infection, underlying disease, and patient age.
[0009] Furthermore, in S4, the resistance phenotype of the pathogen to a specific antibiotic includes: the pathogen's sensitivity classification to the specific antibiotic and the minimum inhibitory concentration trend.
[0010] Furthermore, in S4, the training process of the machine learning model is specifically as follows: Collect genomic sequencing data of historical clinical isolates, corresponding in vitro drug susceptibility test results, and patient clinical information to construct a training dataset; Species information, presence or absence and copy number of drug resistance genes, virulence factors and core genome single nucleotide polymorphism features are extracted from the genome sequencing data, and input feature vectors are generated by combining them with patient clinical information. Using the results of the in vitro drug sensitivity test as labels, the machine learning model is trained using random forest, gradient boosting tree, or deep neural network algorithms.
[0011] This invention also provides a pathogen identification and drug resistance prediction system based on metagenomic sequencing, used to perform a pathogen identification and drug resistance prediction method based on metagenomic sequencing, comprising: The sample processing and sequencing module is used to acquire clinical samples, perform metagenomic sequencing to obtain raw sequencing data, and preprocess the raw sequencing data; wherein the preprocessing includes: quality control and host nucleic acid removal. The intelligent bioinformatics analysis module is used to compare preprocessed sequencing data with a pathogen reference database to identify the types and relative abundance of pathogens in clinical samples. The preprocessed sequencing data were compared with drug resistance gene databases and virulence factor databases to identify drug resistance genes and virulence factors carried by pathogens in clinical samples. Using at least one of the pathogen type, relative abundance, drug resistance genes, and patient clinical information as input features, a trained machine learning model is used to predict the drug resistance phenotype of the pathogen to a specific antibiotic. The clinical decision support and report generation module is used to integrate the pathogen types and relative abundance, drug resistance gene spectrum, virulence factors, and predicted drug resistance phenotypes to generate and output a structured clinical report containing treatment recommendations. The database and management module is used to store the pathogen reference database, the drug resistance gene database, the virulence factor database, patient clinical data, and parameters of the machine learning model.
[0012] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for pathogen identification and drug resistance prediction based on metagenomic sequencing.
[0013] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for pathogen identification and drug resistance prediction based on metagenomic sequencing.
[0014] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects: This invention achieves a closed loop from broad-spectrum pathogen screening to accurate drug resistance prediction and clinical decision-making through the deep integration of metagenomic sequencing and machine learning. Targeted quality control and host nucleic acid removal effectively reduce host background interference, improve the signal-to-noise ratio and effective data utilization of sequencing data, thereby increasing the detection sensitivity of low-abundance pathogens. By integrating multi-dimensional features such as pathogen type and relative abundance, drug resistance genes, and even patient clinical information into the machine learning model, the differences between drug resistance genotypes and phenotypes are effectively bridged, solving the problem of false positives or false negatives easily generated by relying solely on gene comparison, and significantly improving the accuracy and clinical reliability of drug resistance phenotype prediction. By transforming complex metagenomic and bioinformatics analysis results into structured clinical reports containing pathogen drug resistance profiles and treatment recommendations, the barriers between underlying detection data and actual clinical applications are broken down, significantly shortening the anti-infection diagnosis cycle and providing scientific and intuitive decision support for early targeted drug use and precision anti-infection treatment for critical and difficult-to-treat clinical cases. Attached Figure Description
[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.
[0016] The pathogen identification and drug resistance prediction method and system based on metagenomic sequencing of the present invention will be further described below with reference to the accompanying drawings; Figure 1 This is the overall flowchart of the pathogen identification and drug resistance prediction method based on metagenomic sequencing in Embodiment 1 of the present invention. Detailed Implementation
[0017] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.
[0018] To better understand the purpose, structure, and function of this invention, the invention will be described in further detail below with reference to the accompanying drawings.
[0019] Example 1 like Figure 1 As shown, this invention provides a method for pathogen identification and drug resistance prediction based on metagenomic sequencing, comprising the following steps: S1. Obtain clinical samples, perform metagenomic sequencing to obtain raw sequencing data, and preprocess the raw sequencing data; wherein the preprocessing includes: quality control and host nucleic acid removal; In S1, the clinical sample includes at least one of the following: blood, cerebrospinal fluid, bronchoalveolar lavage fluid, tissue, urine, or feces.
[0020] This embodiment, through unbiased metagenomic sequencing, breaks through the limitations of the narrow detection range of traditional culture or targeted PCR, and can comprehensively capture the nucleic acid information of all potential pathogens in the sample. At the same time, through quality control and host nucleic acid removal processing, low-quality data and human background noise interference are effectively eliminated, which greatly improves the accuracy and computational efficiency of subsequent microbial comparison analysis, laying a data foundation for achieving rapid diagnosis within 24-48 hours.
[0021] S2. The preprocessed sequencing data is compared with the pathogen reference database to identify the types and relative abundance of pathogens in clinical samples. This embodiment is based on a large reference database of pathogenic microorganisms for comparison and identification, which can quickly and accurately identify the types of pathogenic microorganisms causing infection and their load (relative abundance), solving the problems of low positive rate and long time consumption of traditional methods. It is especially suitable for the diagnosis of difficult infections with complex etiologies such as fever of unknown origin and sepsis.
[0022] S3. The preprocessed sequencing data is compared with the drug resistance gene database and the virulence factor database to identify the drug resistance genes and virulence factors carried by pathogens in clinical samples. Based on the identification of pathogens, this embodiment further explores the drug resistance genes and virulence factors they carry, revealing the drug resistance potential and pathogenicity of pathogens at the genomic level. This makes up for the deficiency of traditional mNGS, which can only provide species information but cannot directly give the molecular mechanism of drug resistance, and provides key genotypic feature inputs for subsequent drug resistance phenotype prediction.
[0023] S4. Using at least one of the pathogen type, relative abundance, drug resistance gene and patient clinical information as input features, a trained machine learning model is used to predict the drug resistance phenotype of the pathogen to a specific antibiotic. In S4, the resistance phenotype of a pathogen to a specific antibiotic includes: the pathogen’s sensitivity classification to a specific antibiotic and the minimum inhibitory concentration trend.
[0024] In S4, the patient's clinical information includes at least one of the following: site of infection, underlying disease, and patient age.
[0025] In S4, the resistance phenotype of a pathogen to a specific antibiotic includes: the pathogen’s sensitivity classification to a specific antibiotic and the minimum inhibitory concentration trend.
[0026] In S4, the training process of the machine learning model is as follows: Collect genomic sequencing data of historical clinical isolates, corresponding in vitro drug susceptibility test results, and patient clinical information to construct a training dataset; Species information, presence or absence and copy number of drug resistance genes, virulence factors and core genome single nucleotide polymorphism features are extracted from the genome sequencing data, and input feature vectors are generated by combining them with patient clinical information. Using the results of the in vitro drug sensitivity test as labels, the machine learning model is trained using random forest, gradient boosting tree, or deep neural network algorithms.
[0027] The process of building and training the machine learning model in this embodiment is as follows: First, a training dataset was constructed by collecting genomic sequencing data of historical clinical isolates, corresponding gold standard in vitro antimicrobial susceptibility testing (AST) results, and relevant patient clinical information. Second, feature engineering was performed to extract species information, presence / deletion and copy number of resistance genes, virulence factors, and core genomic SNPs from the genomic data, combined with clinical features such as infection type and patient immune status. Subsequently, algorithms such as random forest, XGBoost, or deep neural networks were used to train a classification model with antimicrobial susceptibility test results (susceptible / resistant) as labels, while a regression model was constructed for MIC prediction. Finally, the model's accuracy, recall, and AUC were evaluated using an independent validation set. Once the performance met the standards, the model was integrated into the analysis engine for use in predicting antimicrobial resistance in new samples.
[0028] This embodiment overcomes the limitation of a single drug resistance gene determining phenotype. By integrating multi-dimensional features (pathogen characteristics, gene spectrum, clinical background, etc.) using a machine learning model, it achieves an intelligent leap from "genotype" to "phenotype". It can not only predict the susceptible / resistant classification, but also infer the minimum inhibitory concentration (MIC) trend, which greatly improves the accuracy and clinical reliability of drug resistance prediction.
[0029] S5. Integrate the pathogen types and relative abundance, drug resistance gene spectrum, virulence factors, and predicted drug resistance phenotypes to generate and output a structured clinical report containing treatment recommendations.
[0030] This embodiment transforms complex bioinformatics analysis results into structured, easy-to-read clinical reports and directly provides treatment recommendations including medication suggestions, effectively lowering the professional threshold for clinicians to interpret complex mNGS data; it realizes a closed loop from "sample sequencing" to "clinical decision-making", providing timely and actionable guidance for precision anti-infective therapy and reducing the overuse of empirical drugs.
[0031] Compared with the prior art, the present invention also has the following beneficial effects: (1) Comprehensive and rapid detection: Using mNGS technology to detect all potential pathogens without bias, combined with efficient bioinformatics processes, the traditional diagnosis cycle of several days to several weeks can be shortened to 24-48 hours.
[0032] (2) Accurate drug resistance prediction: It breaks through the limitation that traditional mNGS can only provide species information. By integrating drug resistance genes and machine learning models, it realizes intelligent prediction from genotype to phenotype, providing a direct basis for precision medicine.
[0033] (3) Strong clinical usability: It provides standardized fully automated analysis process and structured clinical reports, which lowers the threshold of bioinformatics analysis and makes the results easier for clinicians to understand and apply.
[0034] (4) Wide range of applications: The system design is universal and can be widely used in the etiological diagnosis of various infectious diseases such as fever of unknown origin, bloodstream infection, central nervous system infection, and respiratory tract infection.
[0035] Example 2 This invention also provides a pathogen identification and drug resistance prediction system based on metagenomic sequencing, used to execute the pathogen identification and drug resistance prediction method based on metagenomic sequencing in Example 1, including: The sample processing and sequencing module is used to acquire clinical samples, perform metagenomic sequencing to obtain raw sequencing data, and preprocess the raw sequencing data; wherein the preprocessing includes: quality control and host nucleic acid removal. In this embodiment, the sample processing and sequencing module receives and processes clinical samples, performing nucleic acid extraction, host nucleic acid removal, library construction, and high-throughput sequencing to generate raw sequencing data. This module is compatible with various sample types (blood, cerebrospinal fluid, bronchoalveolar lavage fluid, tissue, etc.) and mainstream sequencing platforms.
[0036] The intelligent bioinformatics analysis module is used to compare preprocessed sequencing data with a pathogen reference database to identify the types and relative abundance of pathogens in clinical samples. The preprocessed sequencing data were compared with drug resistance gene databases and virulence factor databases to identify drug resistance genes and virulence factors carried by pathogens in clinical samples. Using at least one of the pathogen type, relative abundance, drug resistance genes, and patient clinical information as input features, a trained machine learning model is used to predict the pathogen's sensitivity to a specific antibiotic and the trend of minimum inhibitory concentration.
[0037] In this embodiment, the intelligent bioinformatics analysis module is connected to the sample processing and sequencing module to receive the raw sequencing data and perform automated analysis. The engine further includes: (1) a data quality control and preprocessing unit: performing quality filtering, host removal, and redundancy removal on the raw data. (2) a pathogen rapid screening and identification unit: rapidly identifying the types of microorganisms and their relative abundance in the sample based on the K-mer algorithm, sequence alignment (such as BLAST, Kraken2), and species-specific databases. (3) a drug resistance gene and virulence factor analysis unit: identifying drug resistance genes and virulence factors carried by pathogens by comparing drug resistance gene databases (such as CARD, ResFinder) and virulence factor databases. (4) a drug resistance phenotype prediction unit: integrating machine learning models (such as random forest, gradient boosting tree, or deep learning network) to predict the sensitivity (sensitive, intermediate, resistant) and minimum inhibitory concentration (MIC) trend of the pathogen to a specific antibiotic based on multi-dimensional data such as the identified pathogen type, abundance, drug resistance gene spectrum, and patient clinical information (such as infection site, underlying disease).
[0038] The clinical decision support and report generation module is used to integrate the pathogen types and relative abundance, drug resistance gene spectrum, and predicted antibiotic sensitivity spectrum to generate and output a structured clinical report containing treatment recommendations. In this embodiment, the clinical decision support and report generation module communicates with the intelligent bioinformatics analysis module to integrate analysis results, generate structured clinical reports, and provide treatment recommendations. The report content includes: a list of detected pathogens (sorted by confidence level), drug resistance gene detection results, predicted antibiotic susceptibility profiles, recommended treatment regimens (including first-line and second-line drug selections and dosage references), and necessary clinical warnings.
[0039] The database and management module is used to store the pathogen reference database, the drug resistance gene database, the virulence factor database, patient clinical data, and parameters of the machine learning model.
[0040] In this embodiment, the database and management platform are used to store and manage pathogen reference genome databases, drug resistance gene databases, virulence factor databases, patient clinical data, historical test results, and model parameters, supporting continuous learning and optimization of the system.
[0041] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the pathogen identification and drug resistance prediction method based on metagenomic sequencing in Embodiment 1.
[0042] This system can be deployed on a local server or a cloud platform. The hardware includes: an automated nucleic acid extraction instrument for sample preprocessing (such as the MGISP-NE384), a high-throughput sequencer (such as the Illumina NovaSeq series), and a high-performance computing server (for bioinformatics analysis). The software includes: a sample information management system, a bioinformatics analysis pipeline (built based on the Nextflow or Snakemake framework), a machine learning model for drug resistance prediction, an automated clinical report generation system, and a database management system.
[0043] Example 3 This embodiment provides a specific example of applying the method in Embodiment 1 to the pathogen diagnosis of a patient with fever of unknown origin: Take a patient with fever of unknown origin (FUO) as an example: Clinically, peripheral blood samples were first collected, total nucleic acids were extracted using a matching kit, and human host nucleic acids were removed using a probe capture method. Subsequently, a DNA / cDNA library was constructed and paired-end 150bp sequencing was performed, generating approximately 20-50 million reads of raw data, which were then uploaded to an intelligent bioinformatics analysis engine.
[0044] The intelligent bioinformatics analysis engine performs five intelligent analysis steps sequentially: First, it uses FastQC and Trimmomatic for data quality control and pruning; then, it uses Bowtie2 to align reads to the human reference genome to remove host sequences; next, it uses Kraken2 / Bracken to align the data to an integrated database covering bacteria, viruses, fungi, and parasites, outputting pathogen species classification and relative abundance; simultaneously, it uses the ABRicate tool to align to the Comprehensive Antibiotic Resistance Database (CARD) to identify drug-resistant genes; finally, it inputs the identified pathogen species and abundance, the detected drug-resistant genes, and patient clinical metadata (such as age and infection site) into a pre-trained random forest model to predict the phenotypic sensitivity and confidence level of the pathogen to common antibiotics.
[0045] The clinical decision support module integrated the above results to generate a PDF report, which clearly stated: "High abundance of Klebsiella pneumoniae was detected, carrying the blaKPC carbapenemase gene. The model predicts resistance to meropenem (95% confidence) and sensitivity to ceftazidime / avibactam (88% confidence). Ceftazidime / avibactam is recommended as an empirical treatment option." The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A method for pathogen identification and drug resistance prediction based on metagenomic sequencing, characterized in that, Includes the following steps: S1. Obtain clinical samples, perform metagenomic sequencing to obtain raw sequencing data, and preprocess the raw sequencing data; The pretreatment process includes: quality control and host nucleic acid removal. S2. The preprocessed sequencing data is compared with a pathogen reference database to identify the types and relative abundance of pathogens in clinical samples. S3. The preprocessed sequencing data is compared with the drug resistance gene database and the virulence factor database to identify the drug resistance genes and virulence factors carried by pathogens in clinical samples. S4. Using at least one of the pathogen type, relative abundance, drug resistance gene and patient clinical information as input features, a trained machine learning model is used to predict the drug resistance phenotype of the pathogen to a specific antibiotic. S5. Integrate the pathogen types and relative abundance, drug resistance gene spectrum, virulence factors, and predicted drug resistance phenotypes to generate and output a structured clinical report containing treatment recommendations.
2. The method for pathogen identification and drug resistance prediction based on metagenomic sequencing according to claim 1, characterized in that, In S1, the clinical sample includes at least one of the following: blood, cerebrospinal fluid, bronchoalveolar lavage fluid, tissue, urine, or feces.
3. The method for pathogen identification and drug resistance prediction based on metagenomic sequencing according to claim 1, characterized in that, In S4, the patient's clinical information includes at least one of the following: site of infection, underlying disease, and patient age.
4. The method for pathogen identification and drug resistance prediction based on metagenomic sequencing according to claim 1, characterized in that, In S4, the resistance phenotype of a pathogen to a specific antibiotic includes: the pathogen’s sensitivity classification to a specific antibiotic and the minimum inhibitory concentration trend.
5. The method for pathogen identification and drug resistance prediction based on metagenomic sequencing according to claim 1, characterized in that, In S4, the training process of the machine learning model is as follows: Collect genomic sequencing data of historical clinical isolates, corresponding in vitro drug susceptibility test results, and patient clinical information to construct a training dataset; Species information, presence or absence and copy number of drug resistance genes, virulence factors and core genome single nucleotide polymorphism features are extracted from the genome sequencing data, and input feature vectors are generated by combining them with patient clinical information. Using the results of the in vitro drug sensitivity test as labels, the machine learning model is trained using random forest, gradient boosting tree, or deep neural network algorithms.
6. A pathogen identification and drug resistance prediction system based on metagenomic sequencing, used to execute the pathogen identification and drug resistance prediction method based on metagenomic sequencing as described in any one of claims 1-5, characterized in that, include: The sample processing and sequencing module is used to acquire clinical samples, perform metagenomic sequencing to obtain raw sequencing data, and preprocess the raw sequencing data. The pretreatment process includes: quality control and host nucleic acid removal. The intelligent bioinformatics analysis module is used to compare preprocessed sequencing data with a pathogen reference database to identify the types and relative abundance of pathogens in clinical samples. The preprocessed sequencing data were compared with drug resistance gene databases and virulence factor databases to identify drug resistance genes and virulence factors carried by pathogens in clinical samples. Using at least one of the pathogen type, relative abundance, drug resistance genes, and patient clinical information as input features, a trained machine learning model is used to predict the drug resistance phenotype of the pathogen to a specific antibiotic. The clinical decision support and report generation module is used to integrate the pathogen types and relative abundance, drug resistance gene spectrum, virulence factors, and predicted drug resistance phenotypes to generate and output a structured clinical report containing treatment recommendations. The database and management module is used to store the pathogen reference database, the drug resistance gene database, the virulence factor database, patient clinical data, and parameters of the machine learning model.
7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements the pathogen identification and drug resistance prediction method based on metagenomic sequencing as described in any one of claims 1-5.
8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the pathogen identification and drug resistance prediction method based on metagenomic sequencing as described in any one of claims 1-5.