Single-tube multiple qPCR mutation amplification detection technology based on AI atlas analysis and application of single-tube multiple qPCR mutation amplification detection technology

Abstract

The invention discloses a single-tube multiple qPCR mutation amplification detection technology based on AI atlas analysis, which comprises the following steps: S1, amplifying all DNA targets by using multiple qPCR, obtaining all target sequencing results by using specific reverse primer sequencing, and constructing a reference sequence library; s2, amplifying all the DNA targets by using multiple qPCR, and obtaining a qPCR product mixed atlas by using universal primer sequencing; and S3, analyzing the mixed atlas obtained in the step S2 by using an AI atlas analysis technology, comparing the mixed atlas with the sequences in the reference sequence library constructed in the step S1, and determining the amplified sequence containing mutation, thereby realizing multiple detection of DNA and low-frequency mutation. The method disclosed by the invention not only has the advantages of high sensitivity, high flux, simplicity and convenience in operation and low cost, but also can realize accurate and rapid detection of tumor mutation, and provides powerful support for early diagnosis and treatment of cancers.

Description

Technical Field

[0001] The present invention relates to the field of gene mutation detection, and more specifically to a single-tube multiplex qPCR mutation amplification detection technology based on AI graph analysis and its application. Background Art

[0002] Gene mutation is a key factor in the occurrence and development of diseases, so accurate and timely mutation detection is crucial for disease diagnosis, risk assessment, prognosis and medication guidance. At present, there are many methods for gene mutation detection, including hybridization chemistry (such as FISH and CISH), next-generation sequencing (NGS), quantitative polymerase chain reaction (qPCR), digital PCR, etc. Hybridization chemistry is mainly used to detect chromosome structural variations. The NGS sequencing method is relatively expensive and suitable for high-throughput, comprehensive genome sequencing; the multiple nucleic acid inhibition probe displacement amplification technology based on NGS can achieve 80-fold high throughput and a detection limit of 0.019% at a sequencing depth of 250× (Nat Biomed Eng, 2021, 5, 690–701.).

[0003] qPCR is commonly used for qualitative and quantitative detection of genes, but has limited sensitivity and throughput in detecting DNA mutations; specific allele PCR can improve the sensitivity of qPCR, but is only suitable for detecting specific variants and distinguishing between two known nucleotide variants. Multiplex PCR technology has shown broad application prospects in many fields due to its unique advantages. It uses a single reagent system, which is both economical and efficient, and can achieve simultaneous amplification detection of multiple genes, significantly reducing the amount of reagents and samples required for detection. This has led to the widespread application of multiplex PCR in many fields such as genetic disease screening, pathogen detection, drug resistance gene analysis, and early cancer diagnosis (CN116219071B, CN116516075B, CN115896312A, CN115896312A, CN112592964A, CN100372945C, CN1936019B; CN108823287B, CN114622000A, CN112592964A, CN115109840A; Nat Biomed Eng. 2017, 1, 714–723 2017.).

[0004] However, despite the advantages of multiplex PCR, due to the limitation of the number of fluorescence channels of the instrument, the multiplex PCR technology based on fluorescent probes still needs to adopt a separate detection method in practical applications, which limits its throughput and efficiency to a certain extent. The existing technology also significantly improves the number of targets detected by single-tube multiplex PCR by combining multi-color combination, probe coding and multi-color melting curve analysis, successfully breaking through the 10-fold limit (PLos One, 2011, 6, e16033, EP2436777B1, US 9334533 B2).

[0005] For example, a multiplex nucleic acid inhibition probe displacement amplification technology based on qPCR and Taqman probes designs an inhibition probe for the wild-type template, which is used to completely hybridize with the wild-type template to inhibit wild-type amplification during PCR, while the mutant template does not completely match the inhibition probe, thereby achieving mutation amplification and enrichment during PCR. This method uses nucleic acid inhibition probe displacement amplification technology to enrich low-frequency DNA mutations, and uses TaqMan probes to qualitatively and quantitatively analyze mutations. By mixing and detecting the primers and probes required for the target, this method can detect 3 targets in a single tube reaction, and 3 channels can achieve 9-plex detection (Nat Biomed Eng, 2017, 1, 714–723 2017.). Although this method uses TaqMan-based qPCR detection technology, combined with multiple nucleic acid inhibition probe displacement amplification technology, it can amplify the detection of 0.1% mutations, but multiple detection requires separate management.

[0006] A "MeltArray" multiplex qPCR technology uses a multiplex amplification probe melting curve method combined with an intermediary probe. This technology uses 12 different intermediary primers to complement different predefined regions on the molecular beacon reporter gene ring, and uses molecular beacons and multiple intermediary primers to form a series of fluorescent double strands with different melting points, thereby detecting 12 targets in a single fluorescent channel. This technology improves the detection throughput of qPCR, and each reaction can detect 30 target molecules (PNAS, 2022, 119, e2110672119, see Chinese patent CN108823287B), but this technology can only detect a minimum of 5% mutations, and the sensitivity needs to be optimized.

[0007] A METEOR-dPCR technology uses a target with a half-marker and a right half-marker and adds a barcode sequence for identifying the target, and uses a fluorescent loose molecular beacon probe to combine with the barcode sequence to form a DNA chain with different melting temperatures to achieve multiplexed quantitative gene combination analysis. This method can quantify 60 targets at the same time, and it is expected to reach 216 multiplexes using 3 fluorescent channels (AdvSci, 2023, 10, 2301630.). This technology can detect dozens to hundreds of multiplexes like the existing MeltArray technology, but the single-multiplex detection flux only detects 1 mutation.

[0008] The current multiplex PCR detection technology is mainly based on fluorescent dyes and fluorescent probes. Fluorescent dye-based technology uses fluorescent color and melting curve characteristics to identify different amplified fragment targets through different melting temperatures, thereby achieving multiple detection. However, due to the limited fluorescence channels of the instrument, the number of targets that can be detected in a single tube of the multiplex PCR technology based on fluorescent probes is limited, and increasing the number of targets requires cumbersome "division design", which limits the throughput and increases cost and time. The throughput of multiplex qPCR can be improved by combining different fluorescent channels, fluorescent types and concentrations, melting temperatures, etc. Although these technologies have advantages in multi-target detection, their sensitivity needs to be improved. At the same time, the design and optimization process is cumbersome and costly, which limits their promotion in practical applications.

[0009] It can be seen that the existing multiplex PCR detection technology faces multiple challenges, including high design complexity, high cost, limited sensitivity, and difficulty in improving the throughput of a single reaction. These problems limit the widespread application of multiplex PCR technology in laboratory research and clinical practice. Although the current single-tube multiplex PCR detection technology can achieve single-tube high-throughput gene mutation detection, there are still certain limitations in detection sensitivity, which needs to be further improved. On the other hand, those PCR detection technologies with high sensitivity often find it difficult to achieve the ideal throughput level in single-tube detection. Therefore, it is necessary to develop a simpler, low-cost, and highly sensitive single-tube multiplex PCR technology to meet the needs of higher throughput and low-cost detection methods in fields such as gene mutation detection. Summary of the invention

[0010] The purpose of the present invention is to provide a single-tube multiplex qPCR mutation amplification detection technology based on AI graph analysis and its application, so as to solve the problem that the prior art lacks a simpler, low-cost, and highly sensitive single-tube multiplex PCR technology to meet the demand for higher throughput and low-cost detection methods in the field of gene mutation detection.

[0011] In order to solve the above problems, the present invention adopts the following technical solutions:

[0012] According to a first aspect of the present invention, a single-tube multiplex qPCR mutation amplification detection technology based on AI map analysis is provided, comprising the following steps: S1: using multiplex qPCR to amplify all DNA targets, using specific reverse primer sequencing to obtain sequencing results of all targets, and constructing a reference sequence library; S2: using multiplex qPCR to amplify all DNA targets, using universal primer sequencing to obtain a mixed map of qPCR products; S3: using AI map analysis technology to analyze the mixed map obtained in step S2, comparing the mixed map with the sequence in the reference sequence library constructed in step S1, and determining the sequence containing the mutation to be amplified, thereby realizing multiplex detection of DNA and low-frequency mutations.

[0013] Preferably, steps S1 and S2 include: designing multiple nucleic acid inhibition probes for the wild-type DNA template, suppressing the wild-type DNA template and amplifying the mutant DNA template normally through nucleic acid inhibition probe displacement amplification technology based on qPCR, obtaining a single sequence of the mutant by sequencing with a specific reverse primer, obtaining a mixed sequence of the mutant by sequencing with a universal primer, and achieving initial screening of mutations according to changes in the Ct value of qPCR amplification.

[0014] Preferably, in step S3, the AI ​​graph analysis technology uses a dynamic time warping algorithm and a point-to-point comparison algorithm to calculate similarity.

[0015] Preferably, the dynamic time warping algorithm includes: a DTW single spectrum random combination comparison algorithm and a DTW single spectrum step-by-step comparison and superposition algorithm.

[0016] According to a preferred solution of the present invention, the analytical process of the DTW single spectrum random combination comparison algorithm is as follows:

[0017] 1) Obtain all DNA amplification sequences and length information;

[0018] 2) According to the four channels of A, G, C, and T, 1100 coordinate data of the specific primer sequencing single spectrum and the UP sequencing mixed spectrum are extracted from the 5' to 3' direction starting from the FP end; when reading the mixed spectrum, the longest sequence found is used as a reference to extract 1100 coordinate data, and other sequences are filled with 0 to the length of the longest sequence in the most mixed spectrum;

[0019] 3) Divide by the maximum value of the ordinate of each sequence and multiply by 100 to standardize each sequence to reduce the difference between the mixed spectrum ordinate data and the single spectrum ordinate data;

[0020] 4) Arbitrarily combine the single spectrum sequence coordinate data, the formula is: Where n is the number of single spectrum sequences blended in, and k is the number of any combinations;

[0021] 5) The single spectrum coordinate data of any combination are added together, and the combined sequence coordinate data need to be divided by the maximum value of the combined single spectrum coordinate data and multiplied by 100 to standardize the combined sequence coordinate data;

[0022] 6) According to the four channels A, G, C, and T, the DTW algorithm is used to compare the combined sequence and the mixed spectrum ordinate data. The formula is as follows: Dist(x,y)=|xy| D(i,j)=Dist(i,j)+min[D(i-1,j),D(i,j-1),D(i-1,j-1)] Where x,i is the ordinate data of the mixed spectrum, and y,j is the ordinate data of the combined sequence. If D(i,j) is smaller, it means that the similarity between the combined sequence and the mixed spectrum is higher;

[0023] 7) The DTW distances of the combined sequence and the mixed spectrum sequence of the four channels are added and divided by 4, and the (average) shortest DTW distance between the combined sequence and the mixed spectrum sequence is calculated to obtain the most similar single spectrum combined sequence, and the sequence contained in the mixed spectrum is predicted based on the combined single spectrum sequence with the (average) shortest DTW distance.

[0024] According to a preferred solution of the present invention, the analytical process of the DTW single spectrum step-by-step comparison and superposition algorithm is as follows:

[0025] 1) Obtain all DNA amplification sequences and length information;

[0026] 2) According to the four channels of A, G, C, and T, 1100 coordinate data of the specific primer sequencing single spectrum and the UP sequencing mixed spectrum are extracted from the 5' to 3' direction starting from the FP end; when reading the mixed spectrum, the longest sequence found is used as a reference to extract 1100 coordinate data, and other sequences are filled with 0 to the length of the longest sequence in the most mixed spectrum;

[0027] 3) Divide by the maximum value of the ordinate of each sequence and multiply by 100 to standardize each sequence to reduce the difference between the mixed spectrum ordinate data and the single spectrum ordinate data;

[0028] 4) identifying the longest sequence in the mixed spectrum by naked eyes, comparing the abscissa data of the peaks in the mixed spectrum and the longest sequence, calculating the stretching ratio of the mixed spectrum, and multiplying the abscissa of the mixed spectrum by the stretching ratio to reduce the offset of the abscissa data;

[0029] 5) Using the sequence position information, determine whether the single spectrum sequence exists in the mixed spectrum starting from the longest sequence until the shortest single spectrum sequence.

[0030] 6) According to the four channels A, G, C, and T, the DTW algorithm is used to gradually compare the single spectrum sequence and the mixed spectrum ordinate data at the corresponding position. The position of the DTW comparison is the length of the single spectrum sequence to the next single spectrum sequence. A minimum of 100 data points must be intercepted. The minimum weight of each channel of the single spectrum is found by calculating the minimum DTW distance of each channel. The weight ranges from 0 to 2 and the interval is 0.1. If the minimum DTW distance of three channels calculated by the mixed spectrum comparison is not the single spectrum at this position or two channels of the single spectrum at this position are ≤0.1, it is determined that the single spectrum does not exist in the mixed spectrum and the weight is 0. If it is determined to exist, the sequence is added to the combined sequence according to each channel multiplied by the weight. The formula is as follows: Dist(x,y)=|xy| D(i,j)=Dist(i,j)+min[D(i-1,j),D(i,j-1),D(i-1,j-1)] ifD(i,j)=min[D(i,j)],n=best_n Form=1∶k Combined_Sequence[m]=Combined_Sequence[m-1]+n*sequence[m]

[0031] Where x,i is the vertical coordinate data of the mixed spectrum, and y,j is the vertical coordinate data of the combined sequence. If D(i,j) is smaller, it means that the similarity between the combined sequence and the mixed spectrum is higher, n is the optimal weight to make the DTW distance the shortest, k is the total number of sequences, and m is the sequence number (1-k);

[0032] 7) Predicting the mutation sequence in the mixed spectrum based on the single spectrum determined to be mixed in the mixed spectrum.

[0033] According to a preferred solution of the present invention, the parsing process of the point-to-point comparison algorithm is as follows:

[0034] 1) Obtain all DNA amplification sequences and length information;

[0035] 2) According to the four channels of A, G, C, and T, 1100 coordinate data of the specific primer sequencing single spectrum and the UP sequencing mixed spectrum are extracted from the 5' to 3' direction starting from the FP end; when reading the mixed spectrum, the longest sequence found is used as a reference to extract 1100 coordinate data, and other sequences are filled with 0 to the length of the longest sequence in the most mixed spectrum;

[0036] 3) Divide by the maximum value of the ordinate of each sequence and multiply by 100 to standardize each sequence to reduce the difference between the mixed spectrum ordinate data and the single spectrum ordinate data;

[0037] 4) Based on the four channels of A, G, C, and T, the similarity of each FP sequence data in the UP sequence data is calculated using a point-to-point similarity comparison algorithm. The closer to 1, the more similar. The algorithm formula is: Where f(x) is the point-to-point similarity between the single spectrum sequence and the mixed spectrum sequence, i is the four channels A, G, C, and T, and j is the vertical coordinate data of each channel;

[0038] 5) According to the 5-fold data, if f(x) is less than 6, it means that the single spectrum sequence has a mutation in the mixed spectrum sequence. By calculating all the single spectrum sequences less than 6, the sequence contained in the mixed spectrum is predicted.

[0039] According to a preferred solution of the present invention, in step S3, the AI ​​graph analysis technology further includes constructing a mixed spectrum analysis model so that the model can autonomously learn and effectively identify mutation sequences in the mixed spectrum, wherein the machine learning analysis process is as follows:

[0040] 1) Data collection: Collect all sample data and preprocess them, including feature extraction, selection and standardization;

[0041] 2) Data partitioning: The data set is divided into a training set and a test set in a ratio of 7:3. The training set contains all single spectra, a large number of combined spectra, and a small number of sequenced mixed spectra, while the test set contains a small number of combined spectra and a large number of sequenced mixed spectra.

[0042] 3) Training model: Use single spectra, a large number of combined spectra, and a small number of sequencing mixed spectra from the reference sequence library as training sets to establish a mathematical (mapping) relationship between input X (spectrum) and output Y (existing single spectrum sequence) variables. Such X, Y pairs constitute the label data used to build the model, so as to learn how to predict the output from the input, and then perform parameter tuning and feature selection to train a complete model;

[0043] 4) Model evaluation and optimization: Continuously evaluate and optimize the model, including data set segmentation, hyperparameter tuning, and model replacement, and use performance metrics for quantitative evaluation.

[0044] 5) Test model: The remaining combined spectra and sequencing mixed spectra are used as test sets and input into the final test model to make predictions and evaluate the performance indicators of the model.

[0045] The use of long-fragment nucleic acid inhibition probes allows the enrichment interval to be as long as 44 nt, enabling high-throughput amplification detection of mutations.

[0046] According to a preferred embodiment of the present invention, the qPCR amplification conditions include:

[0047] The qPCR cycle steps can be performed in the denaturation-annealing-extension step, or in the denaturation-annealing + extension mode;

[0048] The annealing temperature range of qPCR is between 50°C and 72°C, and the annealing temperature can also be changed as the number of cycles increases;

[0049] The annealing extension time used in qPCR includes but is not limited to 30s to 5min;

[0050] The number of cycles for qPCR amplification is usually, but not limited to, 30 to 60 cycles;

[0051] DNA polymerases that can be used in qPCR include high-fidelity polymerases and low-fidelity polymerases. High-fidelity polymerases include but are not limited to Q5, Phusion, Kapa, ​​etc., and low-fidelity polymerases include but are not limited to Taq, PowerUp, itaq, HieffUNICON, etc.;

[0052] The primer concentration range for qPCR was between 1 nM and 10 μM;

[0053] The concentration range of nucleic acid inhibition probes for qPCR was between 1 nM and 100 μM;

[0054] qPCR concentrations of various substances, including but not limited to Mg 2+ The concentration is between 1mM and 100mM, the dNTP concentration is between 20μM and 20mM, the DMSO concentration is between 1% and 30%, and the Na + The concentration is between 100mM and 10M.

[0055] According to a preferred embodiment of the present invention, PCR primer design can use but is not limited to NCBI primerdesigning tool, Primer Premier 5.

[0056] According to a preferred embodiment of the present invention, the detection performance can be continuously improved by using optimized key parameters, including the preferred suitable DNA polymerase, premix system and annealing extension time, which helps to increase the enrichment capacity of mutant templates in each round of qPCR amplification and increase the difference between wild-type and mutant signals.

[0057] According to the qPCR-based nucleic acid inhibition probe displacement amplification mutation amplification technology used in the present invention, the core idea of ​​the technology is to design a series of nucleic acid inhibition probes (B) based on the known wild-type DNA sequence, and the inhibition probe can be perfectly hybridized with the wild-type template, thereby inhibiting its amplification during the PCR process. The mutant template is unstable due to the presence of the mutant base, so the inhibition probe can be replaced by the forward primer, so that the mutant template amplification efficiency is greater than that of the wild-type template. After multiple rounds of PCR amplification, the purpose of inhibiting the wild type and amplifying the mutant type is finally achieved. In order to achieve the inhibition of the wild-type template and the enrichment of the mutant template by the nucleic acid inhibition probe displacement amplification technology, the ΔΔG°rxn of the wild-type template reaction will be designed to be a positive value (about +2kcal / mol), and the ΔΔG°rxn of the mutant template reaction will be designed to be a negative value (about -2kcal / mol). The designed nucleic acid inhibition probe 3' end needs to add a blocking group to inhibit DNA extension. The blocking group includes but is not limited to phosphorylation modification, spacer, peptide nucleic acid (PNA), locked nucleic acid (LNA) and DXXDM, where D is a DNA residue, X is a C3 spacer, and M is a mismatch ( Figure 2 ). The length of the nucleic acid inhibition probe is usually 19 to 36 nt, the GC content is between 40% and 60%, and the maximum homopolymer length does not exceed 4; the binding free energy change ΔG of the primer and template ranges from -14kcal / mol to -17kcal / mol. The nucleic acid inhibition probe has a certain overlap range with the FP, and the length is generally 7 to 12 nt.

[0058] According to a preferred embodiment of the present invention, the universal primer cannot hybridize with the primers and inhibitor probes in the panel to form dimers and does not affect the amplification of the target sequence, including but not limited to the sequences used in the present invention.

[0059] According to a preferred embodiment of the present invention, the multiple primers and inhibition probes designed in the present invention are analyzed using the nucleic acid sequence analysis platform NUPACK (but not limited to) to predict the secondary structure, thermodynamic parameters, hybridization and pairing of nucleic acids. The homopolymer generally does not exceed 4, and the maximum number of complementary bases at the 3' end is between 4 and 10.

[0060] According to a preferred embodiment of the present invention, the primer design of the present invention is carried out under conventional PCR conditions (including but not limited to Na + The ion concentration is 0.18M, ​​the temperature is 60°C), and the binding free energy ΔG is between -4kcal / mol and -20kcal / mol.

[0061] The blocking group at the 3' end of the nucleic acid inhibition probe designed in the present invention includes but is not limited to phosphorylation modification, C3 spacer, PNA, LNA, and the blocking group includes but is not limited to phosphorylation modification, spacer, PNA, LNA and DXXDM.

[0062] The primer design method of the present invention uses but is not limited to Primer BLAST to verify the amplification specificity of the primers.

[0063] According to a preferred embodiment of the present invention, the sequencing technology includes but is not limited to Sanger sequencing technology.

[0064] According to the second aspect of the present invention, there is provided a single-tube multiplex qPCR mutation amplification detection technology based on AI graph analysis as described above for use in tumor mutation detection, laboratory cell line contamination identification, genotyping and quantification, genetic testing, and companion diagnosis.

[0065] It is well known that in the field of scientific research, laboratory cell line contamination has become a problem that cannot be ignored. Contaminated cell lines may cause distortion of experimental data, seriously affecting the accuracy and credibility of research conclusions. Therefore, accurate and rapid identification of laboratory cell line contamination is crucial to ensure the authenticity and reliability of scientific research data. The present invention can be used for laboratory cell line contamination identification by optimizing multiplex qPCR technology and combining advanced sequencing and AI spectrum analysis methods. In the field of clinical medicine, tumor mutation detection is of great significance for early screening of cancer, formulation of treatment plans, and improvement of patient survival rates.

[0066] The key invention of the present invention is that, using specific primers (RP), universal primers (UP) and multiple nucleic acid inhibition probe displacement amplification technology, multiple DNA targets are enriched by qPCR, and with AI spectrum analysis technology, multiple mutation sequences in multiple products can be accurately determined by only one sequencing. Multiple nucleic acid inhibition probe displacement amplification technology is to design nucleic acid inhibition probes based on the wild-type template sequence, thereby inhibiting the amplification of the wild-type template, and the mutant template is effectively enriched during each round of qPCR amplification due to mismatches with the inhibition probe. The present invention applies AI spectrum analysis technology to single-tube multiple qPCR mutation amplification detection for the first time, and combines qPCR-based nucleic acid inhibition probe displacement amplification mutation amplification technology, sequencing, AI spectrum analysis and machine learning to achieve single-tube multiple qPCR mutation amplification detection. The multiple nucleic acid inhibition probe displacement amplification technology and AI analysis map technology in the present invention complement each other. The multiple nucleic acid inhibition probe displacement amplification based on qPCR makes the amplification product only a mutant type, thereby reducing the complexity of mixed map analysis. AI analysis map can determine the mutant sequence; qPCR technology is simple and fast to operate, combined with multiple nucleic acid inhibition probe design, it can detect hundreds of potential mutation sites in a single reaction. By monitoring the changes in Ct values, it is possible to quickly and preliminarily determine whether there are related mutations, greatly improving the screening efficiency.

[0067] The single-tube multiplex qPCR mutation amplification detection technology based on AI spectrum analysis provided by the present invention has the following main features: 1) It adopts multiple nucleic acid inhibition probe displacement amplification technology based on qPCR to effectively amplify the mutation signal, and at the same time, performs a rough mutation screening on the sample according to the Ct value of qPCR; 2) It combines universal primer sequencing and mixed spectrum analysis technology to determine the mutation sequence; 3) Multiple nucleic acid inhibition probe displacement amplification technology and AI analysis spectrum technology work together. Since the products after inhibition probe displacement amplification are mainly mutant types, this greatly reduces the complexity of mixed spectrum analysis, and AI analysis spectrum can accurately determine the mutation sequence.

[0068] The method provided by the present invention has the following beneficial effects compared with the prior art:

[0069] High sensitivity and low detection limit: By combining specific primers and universal primers with inhibition probes, the amplification of wild-type templates can be effectively inhibited during qPCR amplification, thereby selectively amplifying mutant templates. The single-plex nucleic acid inhibition probe displacement amplification technology based on qPCR can detect low-frequency mutations of 0.01%, and multiplex can be used for high-throughput detection of DNA mutations. For example, the 16-plex nucleic acid inhibition probe displacement amplification technology based on single-tube qPCR can cover the detection of 342 mutation sites with a minimum detection limit of 0.2%. The 12-plex nucleic acid inhibition probe displacement amplification technology based on single-tube qPCR can cover the detection of 201 gene mutation sites in colorectal cancer.

[0070] Accurately identify mutation sequences: By sequencing the amplification products of multiple nucleic acid inhibition probes based on qPCR and comparing them with the DNA sequence library, the specific sequences contained in the sample can be accurately analyzed. This combination of qPCR and sequencing can not only quickly screen mutations, but also determine the sequence of mutations, greatly reducing the complexity of mixed map analysis.

[0071] Wide range of applications: The present invention can be used to detect laboratory cell line contamination, circulating free DNA (cfDNA) mutations in tumor tissue and blood, etc., covering multiple fields of scientific research and clinical diagnosis, and has broad application prospects.

[0072] In summary, the present invention provides a highly sensitive, high-throughput, easy-to-operate, and cost-effective low-frequency mutation detection solution, which is suitable for various application scenarios such as laboratory cell line contamination identification, blood cfDNA, and tumor tissue mutation detection. It can achieve accurate and rapid detection of tumor gene mutations and provide strong support for the early diagnosis and treatment of cancer. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 :Workflow for mutation analysis by multiplex qPCR amplification and sequencing;

[0074] Figure 2 : The 3' end of the nucleic acid inhibition probe can be modified by a group;

[0075] Figure 3A :DTW single spectrum random combination comparison and spectrum analysis flow chart;

[0076] Figure 3B :DTW single spectrum step-by-step comparison and superposition spectrum analysis flow chart;

[0077] Figure 4 :Point-to-point comparison algorithm spectrum analysis flow chart;

[0078] Figure 5 :Flowchart of spectrum interpretation based on machine learning;

[0079] Fig. 6A :Based on 16-plex nucleic acid inhibition probe amplification detection of low-frequency mutations and qPCR results;

[0080] Figure 6B : Detection of cell line contamination using multiplex nucleic acid inhibition probe displacement amplification technology based on qPCR;

[0081] Fig. 7A :Amplification curve of 16SNP detected by single-plex nucleic acid inhibition probe displacement amplification technology based on qPCR;

[0082] Figure 7B :Summary of detecting point mutations of 16SNPs with different mutation frequencies using single-plex nucleic acid inhibition probe displacement amplification technology based on qPCR;

[0083] Fig. 8A :Summary of 0%, 1%, 100% VAF Ct values ​​detected by 8-plex nucleic acid inhibition probe displacement amplification technology based on qPCR;

[0084] Figure 8B :Summary of the inhibitory effects of different DNA polymerases, premix systems and annealing extension times on the Ct value of mutant DNA templates before and after adding multiple nucleic acid inhibition probes to the 8-plex qPCR system;

[0085] Figure 8C : qPCR-based 8-plex nucleic acid inhibition probe displacement amplification technology to amplify the standard curve of different mutation numbers and VAF templates;

[0086] Fig.9A :Summary of Ct values ​​for identifying contamination in different cell lines using 5-plex and 16-plex nucleic acid inhibition probe displacement amplification techniques based on qPCR;

[0087] Fig. 9B : qPCR-based 5-plex nucleic acid inhibition probe displacement amplification amplifies the wild-type template containing 1% contamination sequencing results;

[0088] Fig. 9C : qPCR-based 5-plex nucleic acid inhibition probe displacement amplification technology was used to detect NA18537 contaminated with HeLa, and the mixed spectrum results were analyzed using a point-to-point alignment algorithm after universal primer sequencing;

[0089] Fig.9D :Point-to-point comparison algorithm analyzes the accuracy of UP sequencing mixed spectrum analysis;

[0090] Fig. 10A :Amplification curve of 16-plex nucleic acid inhibition probe based on qPCR;

[0091] Fig. 10B : Sequencing results of wild-type template amplification and 0.5% mutant template amplification by 16-plex nucleic acid inhibition probe displacement amplification;

[0092] Fig. 10C :16-plex nucleic acid inhibition probe displacement amplification based on qPCR and sequencing results of SNPs in different cell lines;

[0093] Fig.11 :Detection of mutations in colorectal cancer patients using multiplex nucleic acid inhibition probe displacement amplification technology based on qPCR;

[0094] Fig.12 :Using 12-plex nucleic acid inhibition probe replacement technology based on qPCR to detect clinical tissue samples of colorectal cancer patients and interpret the results;

[0095] Fig.13 : Schematic diagram of mutation amplification of normal nucleic acid inhibition probe and long-fragment nucleic acid inhibition probe;

[0096] Fig.14 :The enrichment effect of long-fragment nucleic acid suppression probe on different mutation positions and types within the mutation enrichment interval;

[0097] Fig.15A :KRAS (G13S) mutation standard curve based on qPCR long-fragment nucleic acid inhibition probe displacement amplification mutation amplification technology;

[0098] Fig. 15B :Standard curve of NRAS (G12R) mutation detection based on qPCR long-fragment nucleic acid inhibition probe displacement amplification mutation amplification technology.

[0099] Fig.16A :Detection of KRAS mutation amplification curve and sequencing results in colorectal cancer patients using qPCR-based long-fragment nucleic acid inhibition probe mutation amplification technology;

[0100] Fig. 16B :Amplification curves and sequencing results of NRAS mutations in patients with colorectal cancer detected by long-fragment nucleic acid inhibition probe mutation amplification technology based on qPCR. DETAILED DESCRIPTION

[0101] The present invention will be further described below in conjunction with specific examples. It should be understood that the following examples are only used to illustrate the present invention and are not intended to limit the scope of the present invention. Unless otherwise specified, the technical means used in the examples are conventional operations in the art, or according to the experimental methods recommended by the kit and instrument manufacturers. The reagents and materials used in the examples can be obtained from commercial sources unless otherwise specified.

[0102] The present invention provides a method for detecting multiple DNA mutations, which uses a specific reverse primer (RP) and a universal primer (UP) to amplify the target by qPCR, and then uses RP and UP sequencing, and combines an AI method to analyze the UP sequencing spectrum to obtain multiple mutations ( Figure 1 ). After conventional multiplex qPCR amplification, the qPCR amplification products are sequenced using specific RP to obtain a single sequence sequencing spectrum and build a reference sequence library. After multiple nucleic acid inhibition probe displacement amplification based on qPCR, the amplification products are sequenced using UP to obtain a mixed spectrum of all mutation sequences. Using AI spectrum analysis technology to analyze the sequencing mixed spectrum, multiple mutation sequences can be resolved.

[0103] The technology described in the present invention is based on two methods: ordinary qPCR single-plex amplification and qPCR-based single-plex nucleic acid inhibition probe displacement amplification, as follows:

[0104] Conventional qPCR single-plex amplification technology

[0105] First, design appropriate primers according to the target fragment to be amplified. PCR primer design can use but is not limited to NCBI primer designing tool, Primer Premier 5, the primer length is usually 15-30nt, the GC content is between 40% and 60%, and the maximum homopolymer length does not exceed 4; the binding free energy change ΔG of the primer and template is in the range of -10.5kcal / mol to -12.5kcal / mol.

[0106] The system used for single-plex qPCR amplification is shown in Table 1, including Hieff UNICON premix, forward primer (FP), reverse primer with universal sequence (RP-UP), and DNA template. The program used for qPCR amplification is: 95°C: 2min—[(95°C: 15s—60°C: 60s)×50].

[0107] qPCR-based single-plex nucleic acid inhibition probe displacement amplification mutation amplification technology

[0108] The core idea of ​​this technology is to design a series of nucleic acid inhibition probes (B) based on the known wild-type DNA sequence. The inhibition probe can perfectly hybridize with the wild-type template, thereby inhibiting its amplification during the PCR process. The mutant template is unstable due to the presence of the mutant base, which causes mismatches with the inhibition probe. Therefore, the inhibition probe can be replaced by the forward primer, making the mutant template amplification efficiency greater than that of the wild-type template. After multiple rounds of PCR amplification, the purpose of inhibiting the wild type and amplifying the mutant type is finally achieved. In order to achieve the inhibition of the wild-type template and the enrichment of the mutant template by the nucleic acid inhibition probe displacement amplification technology, the ΔΔG°rxn of the wild-type template reaction will be designed to be a positive value (about +2kcal / mol), while the ΔΔG°rxn of the mutant template reaction will be designed to be a negative value (about -2kcal / mol). A blocking group needs to be added to the 3' end of the designed nucleic acid inhibition probe to inhibit DNA extension. The blocking group includes but is not limited to phosphorylation modification, spacer, peptide nucleic acid (PNA), locked nucleic acid (LNA) and DXXDM, where D is a DNA residue, X is a C3 spacer, and M is a mismatch ( Figure 2 ). The length of nucleic acid inhibition probe is usually 19 to 36 nt, the GC content is between 40% and 60%, and the maximum homopolymer length does not exceed 4; the binding free energy change ΔG range between primer and template is -14kcal / mol to -17kcal / mol. Nucleic acid inhibition probe has a certain overlap range with FP, and the length is generally 7 to 12 nt.

[0109] The system used for qPCR-based nucleic acid inhibition probe displacement amplification is shown in Table 2, including: Hieff UNICON premix, forward primer (FP), reverse primer with universal sequence (RP-UP), DNA template, nucleic acid inhibition probe. The program used for qPCR amplification is: 95℃:2min—[(95℃:15s—60℃:60s)×50]. The experiment has verified that the nucleic acid inhibition probe can amplify the mutant template by 1,000 times (median), and the qPCR single inhibition probe displacement amplification technology can detect low-frequency mutations with an allele frequency of 0.01% to 0.5%.

[0110] The methods used for amplifying multiple targets in the present invention include two methods: ordinary multiple qPCR and multiple nucleic acid inhibition probe displacement amplification based on qPCR to amplify mutations.

[0111] The technology described in the present invention mainly includes the following steps:

[0112] 1. Multiplex qPCR Amplification

[0113] The multiplex qPCR amplification technology described in the present invention is based on single-plex qPCR, and the primer sequences required for multiple target amplification are mixed. The amplification system and conditions are as follows:

[0114] Common multiplex qPCR amplification, amplification system includes but is not limited to the following:

[0115] 1×Hieff UNICON system, annealing and extension at 60℃ for 1min. The system used for qPCR amplification is shown in Table 3, which includes 1×Hieff UNICON premix, multiple FP mixture, multiple RP-UP mixture, multiple DNA templates and enzyme-free water. The program used for qPCR amplification is: 95℃:2min—[(95℃:15s—60℃:1min)×50].

[0116] 0.5×Hieff UNICON system, annealing and extension at 60°C for 2 min. The system used for qPCR amplification is shown in Table 4, which includes 0.5×Hieff UNICON premix, multiple FP mixture, multiple RP-UP mixture, multiple DNA templates and enzyme-free water. The program used for qPCR amplification is: 95°C: 2 min—[(95°C: 15 s—60°C: 2 min)×50].

[0117] 1×PowerUp system, annealing and extension at 60℃ for 1min. The system used for qPCR amplification is shown in Table 5, which includes 1×PowerUp premix, multiple FP mixture, multiple RP-UP mixture, multiple DNA templates and enzyme-free water. The program used for qPCR amplification is: 50℃:2min—95℃:2min—[(95℃:15s—60℃:1min)×50].

[0118] 0.5×PowerUp system, annealing and extension at 60℃ for 2min. The system used for qPCR amplification is shown in Table 6, which includes 1×PowerUp premix, multiple FP mixture, multiple RP-UP mixture, multiple DNA templates and enzyme-free water. The program used for qPCR amplification is: 50℃:2min—95℃:2min—[(95℃:15s—60℃:2min)×50].

[0119] 1×Phusion system, annealing and extension at 72℃ for 0.5min. The system used for qPCR amplification is shown in Table 7, which includes 1×HF reaction solution, Phusion DNA polymerase, multiple FP mixture, multiple RP-UP mixture, multiple DNA templates and enzyme-free water. The program used for qPCR amplification is: 98℃:3min—[(98℃:10s—63℃:15s—72℃:30s)×50].

[0120] Based on qPCR multiple nucleic acid inhibition probe displacement amplification mutation amplification technology, the amplification system includes but is not limited to the following:

[0121] 1×Hieff UNICON nucleic acid inhibition probe system, annealing and extension at 60°C for 1 min. The system used for qPCR amplification is shown in Table 8, which includes 1×Hieff UNICON premix, multiple FP mixture, multiple RP-UP mixture, multiple DNA template, and multiple inhibition probe mixture. The program used for qPCR amplification is: 95°C: 2 min—[(95°C: 15 s—60°C: 1 min)×50].

[0122] 0.5×Hieff UNICON nucleic acid inhibition probe system, annealing and extension at 60°C for 2 min. The system used for qPCR amplification is shown in Table 9, which includes 0.5×Hieff UNICON premix, multiple FP mixture, multiple RP-UP mixture, multiple DNA template, and multiple inhibition probe mixture. The program used for qPCR amplification is: 95°C: 2 min—[(95°C: 15 s—60°C: 2 min)×50].

[0123] 1×PowerUp nucleic acid inhibition probe system, annealing and extension at 60℃ for 1min. The system used for qPCR amplification is shown in Table 10, which includes 1×PowerUp premix, multiple FP mixture, multiple RP-UP mixture, multiple DNA template, and multiple inhibition probe mixture. The program used for qPCR amplification is: 50℃:2min—95℃:2min—[(95℃:15s—60℃:1min)×50].

[0124] 0.5×PowerUp nucleic acid inhibition probe system, annealing and extension at 60°C for 2 min. The system used for qPCR amplification is shown in Table 11, which includes 1×PowerUp premix, multiple FP mixture, multiple RP-UP mixture, multiple DNA template, and multiple inhibition probe mixture. The program used for qPCR amplification is: 50°C: 2 min—95°C: 2 min—[(95°C: 15 s—60°C: 2 min)×50].

[0125] 1×Phusion nucleic acid inhibition probe system, annealing and extension at 72℃ for 0.5min. The system used for qPCR amplification is shown in Table 12, which includes 1×HF reaction solution, Phusion DNA polymerase, multiple FP mixture, multiple RP-UP mixture, multiple DNA template, and multiple inhibition probe mixture. The program used for qPCR amplification is: 98℃: 3min—[(98℃: 10s—63℃: 15s—72℃: 30s)×50].

[0126] In the multiple nucleic acid inhibition probe displacement amplification technology system based on qPCR in the present invention, the wild-type template will be inhibited, so the Ct value is the largest, and the mutant template is amplified normally, and the Ct value is the smallest, so the initial screening of mutations can be achieved according to the Ct value. For example, the 16-plex qPCR nucleic acid inhibition probe displacement amplification technology in the present invention can cover 347 gene loci and detect 0.2% mutations.

[0127] 2. Sequencing

[0128] The amplification products of common multiplex qPCR include wild-type and mutant DNA fragments, while the products of multiplex nucleic acid inhibition probe displacement amplification based on qPCR are only mutant DNA fragments; the above amplification products can be sequenced using specific RP or UP. The results of common multiplex qPCR amplification products using specific RP sequencing are single sequences, and the sequencing map can be added to the reference sequence library. The results of UP sequencing are mixed sequences of amplification; the results of qPCR-based multiplex nucleic acid inhibition probe displacement amplification technology products using specific RP sequencing are single mutant sequences, and the results of UP sequencing are mixed mutant sequences.

[0129] 3. AI Graph Analysis

[0130] Similarity comparison is a key link in sequence analysis, and its methods are diverse, including but not limited to Euclidean distance, Manhattan distance, cosine similarity, Hamming similarity, point-to-point comparison, and dynamic time warping. Each method has its own unique characteristics and applicable scenarios, and the choice of method depends on the specific data characteristics and comparison requirements. The present invention intercepts each data point from the mixed spectrum and the single spectrum, so the methods such as Euclidean distance, Manhattan distance, point-to-point comparison, and dynamic time warping are all suitable choices.

[0131] The spectrum analysis technology of the present invention mainly uses point-to-point comparison method and dynamic time warping method to calculate similarity. Based on the length information of the amplified sequence, if the horizontal coordinate is ignored, the point-to-point comparison method is the most direct method for comparing the coordinate data of two sequences. In addition, although the single spectrum graphs existing in the mixed spectrum will be superimposed, in the case of few mutations, the single spectrum existing in the mixed spectrum can still be distinguished, and the single spectrum without mutation will not appear in the mixed spectrum, so this method is used to compare the similarity between the single spectrum of each sequence and the mixed spectrum.

[0132] Although the length information of each single spectrum sequence is known, its single spectrum may be stretched or shrunk in the mixed spectrum, so the influence of the horizontal axis needs to be considered. The dynamic time warping (DTW) algorithm can effectively handle the step size difference of the time axis (i.e., the horizontal axis of the sequence spectrogram) and accurately capture the similarity between sequences. To further improve the matching accuracy, we introduce an arbitrary single spectrum sequence combination strategy based on DTW to effectively reduce the excessively high DTW distance.

[0133] In the present invention, the UP sequencing results are firstly quality controlled. Theoretically, the length of DNA sequences is greater than 100bp, so the patterns with sequencing length less than 100bp are discarded; then the UP sequencing mixed spectrum is analyzed using AI spectrum analysis technology, and the UP sequencing mixed spectrum is compared with the sequence in the reference sequence library to determine the sequence containing the amplified mutation. The AI ​​spectrum analysis method includes but is not limited to the AI ​​method of dynamic time warping and AI spectrum point-to-point comparison.

[0134] 3.1DTW Algorithm

[0135] DTW algorithm analysis Sequencing mixed spectrum analysis is mainly divided into two methods: a) DTW single spectrum random combination comparison, this method randomly mixes single spectra to form multiple combined spectra, and then compares the sequencing mixed spectrum with a series of combined spectra. The sequence contained in the combined spectrum with the minimum DTW value is the mutation sequence contained in the sequencing mixed spectrum; b) DTW single spectrum step-by-step comparison and superposition, this method is based on the length of the sequence in the sequencing mixed spectrum. First, the longest sequence is compared with the single spectrum and multiplied by the weight of the DTW minimum value of the single spectrum. According to the sequence length in the sequencing mixed spectrum, it is compared with the single spectrum in the reference sequence library in turn. The single spectrum on the comparison is multiplied by the corresponding DTW minimum value weight and superimposed, and finally a single spectrum with the combined spectrum is formed. Figure 1 The sequence contained in the combined spectrum is the mutation sequence contained in the mixed spectrum.

[0136] a) DTW single spectrum random combination comparison

[0137] First, the product after multiple nucleic acid inhibition probe displacement amplification based on qPCR is a mutant DNA sequence. The product is sequenced using UP to obtain information about the mutant sequence in the panel. Secondly, after quality control of the sequencing data, the coordinates of the single spectrum and the mixed spectrum are read as four channels of A, G, C, and T respectively using the code, and then the values ​​are normalized. Then, each spectrum with mutations is filled with 0 to the mixed spectrum binding site according to the position information, and those without mutations are filled with 0 according to the sequence length. After filling with 0, each sequence will be added in any permutation combination, and then compared with the mixed spectrum using the DTW algorithm. The analysis process is as follows ( Figure 3A ):

[0138] Obtain all DNA amplification sequences and length information;

[0139] Starting from the FP end sequence, 1100 coordinate data of the specific primer sequencing single spectrum and UP sequencing mixed spectrum were extracted from 5' to 3' direction; when reading the mixed spectrum, the longest sequence found was used as a reference to extract 1100 coordinate data, and other sequences were filled with 0 to the length of the longest sequence in the mixed spectrum;

[0140] Divide by the maximum value of the ordinate of each sequence and multiply by 100 to standardize each sequence to reduce the difference between the mixed spectrum ordinate data and the single spectrum ordinate data;

[0141] The single spectrum sequence coordinate data can be combined arbitrarily, and the formula is: Where n is the number of single spectrum sequences blended in, and k is the number of any combinations;

[0142] The single spectrum coordinate data of any combination are added together, and the combined sequence coordinate data need to be divided by the maximum value of the combined single spectrum coordinate data and multiplied by 100 to standardize the combined sequence coordinate data;

[0143] According to the four channels of A, G, C, and T, the DTW algorithm is used to compare the combined sequence and the mixed spectrum ordinate data. The formula is as follows: Dist(x,y)=|xy| D(i,j)=Dist(i,j)+min[D(i-1,j),D(i,j-1),D(i-1,j-1)]

[0144] Where x,i is the vertical coordinate data of the mixed spectrum, and y,j is the vertical coordinate data of the combined sequence. If D(i,j) is smaller, it means that the similarity between the combined sequence and the mixed spectrum is higher;

[0145] The DTW distances of the combined sequence and the mixed spectrum sequence of the four channels are added and divided by 4, and the (average) shortest DTW distance between the combined sequence and the mixed spectrum sequence is calculated to obtain the most similar single spectrum combined sequence, and the sequence contained in the mixed spectrum is predicted based on the combined single spectrum sequence with the (average) shortest DTW distance.

[0146] b) DTW single spectrum step-by-step comparison and superposition

[0147] First, the product after qPCR-based multiple nucleic acid inhibition probe displacement amplification is a mutant DNA sequence. The product is sequenced using UP to obtain information about the mutant sequence in the panel. Secondly, after quality control of the sequencing data, the coordinates of the single spectrum and the mixed spectrum are read as four channels of A, G, C, and T respectively using code, and then the values ​​are normalized. Each spectrum with mutation is then filled with 0 to the mixed spectrum binding site according to the position information, and those without mutation are filled with 0 according to the sequence length. After filling with 0, each sequence will be gradually compared with the mixed spectrum using the DTW algorithm according to the sequence length, and the corresponding weight is found by finding the minimum DTW distance. The analysis process is as follows ( Figure 3B ):

[0148] Obtain all DNA amplification sequences and length information;

[0149] Starting from the FP end sequence, 1100 coordinate data of the specific primer sequencing single spectrum and UP sequencing mixed spectrum were extracted from 5' to 3' direction; when reading the mixed spectrum, the longest sequence found was used as a reference to extract 1100 coordinate data, and other sequences were filled with 0 to the longest sequence length in the most mixed spectrum;

[0150] Divide by the maximum value of the ordinate of each sequence and multiply by 100 to standardize each sequence to reduce the difference between the mixed spectrum ordinate data and the single spectrum ordinate data;

[0151] Identify the longest sequence in the mixed spectrum by naked eyes, compare the abscissa data of the peaks in the mixed spectrum and the longest sequence, calculate the stretching ratio of the mixed spectrum, and multiply the abscissa of the mixed spectrum by the stretching ratio to reduce the offset of the abscissa data;

[0152] The sequence position information is used to determine whether the single spectrum sequence exists in the mixed spectrum, starting from the longest sequence, until the shortest single spectrum sequence.

[0153] According to the four channels A, G, C, and T, the DTW algorithm is used to gradually compare the single spectrum sequence and the mixed spectrum ordinate data at the corresponding position. The position of the DTW comparison is the length of the single spectrum sequence to the next single spectrum sequence. A minimum of 100 data points must be intercepted. The minimum weight of each channel of the single spectrum is found by calculating the minimum DTW distance of each channel. The weight ranges from 0 to 2, and the interval is 0.1. If the lowest DTW distance of three channels calculated by the mixed spectrum comparison is not the single spectrum at this position or two channels of the single spectrum at this position are ≤0.1, it is determined that the single spectrum does not exist in the mixed spectrum, and the weight is 0. If it is determined to exist, the sequence is added to the combined sequence according to each channel multiplied by the weight. The formula is as follows: Dist(x,y)=|xy| D(i,j)=Dist(i,j)+min[D(i-1,j),D(i,j-1),D(i-1,j-1)] ifD(i,j)=min[D(i,j)],n=best_n Form=1∶k Combined_Sequence[m]=Combined_Sequence[m-1]+n*sequence[m]

[0154] Where x,i is the vertical coordinate data of the mixed spectrum, and y,j is the vertical coordinate data of the combined sequence. If D(i,j) is smaller, it means that the similarity between the combined sequence and the mixed spectrum is higher, n is the optimal weight to make the DTW distance the shortest, k is the total number of sequences, and m is the sequence number (1-k);

[0155] Based on the single spectra determined to be mixed into the mixed spectrum, the mutation sequence in the mixed spectrum is predicted.

[0156] 3.2 Point-to-point comparison

[0157] First, the product after qPCR-based multiple nucleic acid inhibition probe displacement amplification is a mutant DNA sequence. The product is sequenced using UP to obtain information about the mutant sequence in the panel. Secondly, after quality control of the sequencing data, the coordinates of the sequencing spectra of each multiplex and UP are read as four channels, A, G, C, and T, respectively, using code. Then, each spectrum with mutations is filled with 0 to the UP sequencing mixed spectrum binding site according to the position information, and the single spectrum without mutations is filled with 0 according to the sequence length. After filling with 0, each single spectrum will be compared point-to-point with the mixed spectrum according to the four channels and then the average value is taken. The analysis process is as follows ( Figure 4 ):

[0158] Obtain all DNA amplification sequences and length information;

[0159] According to the four channels of A, G, C, and T, 1100 coordinate data of the specific primer sequencing single spectrum and UP sequencing mixed spectrum are extracted from the FP end in the 5' to 3' direction; when reading the mixed spectrum, the longest sequence found is used as a reference to extract 1100 coordinate data, and other sequences are filled with 0 to the longest sequence length in the most mixed spectrum;

[0160] Divide by the maximum value of the ordinate of each sequence and multiply by 100 to standardize each sequence to reduce the difference between the mixed spectrum ordinate data and the single spectrum ordinate data;

[0161] Based on the four channels of A, G, C, and T, the similarity of each FP sequence data in the UP sequence data is calculated using a point-to-point similarity comparison algorithm. The closer to 1, the more similar. The algorithm formula is: Where f(x) is the point-to-point similarity between the single spectrum sequence and the mixed spectrum sequence, i is the four channels A, G, C, and T, and j is the vertical coordinate data of each channel;

[0162] According to the 5-fold data, f(x) less than 6 indicates that the single spectrum sequence has a mutation in the mixed spectrum sequence. By calculating all single spectrum sequences less than 6, the sequence contained in the mixed spectrum is predicted.

[0163] 3.3 Machine Learning Spectral Interpretation

[0164] Considering the signal diversity of the mixed sequencing spectrum, we can build a mixed spectrum analysis model through machine learning methods, so that the model can autonomously learn and effectively identify the mutant sequences in the mixed spectrum. First, we use a PCR amplification system that does not contain nucleic acid inhibition probes to amplify the target sequence, and sequence all the amplified products using the RP and UP of the sequence to construct a sequence library containing all single spectrum sequences, combined spectrum sequences, and sequencing mixed spectrum sequences. The reference sequence library is composed of single spectra sequenced with the corresponding RP of the sequence. Next, we select single spectra, a large number of combined spectra, and a small number of sequencing mixed spectra of the reference sequence library as training sets, and use machine learning algorithms for analysis. Samples that are successfully analyzed will be marked as 1, and samples that fail to be analyzed will be marked as 0. Then, we predict and evaluate the model, and optimize the model through multiple verifications and evaluations to determine the final algorithm model. Finally, we input the remaining combined spectra and sequencing mixed spectra into the model as test sets, test the model, determine the single spectrum sequences present in it, and evaluate the results. The ultimate goal is to achieve automatic identification of mixed spectra of unknown mutations to ensure the accuracy and reliability of the model. The entire machine learning analysis process is as follows ( Figure 5 ):

[0165] Data collection: Collect all sample data and preprocess them, including feature extraction, selection, and standardization.

[0166] Data division: The data set is divided into a training set and a test set in a ratio of 7:3. The training set contains all single spectra, a large number of combination spectra and a small number of sequencing mixed spectra, while the test set contains a small number of combination spectra and a large number of sequencing mixed spectra.

[0167] Training model: Use single spectra, a large number of combined spectra, and a small number of sequencing mixed spectra of the reference sequence library as training sets to establish a mathematical (mapping) relationship between input X (spectrum) and output Y (existing single spectrum sequence) variables. Such X, Y pairs constitute the label data used to build the model, so as to learn how to predict the output from the input, and then perform parameter tuning and feature selection to train a complete model.

[0168] Model evaluation and optimization: Continuously evaluate and optimize the model, including data set segmentation, hyperparameter tuning, and model replacement, and use performance metrics for quantitative evaluation.

[0169] Test model: The remaining combined spectra and sequencing mixed spectra are used as test sets and input into the final test model to make predictions and evaluate the performance indicators of the model.

[0170] Application Example 1: Cell Line Contamination Identification

[0171] This application example first uses the NCBI 1000Genome database, the NA18537 genome as the wild-type template, and the NA18562 genome as the mutant template to perform single- and multiple-plex verification of 16 single nucleotide polymorphism (SNP) sites; this system can be applied to cell line contamination identification. Schematic diagram of low-frequency mutation detection based on 16-plex nucleic acid inhibition probe amplification and qPCR results are shown in Fig. 6A , a schematic diagram for detecting cell line contamination is shown in Figure 6B .

[0172] The experimental process of cell line contamination identification is as follows:

[0173] 1.1 qPCR-based single-plex nucleic acid inhibition probe displacement amplification mutation amplification

[0174] The nucleic acid primers and nucleic acid inhibition probes designed for 16 SNPs in this example are shown in Tables 13 and 14 respectively. First, the artificially synthesized wild-type DNA template and the mutant DNA template were mixed in different proportions to prepare templates with different variant allele frequencies (VAF). Using 20 ng of artificially synthesized DNA template, the designed primers and nucleic acid inhibition probes were tested according to the PCR amplification system and amplification conditions described in the DNA multiple mutation amplification detection process. The primers and nucleic acid inhibition probes designed for each SNP were amplified and tested on the synthetic templates with different VAFs. The qPCR amplification curves are shown in Tables 13 and 14. Fig. 7A Using only 20 ng of DNA, the qPCR-based single-plex nucleic acid inhibition probe displacement amplification technology can detect ultra-low frequency mutations as low as 0.01%. Figure 7B The Ct values ​​of qPCR amplification using 0%, 1%, and 100% VAF DNA templates are summarized. The greater the difference between 0% and 100% VAF Ct, the higher the detection sensitivity.

[0175] 1.2 Testing and optimization of 8-plex nucleic acid inhibition probe displacement amplification based on qPCR

[0176] Based on the test results of 16 sets of primers and nucleic acid inhibition probes in single-plex amplification in this embodiment, 8 SNPs (m1, m2, m5, m7, m8, m9, m11, m14) and 8 types of FP, RP-UP and nucleic acid inhibition probes required for amplification were first screened, and mixed solutions of FP, RP-UP and nucleic acid inhibition probes were prepared respectively. The system and procedure used for amplification include the following:

[0177] ①1×Hieff UNICON nucleic acid inhibition probe system, annealing and extension at 60℃ for 1min. The system used for amplification is shown in Table 8, including 1×Hieff UNICON premix, 8 FP mixtures, 8 RP-UP mixtures, 8 DNA templates, and 8 inhibition probe mixtures. The program used for amplification is: 95℃:2min—[(95℃:15s—60℃:1min)×50].

[0178] ②0.5×Hieff UNICON nucleic acid inhibition probe system, annealing and extension at 60℃ for 2min. The system used for amplification is shown in Table 9, which specifically includes 0.5×Hieff UNICON premix, 8 FP mixtures, 8 RP-UP mixtures, 8 DNA templates, 8 inhibition probe mixtures and enzyme-free water. The program used for amplification is: 95℃:2min—[(95℃:15s—60℃:2min)×50].

[0179] ③1×PowerUp nucleic acid inhibition probe system, annealing and extension at 60℃ for 1min. The system used for amplification is shown in Table 10, including 1×PowerUp premix, 8 FP mixtures, 8 RP-UP mixtures, 8 DNA templates, and 8 inhibition probe mixtures. The program used for amplification is: 50℃:2min—95℃:2min—[(95℃:15s—60℃:1min)×50].

[0180] ④0.5×PowerUp nucleic acid inhibition probe system, annealing and extension at 60℃ for 2min. The system used for amplification is shown in Table 11, including 1×PowerUp premix, 8 FP mixtures, 8 RP-UP mixtures, 8 DNA templates, 8 inhibition probe mixtures and enzyme-free water. The program used for amplification is: 50℃:2min—95℃:2min—[(95℃:15s—60℃:2min)×50].

[0181] ⑤1×Phusion nucleic acid inhibition probe system, annealing and extension at 72℃ for 0.5min. The system used for amplification is shown in Table 12, including 1×HF reaction solution, Phusion DNA polymerase, 8 FP mixed solutions, 8 RP-UP mixed solutions, 8 DNA templates, and 8 inhibition probe mixed solutions. The program used for amplification is: 98℃: 3min—[(98℃: 10s—63℃: 15s—72℃: 30s)×50].

[0182] ⑥1×Hieff UNICON system, annealing and extension at 60℃ for 1min. The system used for amplification is shown in Table 8, including 1×Hieff UNICON premix, 8 FP mixtures, 8 RP-UP mixtures, 8 DNA templates and enzyme-free water. The program used for amplification is: 95℃:2min—[(95℃:15s—60℃:1min)×50].

[0183] ⑦0.5×Hieff UNICON system, annealing and extension at 60℃ for 2min. The system used for amplification is shown in Table 9, including 0.5×Hieff UNICON premix, 8 FP mixtures, 8 RP-UP mixtures, 8 DNA templates and enzyme-free water. The program used for amplification is: 95℃:2min—[(95℃:15s—60℃:2min)×50].

[0184] ⑧1×PowerUp system, annealing and extension at 60℃ for 1min. The system used for amplification is shown in Table 10, including 1×PowerUp premix, 8 FP mixed solutions, 8 RP-UP mixed solutions, 8 DNA templates and enzyme-free water. The program used for amplification is: 50℃:2min—95℃:2min—[(95℃:15s—60℃:1min)×50].

[0185] ⑨0.5×PowerUp system, annealing and extension at 60℃ for 2min. The system used for amplification is shown in Table 11, including 1×PowerUp premix, 8 FP mixed solutions, 8 RP-UP mixed solutions, 8 DNA templates and enzyme-free water. The program used for amplification is: 50℃:2min—95℃:2min—[(95℃:15s—60℃:2min)×50].

[0186] ⑩1×Phusion system, annealing and extension at 72℃ for 0.5min. The system used for amplification is shown in Table 12, including 1×HF reaction solution, Phusion DNA polymerase, 8 FP mixed solutions, 8 RP-UP mixed solutions, 8 DNA templates and enzyme-free water. The program used for amplification is: 98℃: 3min—[(98℃: 10s—63℃: 15s—72℃: 30s)×50].

[0187] In this embodiment, the above ①─⑩ include an 8-plex qPCR amplification reaction system and an 8-plex nucleic acid inhibition probe displacement amplification reaction system based on qPCR. This embodiment uses 20ng DNA templates with 0%, 1% and 100% VAF, and based on the above method, the 8-plex nucleic acid inhibition probe displacement amplification technology based on qPCR is tested and optimized respectively. The intended purpose of this embodiment is that after using the nucleic acid inhibition probe, the amplification of the wild-type DNA template will be inhibited and the Ct will increase; the amplification of the mutant DNA template will not be affected, and the Ct will not change significantly.

[0188] When the synthetic DNA templates were amplified using the above ①─⑤, the Ct value of the wild-type DNA template (0% VAF) qPCR amplification was the largest, while the Ct value of the mutant DNA template (100% VAF) qPCR amplification became smaller. When the synthetic DNA templates were amplified using the above ①─⑤, the wild-type and mutant DNA templates could be distinguished based on the Ct value. When the synthetic DNA templates were amplified using method ②, the wild-type and mutant DNA templates were best distinguished, with a ΔCt of 11, and it is expected that a 2048-fold enrichment of the mutant DNA template could be achieved. When the synthetic DNA templates were amplified using method ⑤, the wild-type and mutant DNA templates were worst distinguished, with a ΔCt of 0.6, and it is expected that a 1.5-fold enrichment of the mutant DNA template could be achieved. By using optimized key parameters, the detection performance can be continuously improved, including the selection of suitable DNA polymerases, premix systems, and annealing extension times, which helps to increase the enrichment capacity of mutant templates in each round of qPCR amplification and increase the difference between wild-type and mutant signals. Fig. 8A shown.

[0189] Compared with ⑥─⑩, the above ①─⑤ used 8-fold nucleic acid inhibition probes. For the amplification of mutant DNA templates, this example compares the changes in Ct values ​​for qPCR amplification of mutant DNA templates using 8-fold nucleic acid inhibition probes and without nucleic acid inhibition probes. The intended purpose is that after using the nucleic acid inhibition probe, the amplification of the wild-type DNA template will be inhibited and the Ct will increase; the amplification of the mutant DNA template will not be affected, the Ct will not change significantly, and the Ct value will not lag (delay). The acceptable ΔCt change for the inhibition of mutant DNA template amplification is 1. Comparing the use of the above methods ① and ⑥, 1×HieffUNICON nucleic acid inhibition probe system, 60°C annealing and extension for 1min to amplify the mutant DNA template, there was no significant change in the Ct value. Comparing the use of the above methods ② and ⑦, 0.5×Hieff UNICON nucleic acid inhibition probe system, 60°C annealing and extension for 2min to amplify the mutant DNA template, the ΔCt was 1.6, indicating that the nucleic acid inhibition probe slightly inhibited the amplification of the mutant DNA template. Comparing the use of methods ③ and ⑧ above, when the mutant DNA template was amplified by 1×PowerUp nucleic acid inhibition probe system and annealing and extending at 60°C for 1 min, the Ct value did not show a significant change. Comparing the use of methods ④ and ⑨ above, when the mutant DNA template was amplified by 0.5×PowerUp nucleic acid inhibition probe system and annealing and extending at 60°C for 1 min, the ΔCt was 1.8, indicating that the nucleic acid inhibition probe slightly inhibited the amplification of the mutant DNA template. Comparing the use of methods ⑤ and ⑩ above, when the mutant DNA template was amplified by 0.5×PowerUp nucleic acid inhibition probe system and annealing and extending at 60°C for 2 min, the ΔCt was 2.2, indicating that the nucleic acid inhibition probe slightly inhibited the amplification of the mutant DNA template. Figure 8B shown.

[0190] Based on the 8 DNA synthesis templates m1, m2, m5, m7, m8, m9, m11, and m14, this embodiment mixes several wild-type templates with wild-type templates to prepare DNA mixed templates containing 1, 24, and 8 mutant templates, respectively. For example, the m1 mutant sequence is mixed with the remaining 7 wild-type templates to prepare a mixed template containing 1 mutation; the m1, m2, m5, and m7 mutant sequences are mixed with the remaining 4 wild-type templates to prepare a mixed template containing 4 mutations. In the above mixed templates, the concentration of each template is 6000 copies / μL, and then it is mixed with 8 wild-type templates with a concentration of 6000 copies / μL in different proportions to obtain DNA mixed templates with different VAFs and mutation numbers, and then an 8-plex nucleic acid inhibition probe based on qPCR is used for displacement amplification. The standard curve results show that as the number of mutant templates increases, the Ct value becomes smaller, such as Figure 8C shown.

[0191] 1.3 16-plex qPCR test

[0192] Laboratory cell line contamination is an issue that cannot be ignored in scientific research. Cell line contamination may cause distortion of experimental data, thereby affecting the accuracy and credibility of research conclusions. Therefore, timely detection and identification of laboratory cell line contamination is crucial to ensure the reliability of scientific research. The present invention provides a cell line contamination identification method, which uses a multiple nucleic acid inhibition probe displacement amplification technology based on qPCR, performs a preliminary screening of whether the sample is contaminated according to the change in the qPCR amplification Ct value, and then combines sequencing and spectrum analysis results to determine the specific contamination type. It is expected that laboratory cell line contamination identification will show great potential.

[0193] This example provides a 5-plex small panel containing rs10230708 (m1), rs10964389 (m4), rs7104025 (m9), rs9466035 (m11), and rs1338945 (m13) for genotyping of East Asians and a 16-plex panel, which is of great significance in analyzing individual genotype differences and identifying cell contamination in biological laboratories. This example uses 6000 copies / μL of DNA template to mix NA18537 wild-type DNA with DNA from different cell lines at a ratio of 99:1. In this example, the 5-plex nucleic acid inhibition probe displacement amplification technology based on qPCR uses a 1×Hieff UNICON nucleic acid inhibition probe system, annealing and extending at 60°C for 1 min. The system used for amplification is shown in Table 12, specifically including 1×Hieff UNICON premix, 5 FP mixtures, 5 RP-UP mixtures, 5 DNA templates, and 5 inhibition probe mixtures. The amplification program was 95°C: 2 min—[(95°C: 15 s—60°C: 1 min)×50].

[0194] In this example, 6 cell lines were detected using the qPCR-based 5-plex nucleic acid inhibition probe displacement amplification technique, and 2 cell lines were detected using the qPCR-based 16-plex nucleic acid inhibition probe displacement amplification technique. When the qPCR-based 5-plex and 16-plex nucleic acid inhibition probe displacement amplification techniques were used to detect NA18537 mixed with DNA from different cell lines, the Ct value of the contaminated NA18537 cell line was significantly smaller than that of the uncontaminated NA18537 cell line, as shown in Figure 2. Fig.9A Therefore, this embodiment provides a method for determining whether a sample is contaminated and the degree of contamination based on the Ct value, and the specific contamination type can be obtained through sequencing.

[0195] In this example, after DNA was amplified by a 5-plex nucleic acid inhibition probe replacement technique based on qPCR, NA18537 and contaminated NA18537 were sequenced using a specific RP. Among them, the number of mutations in the five cell lines S1-S5 were 3, 4, 4, 2, and 4, respectively. Fig. 9B shown.

[0196] Taking the HeLa cell line as an example, after the NA18537 cell line DNA is amplified using the 5-fold nucleic acid inhibition probe replacement technology based on qPCR, if the Ct value of qPCR becomes smaller than that of the wild type, it can be preliminarily judged that the wild type is contaminated; then the PCR amplification product is sequenced using UP, and the mutation sequence contained is analyzed using the AI ​​spectrum analysis technology provided by the present invention ( Fig. 9C The accuracy of the six sequencing results of each mixed spectrum analyzed by the point-to-point comparison algorithm was 100% ( Fig.9D ).

[0197] The products of ordinary multiplex qPCR amplification are sequenced using specific RP sequencing to produce a single sequence, and the sequencing map can be added to the reference sequence library. The results of UP sequencing are mixed sequences of amplification. The products of qPCR-based multiplex nucleic acid inhibition probe displacement amplification technology are sequenced using specific RP sequencing to produce a single mutant sequence, and the results of UP sequencing are mixed sequences of mutants.

[0198] In this example, NA18537 was used as the wild type. Based on the above-mentioned 16SNP primers (Table 13) and nucleic acid inhibition probes (Table 14), the Ct values ​​of the wild type and the cell line to be tested based on qPCR multiple nucleic acid inhibition probe amplification can be used to preliminarily determine whether the cell line is contaminated. If the cell line is contaminated, the Ct value will become smaller, and then the genotype and contamination type can be determined in combination with the Sanger sequencing results.

[0199] The 16-plex nucleic acid inhibition probe displacement amplification technology based on qPCR in this embodiment uses a 1×Hieff UNICON nucleic acid inhibition probe system, annealing and extension at 60°C for 1 min. The system used for amplification is shown in Table 12, specifically including 1×Hieff UNICON premix, 16 FP mixtures, 16 RP-UP mixtures, 16 DNA templates, and 16 inhibition probe mixtures. The amplification program is 95°C: 2min—[(95°C: 15s—60°C: 1min)×50]. The amplification curves of DNA of different VAFs amplified by the 16-plex nucleic acid inhibition probe displacement technology based on qPCR in this embodiment are shown in Table 12. Fig. 10A. In this embodiment, the 16-plex nucleic acid inhibition probe displacement amplification technology based on qPCR can detect 16 SNP mutations with a minimum of 0.2% according to the change in the Ct value of qPCR. By comparing the sequencing results of the wild-type template and the 0.5% mutant template, the Sanger sequencing results can distinguish mutations as low as 0.5%, indicating that the multiple nucleic acid inhibition probe displacement amplification technology based on qPCR can effectively inhibit wild-type and amplify mutant signals ( Fig. 10B This example provides a 16-plex panel that can be used to distinguish (but not limited to) 37 cell lines ( Fig. 10C ).

[0200] Application Example 2: High-throughput Detection of Low-Frequency Tumor Mutations

[0201] Colorectal cancer is the third most common cancer in the world and the second most deadly cancer. Gene mutation detection plays an important role in the diagnosis, targeted drug treatment, and prognosis and recurrence of colorectal cancer. It is of great significance for early cancer screening, improving the success rate of treatment, and improving patient survival. This application example designs a multiple colorectal panel, covering the mutation sites recommended by the Chinese expert consensus on clinical detection of colorectal cancer molecular markers, including but not limited to APC, BRAF, ERBB2, KRAS, NRAS, HRAS, PIK3CA and other genes. These gene mutations are mutually exclusive in most cases, such as RAS, BRAF, and ERBB2 mutations. Therefore, the multiple nucleic acid inhibition probe displacement amplification technology based on qPCR is used to amplify and enrich the mutations in colorectal cancer samples, and the use of UP one-time sequencing can realize the detection of multiple low-frequency mutations in colorectal cancer. Taking the gene mutation of colorectal cancer as an example, this example constructs a 12-plex panel, which detects 5 common mutation genes of colorectal cancer in a single tube, covering 201 gene loci, and is applied to the detection of DNA mutations in clinical tissue samples of colorectal cancer patients. The mutations are enriched by multiple nucleic acid inhibition amplification based on qPCR and then sequenced using UP. The mutation sequences analyzed by sequencing are consistent with the NGS results. The results of detecting mutations in colorectal cancer patients using multiple nucleic acid inhibition probe displacement amplification technology based on qPCR are as follows Fig.11 In addition, this method uses less clinical samples, making it of great significance for the detection of cfDNA in blood.

[0202] The primers and sequences used in the colorectal cancer multiple mutation detection panel are shown in Table 15, and the nucleic acid inhibition probes and sequences are shown in Table 16. The contents disclosed in Chinese invention patent application 202410198644X are also incorporated herein by reference.

[0203] This example uses 12-plex nucleic acid inhibition probe displacement technology based on qPCR to amplify 12 synthetic templates. The system and procedures used include the following:

[0204] 1×Hieff UNICON nucleic acid inhibition probe system, annealing and extension at 60°C for 1 min. The system used for amplification is shown in Table 8, including 1×Hieff UNICON premix, 12 FP mixtures, 12 RP-UP mixtures, 12 DNA templates, and 12 inhibition probe mixtures. The program used for amplification is: 95°C: 2 min—[(95°C: 15 s—60°C: 1 min×50].

[0205] 0.5×Hieff UNICON nucleic acid inhibition probe system, annealing and extension at 60°C for 2 min. The system used for amplification is shown in Table 9, including 0.5×Hieff UNICON premix, 12 FP mixtures, 12 RP-UP mixtures, 12 DNA templates, 12 inhibition probe mixtures and enzyme-free water. The program used for amplification is: 95°C: 2 min—[(95°C: 15 s—60°C: 2 min)×50].

[0206] In this example, DNA from colorectal cancer patients was amplified by common 12-plex qPCR, and then the qPCR amplification products were sequenced using specific RP to obtain a sequencing map of a single sequence and to construct a reference sequence library. After 12-plex nucleic acid inhibition probe displacement amplification based on qPCR, the amplification products were sequenced using UP to obtain a mixed spectrum of all mutation sequences. The analysis of the sequencing mixed spectrum can be achieved using AI spectrum analysis technology. This example provides the tissue DNA test results of three colorectal cancer patients. The test results are as follows: Fig.12 As shown. UP sequencing of colorectal cancer patients 1 and 2 showed a single sequence map, each with one mutation, KRAS c.35G>A and KRAS c.175G>A. The sequencing result of colorectal cancer patient 3 was a mixed map, and the analysis results of the mixed map showed that the mutations were located in exon 2 of KRAS and exon 10 of PIK3CA, respectively.

[0207] There are hotspot mutation gene sites in colorectal cancer patients, for example, KRAS gene G12, G13, NRAS gene G12, G13, BRAF gene V600, TP53 gene R245, R248, etc. These mutation hotspot clusters can also use long-fragment nucleic acid inhibition probes to achieve high-throughput enrichment detection of mutations. For multiple mutations in colorectal cancer DNA, another preferred solution in this embodiment is to use long-fragment nucleic acid inhibition probe mutation enrichment technology. The mutation enrichment interval of ordinary nucleic acid inhibition probe displacement amplification technology is 19-25nt, and the enrichment interval of long-fragment nucleic acid inhibition probes can be up to 44nt ( Fig.13), so long-fragment nucleic acid inhibition probes can also achieve high-throughput amplification detection of mutations. Taking the KRAS gene as an example, a single-tube single-strip long-fragment nucleic acid inhibition probe designed for mutations on exon 2 can cover 81 mutation types including G12 and G13, and can be used for multiple mutation detection in colorectal cancer patients (Angew Chem Int Ed, 2024, 63, e202400551).

[0208] The 3' end blocking group of the long nucleic acid inhibition probe includes but is not limited to phosphorylation modification, spacer, peptide nucleic acid (PNA), locked nucleic acid (LNA) and DXXDM, where D is a DNA residue, X is a C3 spacer, and M is a mismatch. The DNA polymerase used in the long nucleic acid inhibition probe displacement amplification technology includes but is not limited to Phusion and Q5 high-fidelity enzymes.

[0209] See Table 17 for primers and sequences used in the long-fragment nucleic acid inhibition probe displacement amplification technique to detect RAS gene mutations.

[0210] The system used for long-fragment nucleic acid inhibition probe amplification is shown in Table 18, which specifically includes 5×Q5 reaction solution, FP, RP, inhibition probe, dNTPs, Q5 hot start high-fidelity DNA polymerase, SYBR green, DNA template and enzyme-free water. The amplification conditions used are: 98℃:30s—[(98℃:10s—70℃:30s—72℃:3min)×50].

[0211] The long-fragment nucleic acid inhibition probe displacement amplification mutation amplification technology described in this example can amplify mutations of different positions and types within the enrichment interval. The template addition amount was 6000 copies, and a KRAS long-fragment nucleic acid inhibition probe was used to test 7 mutation templates with different mutations at different positions within the mutation enrichment interval, and their mutation enrichment factor (EF) was calculated. The enrichment factor is defined as 2^[Ct WT -Ct MT ], where CtWT and CtMT represent the median of Ct of three repeated experiments on wild-type and mutant DNA, respectively. The KRAS long-fragment nucleic acid inhibition probe has the best enrichment effect on G13S (G>A) after qPCR amplification, with an enrichment multiple of 4100 and a mutation detection limit of 0.22%. The KRAS long-fragment nucleic acid inhibition probe has a relatively weak enrichment effect on mutations located at the 3' end after qPCR amplification. The KRAS long-fragment nucleic acid inhibition probe has a median enrichment of 650 times for single mutations within the enrichment interval after qPCR amplification. The present embodiment provides a qPCR-based long-fragment nucleic acid inhibition probe mutation enrichment technology, which can cover the detection of 81 mutation types for a single long-fragment nucleic acid inhibition probe for KRAS, such as Fig.14 shown.

[0212] In this example, a standard curve for G13S (G>A) detection was constructed by using 6000 copies of different VAF KRAS templates using a long-fragment nucleic acid inhibition probe displacement amplification mutation amplification technique based on qPCR. The standard curve for G13S (G>A) detection by the long-fragment nucleic acid inhibition probe displacement amplification mutation amplification technique based on qPCR is: Ct = -3.4811LogVAF + 24.453, as shown in Figure 2. Fig.15A shown.

[0213] In this example, a standard curve for G12R (G>C) detection was constructed by using 6000 copies of different VAF NRAS templates using a long-fragment nucleic acid inhibition probe displacement amplification mutation amplification technique based on qPCR. The standard curve for G12R (G>C) detection by the long-fragment nucleic acid inhibition probe displacement amplification mutation amplification technique based on qPCR is: Ct = -3.1008LogVAF + 26.414, as shown in FIG. Fig. 15B shown.

[0214] This example uses a qPCR-based long-fragment nucleic acid inhibition probe displacement amplification mutation amplification technique to detect KRAS and NRAS gene mutations in patients with colorectal cancer. This example shows the qPCR detection and sequencing results of two patients with colorectal cancer with KRAS mutations ( Fig.16A ), qPCR detection and sequencing results of 2 patients with colorectal cancer with NRAS mutation ( Fig. 16B ). Both colorectal cancer patients 4 and 5 had KRAS:G12A (G>C) mutations, with mutation frequencies of 1.12% and 1.80%, respectively. Colorectal cancer patient 6 had NRAS:G12D (G>A) mutations, with a mutation frequency of 58.40%. Colorectal cancer patient 7 had NRAS:G13R (G>C) mutations, with a mutation frequency of 53.21%.

[0215] The following are the details of Tables 1 to 18:

[0216] Table 1: 1×Hieff UNICON singleplex qPCR amplification reaction system. Element Volume (μL) Final concentration 2×Hieff UNICON master mix 5 1× FP (2 μM) 1 0.2μM RP-UP (2 μM) 1 0.2μM Enzyme-free water 2 - DNA template (20ng) 1 2ng / μl

[0217] Table 2: 1×Hieff UNICON singleplex nucleic acid inhibition probe displacement amplification qPCR reaction system. Element Volume (μL) Final concentration 2×Hieff UNICON master mix 5 1× FP (2 μM) 1 0.2μM RP-UP (2 μM) 1 0.2μM B(10μM) 2 2μM DNA template (20ng) 1 2ng / μl

[0218] Table 3: 1×Hieff UNICON multiplex qPCR amplification reaction system. Element Volume (μL) Final concentration 2×Hieff UNICON master mix 5 1× FP mixed solution (2 μM each) 1 0.2μM RP-UP mixed solution (2 μM each) 1 0.2μM Enzyme-free water 2 - DNA template (20ng) 1 2ng / μl

[0219] Table 4: 0.5×Hieff UNICON multiplex qPCR amplification reaction system. Element Volume (μL) Final concentration 2×Hieff UNICON master mix 2.5 0.5× Enzyme-free water 4.5 - FP mixed solution (2 μM each) 1 0.2μM RP-UP mixed solution (2 μM each) 1 0.2μM DNA template (20ng) 1 2ng / μl

[0220] Table 5: 1×PowerUp multiplex qPCR reaction amplification system. Element Volume (μl) Final concentration 2×PowerUp premix 2.5 0.5× Enzyme-free water 4.5 - FP mixed solution (2 μM each) 1 0.2μM RP-UP mixed solution (2 μM each) 1 0.2μM DNA template (20ng) 1 2ng / μl

[0221] Table 6: 0.5×PowerUp multiplex qPCR reaction amplification system. Element Volume (μl) Final concentration 2×PowerUp premix 5 1× FP mixed solution (2 μM each) 1 0.2μM RP-UP mixed solution (2 μM each) 1 0.2μM Enzyme-free water 2 - DNA template (20ng) 1 2ng

[0222] Table 7: 1×Phusion multiplex qPCR reaction amplification system. Element Volume (μL) Final concentration 5×HF reaction solution 2 1× Enzyme-free water 4.6 - FP mixed solution (2 μM each) 1 0.2μM RP-UP mixed solution (2 μM each) 1 0.2μM dNTPs(10mM) 0.2 200μM 100×SYBR green 0.1 1× Phusion DNA Polymerase 0.1 0.02U / μl DNA template (20ng) 1 2ng / μl

[0223] Table 8: 1×Hieff UNICON multiplex nucleic acid inhibition probe displacement amplification qPCR reaction system.

[0224] Table 9: 0.5×Hieff UNICON multiplex nucleic acid inhibition probe amplification qPCR reaction system. Element Volume (μL) Final concentration 2×Hieff UNICON master mix 2.5 0.5× Enzyme-free water 2.5 - FP mixed solution (2 μM each) 1 0.2μM RP-UP mixed solution (2 μM each) 1 0.2μM Inhibitor probe mixture (10 μM each) 2 2μM DNA template (20ng) 1 2ng / μl

[0225] Table 10: 1×PowerUp multiplex nucleic acid inhibition probe displacement amplification qPCR reaction system. Element Volume (μl) Final concentration 2×PowerUp premix 5 1× FP mixed solution (2 μM each) 1 0.2μM RP-UP mixed solution (2 μM each) 1 0.2μM Inhibitor probe mixture (10 μM each) 2 2μM DNA template (20ng) 1 2ng / μl

[0226] Table 11: 0.5×PowerUp multiplex nucleic acid inhibition probe displacement amplification qPCR reaction system. Element Volume (μl) Final concentration 2×PowerUp premix 2.5 0.5× Enzyme-free water 2.5 - FP mixed solution (2 μM each) 1 0.2μM RP-UP mixed solution (2 μM each) 1 0.2μM Inhibitor probe mixture (10 μM each) 2 2μM DNA template (20ng) 1 2ng / μl

[0227] Table 12: 1×Phusion multiplex nucleic acid inhibition probe displacement amplification qPCR reaction system. Element Volume (μL) Final concentration 5×HF reaction solution 2 1× Enzyme-free water 2.6 - FP mixed solution (2 μM each) 1 0.2μM RP-UP mixed solution (2 μM each) 1 0.2μM Inhibitor probe mixture (10 μM each) 2 2 dNTPs(10mM) 0.2 200μM 100×SYBR green 0.1 1× Phusion DNA Polymerase 0.1 0.02U / μl DNA template (20ng) 1 2ng / μl

[0228] Table 13: Primers and sequences used for 16SNP amplification (SEQ ID NO. 1-33). Primers sequence rs10230708(m1,C>T)_FP ACCAATGGGAGTCACTGCTG rs17149369(m2,A>G)_FP GGATTCCCTAAGCTCTTCAATATTGC rs10741037(m3,C>T)_FP CACTTTATCAGACACAGTTATGTGCT rs10964389(m4,C>T)_FP CAAAGTTGATAAATTAAAGGACTAAGGCAC rs12259813(m5,A>G)_FP TGTAGGAGAGATTGGGCTAGAGAG rs3789806(m6,C>G)_FP CTTGTATAGACGGTAAAATAAACACCAAGA rs2510152(m7,G>T)_FP ACCCAGGTGAGTTTTGTTTCACAT rs10510620(m8,C>G)_FP TCCGCAAAACCTACAATCTCTGAA rs7104025(m9,C>G)_FP TCAGATGCTTTAGGCTCATGAGTTA rs10508599(m10,T>G)_FP GGGTTAAAATCTTTTGCTTTCATATTGAGC rs9466035(m11,A>G)_FP CCTAACACCAGTTCTTCCTCCAC rs4712498(m12,A>T)_FP GTTTTTATATGTTAGTGTCCCCATGGTA rs1338945(m13,C>A)_FP GAAATATTGCTGGGGTCAGCG rs12478327(m14,C>A)_FP ACTTCTGCCAACATTCAAATTCAGG rs12213948(m15,C>T)_FP TGAAAGACGTCACAGCAAGGT rs12541300(m16,C>G)_FP ACAGAAACCAATTACCTATGAGGAGTAA rs10230708(m1,C>T)_RP TAAGTGGAAAGAACTGGGGTGTC rs17149369(m2,A>G)_RP AAATGTAGTTCTATTATGGTCAGCACAC rs10741037(m3,C>T)_RP GCAAAAATTACTATACCGACTTTAATAACGAAA rs10964389(m4,C>T)_RP ACTTACTGAGCACATGGCCTG rs12259813(m5,A>G)_RP GCTGAGCTGTCACATCACTTCA rs3789806(m6,C>G)_RP AGGCACCAGAAGTCATCAGAATG rs2510152(m7,G>T)_RP TGAAACCACATACACACAAATTCACT rs10510620(m8,C>G)_RP AAAAGATAATGTTCTTGTTTATATGCCCTTG rs7104025(m9,C>G)_RP TACAGCAACTCACAAACTAATGACTCT rs10508599(m10,T>G)_RP TGCAAAATGAAGCACAGCCC rs9466035(m11,A>G)_RP TGGGCTCAAGTGATCCACCTA rs4712498(m12,A>T)_RP AGACCGAACTTGTTGCGAATCA rs1338945(m13,C>A)_RP AGGAGAGGGAGGAGCATGG rs12478327(m14,C>A)_RP GTGCAAGCTGGAGGCACT rs12213948(m15,C>T)_RP AATCCAGTGACATTCTTTAAACTGTCTT rs12541300(m16,C>G)_RP GCTGTGTAGCTTGGCAAATTAACTA UP GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAG

[0229] Table 14: Nucleic acid inhibition probes and sequences used for 16SNP amplification (SEQ ID NO. 34-49).

[0230] Table 15: Primers and sequences used in the colorectal cancer multiple mutation detection panel (SEQ ID NO.50-77). Primer Sequence APC_R213_FP GGTACCTGCCAGGATATGGAAAA APC_R876_FP AGTGGTGGAGATCTGCAAACC APC_R1450_FP GCAGTAGGTGCTTTATTTTTAGGTAC BRAF_V600E_FP TGGGACCCACTCCATCGAGAT ERBB2_S310_FP GGCAGACGAGGGTGCAG ERBB2_V777_FP CCCCAGGAAGCATACGTGATG ERBB2_V842_FP GGGATGAGCTACCTGGAGGATG KRAS_A146_FP GATCTGTATTTATTTCAGTGTTACTTACCTGT PIK3CA_E542_FP CAATTTCTACACGAGATCCTCTCTC PIK3CA_H1047_FP GAGTATTTCATGAAACAAATGAATGATGC TP53_R196_FP TGGCCCCTCCTCAGCATC TP53_G245_FP TGTAACAGTTCCTGCATGGGC TP53_R273_FP CTGGGACGGAACAGCTTTGA TP53_R280_FP TGTTTGTGCCTGTCCTGGGA APC_R213_RP GCCAAAATAAACACAGCCTTATGTCT APC_R876_RP AGATAGAAGTTTGGAGAGAGAACGC APC_R1450_RP GTGATCTTCCAGATAGCCCTGG BRAF_V600E_RP TGCTTGCTCTGATAGGAAAATGAGATC ERBB2_S310_RP CGGTAATGCTGCTCATGGTG ERBB2_V777_RP CCGGACATGGTCTAAGAGGC ERBB2_V842_RP TCTGTCTCGTCAATGTCCAGC KRAS_A146_RP CTAGAACAGTAGACACAAAACAGGC PIK3CA_E542_RP ACATGCTGAGATCAGCCAAATTC PIK3CA_H1047_RP AGTGTGGAATCCAGAGTGAGC TP53_R196_RP GAGACCCCAGTTGCAAACCA TP53_G245_RP GGTCAGAGGCAAGCAGAGG TP53_R273_RP CCTGCTTGCTTACCTCGCTT TP53_R280_RP GTCCTGCTTGCTTACCTCGC

[0231] Table 16: Nucleic acid inhibition probes and sequences used in the colorectal cancer multiple mutation detection panel (SEQ ID NO.78-91).

[0232] Table 17: Primers and sequences used for detecting RAS gene mutations using long-fragment nucleic acid inhibition probe displacement amplification technology (SEQ ID NOs. 92-97). Primer and probe Sequence KRAS_FP TGACTGAATATAAACTTGTGGTAGTTGGAGC KRAS_RP TCTGTATCAAAGAATGGTCCTGCACCAGTAATATGCA KRAS_B GTAGTTGGAGCTGGTGGCGTAGGCAAGAGT / iSpC3 / / iSpC3 / CA NRAS_FP GTGTGAAATGACTGAGTACAAACTGGTGGTGGTTG NRAS_RP CCTCACCTCTATGGTGGGATCATATTCATCTACAAAGTG NRAS_B TGGTGGTTGGAGCAGGTGGTGTTGGGAAA / iSpC3 / / iSpC3 / CA

[0233] Table 18: 1×Q5 system long fragment nucleic acid inhibition probe displacement amplification qPCR reaction system. Element Volume (μL) Final concentration Enzyme-free water 2.6 - 5×Q5 reaction solution 2 1× FP (4 μM) 1 0.4μM RP(4μM) 1 0.4μM B(10μM) 2 2μM 10mM dNTPs 0.2 200μM Q5 Hot Start High-Fidelity DNA Polymerase 0.1 0.02U / μl 100×SYBR green 0.1 1× DNA template (20ng) 1 2ng / μl

[0234] The above is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. The above embodiments of the present invention can also be modified in various ways. All simple, equivalent changes and modifications made according to the claims and the description of the present invention fall within the scope of protection of the claims of the present invention. The contents not described in detail in the present invention are all conventional technical contents.

Claims

1. A single-tube multiplex qPCR mutation amplification detection technology based on AI graph analysis, It is characterized in that The following steps are involved: S1: Use multiplex qPCR to amplify all DNA targets, use specific reverse primer sequencing to obtain the sequencing results of all targets, and construct a reference sequence library; S2: Amplify all DNA targets using multiplex qPCR and obtain a mixed profile of qPCR products using universal primer sequencing; S3: Use AI map analysis technology to analyze the mixed map obtained in step S2, compare the mixed map with the sequences in the reference sequence library constructed in step S1, and determine the sequences containing amplified mutations, thereby achieving multiple detection of DNA and low-frequency mutations.

2. According to claim 1, the single-tube multiplex qPCR mutation amplification detection technology based on AI map analysis, It is characterized in that Steps S1 and S2 include: designing multiple nucleic acid inhibition probes for the wild-type DNA template, using nucleic acid inhibition probe displacement amplification technology based on qPCR to inhibit the wild-type DNA template and amplify the mutant DNA template normally, using specific reverse primer sequencing to obtain a single sequence of the mutant, using universal primer sequencing to obtain a mixed sequence of the mutant, and achieving initial screening of mutations based on changes in qPCR amplification Ct values.

3. According to claim 1, the single-tube multiplex qPCR mutation amplification detection technology based on AI map analysis, It is characterized in that In step S3, the AI ​​graph analysis technology uses a dynamic time warping algorithm and a point-to-point comparison algorithm to calculate similarity.

4. According to claim 3, the single-tube multiplex qPCR mutation amplification detection technology based on AI map analysis, It is characterized in that The dynamic time warping algorithm includes: a DTW single spectrum random combination comparison algorithm and a DTW single spectrum step-by-step comparison and superposition algorithm.

5. According to claim 4, the single-tube multiplex qPCR mutation amplification detection technology based on AI map analysis, It is characterized in that The analytical process of the DTW single spectrum random combination comparison algorithm is as follows: 1) Obtain all DNA amplification sequences and length information; 2) According to the four channels of A, G, C, and T, 1100 coordinate data of the specific primer sequencing single spectrum and the UP sequencing mixed spectrum are extracted from the 5' to 3' direction starting from the FP end; when reading the mixed spectrum, the longest sequence found is used as a reference to extract 1100 coordinate data, and other sequences are filled with 0 to the length of the longest sequence in the mixed spectrum; 3) Divide by the maximum value of the ordinate of each sequence and multiply by 100 to standardize each sequence to reduce the difference between the ordinate data of the mixed spectrum and the ordinate data of each single spectrum; 4) Arbitrarily combine the single spectrum sequence coordinate data, the formula is: Where n is the number of single spectrum sequences blended in, and k is the number of any combinations; 5) The single spectrum coordinate data of any combination are added together, and the combined sequence coordinate data need to be divided by the maximum value of the combined single spectrum coordinate data and multiplied by 100 to standardize the combined sequence coordinate data; 6) According to the four channels of A, G, C, and T, the DTW algorithm is used to compare the ordinate data of the combined sequence and the mixed spectrum. The formula is as follows: Dist(x,y)=|xy| D(i,j)=Sist(i,j)+min[D(i-1,j),D(i,j-1),D(i-1,j-1)] Among them, x,i is the ordinate data of the mixed spectrum, and y,j is the ordinate data of the combined sequence. If D(i,j) is smaller, it means that the similarity between the combined sequence and the mixed spectrum is higher; 7) Add the DTW distances of the obtained 4-channel combined sequence and the mixed spectrum sequence and divide them by 4, calculate the (average) shortest DTW distance between the combined sequence and the mixed spectrum sequence, obtain the most similar single spectrum combined sequence, and predict the sequence contained in the mixed spectrum based on the combined single spectrum sequence with the (average) shortest DTW distance.

6. According to claim 4, the single-tube multiplex qPCR mutation amplification detection technology based on AI map analysis, It is characterized in that The analytical process of the DTW single spectrum step-by-step comparison and superposition algorithm is as follows: 1) Obtain all DNA amplification sequences and length information; 2) According to the four channels of A, G, C, and T, 1100 coordinate data of the specific primer sequencing single spectrum and the UP sequencing mixed spectrum are extracted from the 5' to 3' direction starting from the FP end; when reading the mixed spectrum, the longest sequence found is used as a reference to extract 1100 coordinate data, and other sequences are filled with 0 to the length of the longest sequence in the mixed spectrum; 3) Divide by the maximum value of the ordinate of each sequence and multiply by 100 to standardize each sequence to reduce the difference between the ordinate data of the mixed spectrum and the ordinate data of each single spectrum; 4) identifying the longest sequence in the mixed spectrum by naked eyes, comparing the abscissa data of the peaks in the mixed spectrum and the longest sequence, calculating the stretching ratio of the mixed spectrum, and multiplying the abscissa of the mixed spectrum by the stretching ratio to reduce the offset of the abscissa data; 5) Using the sequence position information, starting from the longest sequence, determine whether the single spectrum sequence exists in the mixed spectrum until the shortest single spectrum sequence; 6) According to the four channels A, G, C, and T, the DTW algorithm is used to gradually compare the single spectrum sequence and the mixed spectrum ordinate data at the corresponding position. The position of the DTW comparison is the length of the single spectrum sequence to the next single spectrum sequence. A minimum of 100 data points must be intercepted. The minimum weight of each channel of the single spectrum is found by calculating the minimum DTW distance of each channel. The weight ranges from 0 to 2, and the interval is 0.

1. If the minimum DTW distance of three channels calculated by the mixed spectrum comparison is not the single spectrum at this position or two channels of the single spectrum at this position are ≤0.1, it is determined that the single spectrum does not exist in the mixed spectrum, and the weight is 0. If it is determined to exist, the sequence is added to the combined sequence according to each channel multiplied by the weight. The formula is as follows: Dist(x,y)=|xy| D(i,j)=Dist(i,j)+min[D(i-1,j),D(i,j-1),D(i-1,j-1)] if D(i,j)=min[D(i,j)],n=best_n For m=1∶k Combined_Sequence[m]=Combined_Sequence[m-1]+n*sequence[m] Where x,i is the ordinate data of the mixed graph, and y,j is the ordinate data of the combined sequence. If D(i,j) is smaller, it means that the similarity between the combined sequence and the mixed graph is higher, n is the optimal weight to make the DTW distance the shortest, k is the total number of sequences, and m is the sequence number (1-k); 7) Predicting the mutation sequence in the mixed spectrum based on the single spectrum determined to be mixed in the mixed spectrum.

7. According to claim 3, the single-tube multiplex qPCR mutation amplification detection technology based on AI map analysis, It is characterized in that The parsing process of the point-to-point comparison algorithm is as follows: 1) Obtain all DNA amplification sequences and length information; 2) According to the four channels of A, G, C, and T, 1100 coordinate data of the specific primer sequencing single spectrum and the UP sequencing mixed spectrum are extracted from the 5' to 3' direction starting from the FP end; when reading the mixed spectrum, the longest sequence found is used as a reference to extract 1100 coordinate data, and other sequences are filled with 0 to the length of the longest sequence in the mixed spectrum; 3) Divide by the maximum value of the ordinate of each sequence and multiply by 100 to standardize each sequence to reduce the difference between the ordinate data of the mixed spectrum and the ordinate data of each single spectrum; 4) Based on the four channels of A, G, C, and T, the similarity of each FP sequence data in the UP sequence data is calculated using a point-to-point similarity comparison algorithm. The closer to 1, the more similar. The algorithm formula is: Where f(x) is the point-to-point similarity between the single spectrum sequence and the mixed spectrum sequence, i is the four channels of A, G, C, and T, and j is the ordinate data of each channel; 5) According to the 5-fold data, if f(x) is less than 6, it means that the single spectrum sequence has a mutation in the mixed spectrum sequence. By calculating all the single spectrum sequences less than 6, the sequence contained in the mixed spectrum is predicted.

8. According to claim 3, the single-tube multiplex qPCR mutation amplification detection technology based on AI map analysis, It is characterized in that In step S3, the AI ​​graph analysis technology also includes constructing a hybrid graph analysis model so that the model can autonomously learn and effectively identify mutation sequences in the hybrid graph, wherein the machine learning analysis process is as follows: 1) Data collection: Collect all sample data and preprocess them, including feature extraction, selection and standardization; 2) Data partitioning: The data set is divided into a training set and a test set in a ratio of 7:

3. The training set contains all single spectra, a large number of combined spectra, and a small number of sequencing mixed spectra, while the test set contains a small number of combined spectra and a large number of sequencing mixed spectra. 3) Training model: Use single spectra of the reference sequence library, a large number of combined spectra, and a small number of sequencing mixed spectra as training sets to establish a mathematical (mapping) relationship between input X (spectrum) and output Y (existing single spectrum sequence) variables. Such X, Y pairs constitute the label data for building the model, so as to learn how to predict the output from the input, and then perform parameter tuning and feature selection to train a complete model; 4) Model evaluation and optimization: Continuously evaluate and optimize the model, including data set segmentation, hyperparameter tuning, and model replacement, and use performance metrics for quantitative evaluation. 5) Test model: The remaining combined spectra and sequencing mixed spectra are used as test sets and input into the final test model to make predictions and evaluate the performance indicators of the model.

9. According to claim 1, the single-tube multiplex qPCR mutation amplification detection technology based on AI map analysis, It is characterized in that In steps S1 and S2, long-fragment nucleic acid inhibition probes are used to extend the enrichment interval to 44 nt, thereby achieving high-throughput amplification detection of mutations.

10. An application of the single-tube multiplex qPCR mutation amplification detection technology based on AI map analysis according to any one of claims 1 to 9, It is characterized in that The applications include: tumor mutation detection, laboratory cell line contamination identification, genotyping and quantification, genetic testing, and companion diagnostics.

Images