Methods for obtaining and correcting sequence information of a target polynucleotide

Abstract

The present disclosure provides a method for obtaining sequence information of a target polynucleotide. Nucleotide substrate molecules comprising at least two different bases are included in a sequencing reaction solution; degenerate genetic code information is obtained. Two or more degenerate code information are compared with each other, and a person skilled in the art can determine whether the codes at the position are contradictory to each other. Any minor improvement to the sequencing method that can reduce the error rate of the original sequencing data can significantly reduce the error rate of the corrected code data by using the method described in the present disclosure.

Description

[0001] This application is a divisional application of the application filed on November 16, 2016, with application number 201680079417.9 and invention title "Method for Obtaining and Correcting Biological Sequence Information". Technical Field

[0002] This invention relates to a method for obtaining and correcting sequence information of target polynucleotides, belonging to the field of gene sequencing. Background Technology

[0003] High-throughput sequencing is a technology that has developed rapidly in recent years. Compared to traditional Sanger sequencing, the biggest advantage of high-throughput sequencing is that it can read massive amounts of sequence information simultaneously. Although its accuracy is not as high as traditional sequencing methods, the analysis of massive amounts of data can yield information beyond the sequence itself, such as gene expression levels and copy number changes.

[0004] Most mainstream sequencers today use the SBS (sequencing-by-synthesis) method, such as Solexa / Illumina, 454, and IonTorrent. These sequencers share similar structures, including fluidic systems, optical systems, and chip systems. The sequencing reaction occurs within the chip. The sequencing process is also similar, involving: introducing the reaction solution into the chip, initiating the SBS reaction, acquiring the signal, and washing. Then, a new round of sequencing is performed. This is a cyclical process. With each cycle, consecutive single-base non-degenerate sequence information (such as ACTGACTG) is obtained. However, high-throughput sequencers cannot completely eliminate sequencing errors. Sequencing errors can originate from: accidental or cumulative reaction errors, signal acquisition errors, errors from signal correction, etc. In existing sequencers, these chemical, optical, or software errors can become noise, which cannot be identified at a single readout site. They can only be eliminated through deep sequencing, utilizing multiple readouts of the same sequence at different sites. More accurate readouts are an important direction for the development of high-throughput sequencing. However, current technologies for improving accuracy primarily focus on optimizing the chemical reaction itself and subsequent image signal processing, without innovating the sequencing logic. Therefore, there is a need for improved sequencing methods. Summary of the Invention

[0005] This application claims priority to the following Chinese patent applications: Chinese patent application CN201510822361.9, filed November 19, 2015, entitled "A method for sequencing nucleotide molecules using phosphate-modified fluorophores"; Chinese patent application CN201510815685.X, filed November 19, 2015, entitled "A method for sequencing nucleotide substrate molecules using fluorophores with fluorescence switching properties"; Chinese patent application CN201510944878.5, filed December 12, 2015, entitled "A method for detecting and correcting sequence data errors in sequencing results"; and Chinese patent application CN201610899880.X, filed October 14, 2016, entitled "A method for reading sequence information from raw signals of high-throughput DNA sequencing," the entire contents of which are incorporated herein by reference.

[0006] The summary of this invention is not intended to limit the scope of the claimed subject matter. Other features, details, utility, and advantages of the claimed subject matter will become apparent from the detailed description of those aspects disclosed in the drawings and appended claims.

[0007] On the one hand, this document discloses a method for obtaining sequence information of a target polynucleotide, the method comprising: a) providing a first sequencing reagent to the target polynucleotide in the presence of a first polynucleotide replication catalyst, wherein the first sequencing reagent comprises two different nucleotide monomers each conjugated to a first label, and the nucleotide monomer / first label conjugate is substantially non-fluorescent until the nucleotide monomer is incorporated into the target polynucleotide according to its complementarity with the target polynucleotide; and b) providing a second sequencing reagent to the target polynucleotide in the presence of a second polynucleotide replication catalyst, wherein the second sequencing reagent comprises two different nucleotide monomers each conjugated to a second label, and the nucleotide monomer / second label conjugate is substantially non-fluorescent until the nucleotide monomer is incorporated into the target polynucleotide according to its complementarity with the target polynucleotide, and wherein the second sequencing reagent is provided subsequently after the first sequencing reagent is provided; and c) by means of steps a) and b). After incorporating a nucleotide monomer into a polynucleotide, fluorescence emission caused by a first label and a second label is detected to obtain at least partial sequence information of the target polynucleotide. The nucleotide monomers in the first and second sequencing reagents are selected from the group consisting of: 1) an adenine (A) nucleotide monomer and a thymine (T) / uracil (U) nucleotide monomer in one sequencing reagent and a cytosine (C) nucleotide monomer and a guanine (G) nucleotide monomer in another sequencing reagent; 2) an adenine (A) nucleotide monomer and a guanine (G) nucleotide monomer in one sequencing reagent and a cytosine (C) nucleotide monomer and a thymine (T) / uracil (U) nucleotide monomer in another sequencing reagent; and 3) an adenine (A) nucleotide monomer and a cytosine (C) nucleotide monomer in one sequencing reagent and a guanine (G) nucleotide monomer and a thymine (T) / uracil (U) nucleotide monomer in another sequencing reagent. In one embodiment, the first label of the two different nucleotide monomers in step a) and the second label of the two different nucleotide monomers in step b) are the same label. In another embodiment, the first marker comprises two distinct markers, wherein one of the first markers is identical to one of the second markers, and the other of the first markers is identical to the other of the second markers. In any of the foregoing embodiments, multiple rounds of steps a) and b) are performed, each round using a combination selected from combinations 1)-3). In another embodiment, obtaining at least two or three sets of sequence information in step c) comprises: using combination 1) to perform multiple rounds of steps a) and b) in a first sequencing reaction volume to obtain a first set of sequence information, using combination 2) to perform multiple rounds of steps a) and b) in a second sequencing reaction volume to obtain a second set of sequence information, and / or using combination 3) to perform multiple rounds of steps a) and b) in a third sequencing reaction volume to obtain a third set of sequence information. In one embodiment, the first, second, and third sets of sequence information are obtained in parallel from separate sequencing reaction volumes.In another embodiment, the first, second, and third sets of sequence information are obtained sequentially from the same sequencing reaction volume, and the product of the previous sequencing reaction is removed before starting the next sequencing reaction. In any of the foregoing embodiments, the method further includes comparing at least two or three sets of sequence information to reduce or eliminate sequence errors. In one embodiment, the comparison indicates that there are no errors in the obtained target polynucleotide sequence when at least two or three sets of sequence information are consistent with each other. In another embodiment, the comparison indicates that there are errors in the obtained target polynucleotide sequence when at least two or three sets of sequence information contain differences in at least one nucleotide residue of the target polynucleotide sequence. In one embodiment, the method further includes correcting at least one nucleotide residue in the obtained target polynucleotide sequence such that, after correction, at least two or three sets of sequence information are consistent with each other.

[0008] In another aspect, this document discloses a method for obtaining sequence information of a target polynucleotide, the method comprising: a) providing a first sequencing reagent to the target polynucleotide in the presence of a first polynucleotide replication catalyst, wherein the first sequencing reagent comprises three different nucleotide monomers each conjugated to a first label, and the nucleotide monomer / first label conjugate is substantially non-fluorescent until the nucleotide monomer is incorporated into the target polynucleotide according to its complementarity with the target polynucleotide; and b) providing a second sequencing reagent to the target polynucleotide in the presence of a second polynucleotide replication catalyst, wherein the second sequencing reagent comprises a nucleotide monomer conjugated to a second label, and the nucleotide monomer / second label conjugate is substantially non-fluorescent until the nucleotide monomer is incorporated into the target polynucleotide according to its complementarity with the target polynucleotide, and wherein the second sequencing reagent is provided before or after the provision of the first sequencing reagent; and c) detecting the first label and the second label after incorporation of the nucleotide monomer into the polynucleotide in steps a) and b). The fluorescence emission of the sequencing reagent yields sequence information of at least a portion of the target polynucleotide, wherein the nucleotide monomers in the first and second sequencing reagents are selected from the group consisting of: 1) a sequencing reagent containing cytosine (C) nucleotide monomers, guanine (G) nucleotide monomers, thymine (T) / uracil (U) nucleotide monomers, and adenine (A) nucleotide monomers; 2) a sequencing reagent containing adenine (A) nucleotide monomers, guanine (G) nucleotide monomers, thymine (T) / uracil (U) nucleotide monomers, and cytosine (C) nucleotide monomers; and 3) a sequencing reagent containing adenine (A) nucleotide monomers, cytosine (C) nucleotide monomers, thymine (T) / uracil (U) nucleotide monomers, and guanine (G) nucleotide monomers; and 4) a sequencing reagent containing adenine (A) nucleotide monomers, cytosine (C) nucleotide monomers, and guanine (G) nucleotide monomers, and another sequencing reagent containing thymine (T) / uracil (U) nucleotide monomers. In one embodiment, the first label of the three different nucleotide monomers in step a) and the second label of one nucleotide monomer in step b) are the same label. In any of the foregoing embodiments, multiple rounds of steps a) and b) are performed, each round using a combination selected from combinations 1)-4). In one embodiment, obtaining at least two, three, or four sets of sequence information in step c) is achieved by: performing multiple rounds of steps a) and b) in a first sequencing reaction volume using combination 1) to obtain a first set of sequence information; performing multiple rounds of steps a) and b) in a second sequencing reaction volume using combination 2) to obtain a second set of sequence information; performing multiple rounds of steps a) and b) in a third sequencing reaction volume using combination 3) to obtain a third set of sequence information; and / or performing multiple rounds of steps a) and b) in a fourth sequencing reaction volume using combination 4) to obtain a fourth set of sequence information.In one implementation, the first, second, third, and fourth sets of sequence information are obtained in parallel from separate sequencing reaction volumes. In another implementation, the first, second, third, and fourth sets of sequence information are obtained sequentially from the same sequencing reaction volume, and the product of the previous sequencing reaction is removed before starting the next sequencing reaction. In any of the foregoing implementations, the method further includes comparing at least two, three, or four sets of sequence information to reduce or eliminate sequence errors. In one implementation, the comparison indicates that there are no errors in the obtained target polynucleotide sequence when at least two, three, or four sets of sequence information are consistent with each other. On the one hand, when using a single-color sequencing method, at least three sets of sequence information are required to monitor sequencing errors. On the other hand, when using a two-color sequencing method, only two sets of sequence information are required to detect sequencing errors because the information from the two fluorescent labels provides an additional set of information for sequence comparison.

[0009] In another embodiment, when at least two, three, or four sets of sequence information contain differences at at least one nucleotide residue in the target polynucleotide sequence, the comparison indicates an error in the obtained target polynucleotide sequence. In one embodiment, the method further includes correcting at least one nucleotide residue in the obtained target polynucleotide sequence such that, after correction, at least two, three, or four sets of sequence information are consistent with each other. On one hand, at least one nucleotide residue is corrected by deletion or insertion at the erroneous position to achieve the correct sequence. On the other hand, each insertion at the erroneous position extends the sequence by at least one nucleotide, and sequence information from one or more rounds of sequencing is compared with the extended sequence to achieve the corrected sequence. On the other hand, each deletion at the erroneous position shortens the sequence by at least one nucleotide, and sequence information from one or more rounds of sequencing is compared with the shortened sequence to achieve the corrected sequence.

[0010] In another aspect, this document discloses a computer-readable medium including instructions for correcting sequencing information errors. In one aspect, the instructions include: a) receiving sequencing information of a target polynucleotide and a reference polynucleotide; and b) correcting the sequencing information of the target polynucleotide using any of the methods disclosed herein for correcting sequencing information.

[0011] On the other hand, a computer system for sequencing is provided, the system including the computer-readable medium disclosed herein. Attached Figure Description

[0012] Figure 1 The method for correcting sequence data errors is shown.

[0013] Figure 2The data distribution for groups 1 through 5 is shown in violin and box plots. Black represents encoding accuracy, and gray represents decoding accuracy. Groups 1 through 5 are presented from left to right in the sequence.

[0014] Figure 3 A frequency distribution histogram is displayed, showing the number of signals modified during decoding for each of the 5000 sequence data.

[0015] Figure 4 This displays the correlation between the number of signals that are erroneous during encoding and the number of signals that are erroneously modified during decoding. The horizontal axis represents the number of signals that are erroneous during encoding, and the vertical axis represents the correlation between the number of signals that are erroneously modified during decoding. The grayscale of the color indicates the proportion of times that point is counted out of all sequences.

[0016] Figure 5 The solid-phase PCR process is shown.

[0017] Figure 6 According to one aspect of the present invention, the principle of degenerate base fluorescence gene sequencing is shown.

[0018] Figure 7 According to one aspect of the invention, degenerate base-calling results are shown.

[0019] Figure 8 According to one aspect of the present invention, an information communication model for ECC sequencing is shown.

[0020] Figure 9 According to one aspect of the invention, a sequence decoding result using dynamic programming is shown.

[0021] Figure 10 According to one aspect of the invention, display decoding improves the accuracy of ECC sequencing.

[0022] Figure 11 According to one aspect of the invention, a cyclic range distribution of three base combinations is shown.

[0023] Figure 12 This shows an example of the layer and node traversal order of the rating matrix structure.

[0024] Figure 13 According to one aspect of the invention, a state transition network for a hidden Markov model of ECC decoding is shown.

[0025] Figure 14 The diagram shows a simulated distribution of accuracy before and after decoding, according to one aspect of the invention.

[0026] Figure 15An example decoding result is shown. Detailed Implementation

[0027] The following provides a detailed description of one or more embodiments of the claimed subject matter, along with accompanying drawings illustrating the principles of the claimed subject matter. The claimed subject matter is described in conjunction with such embodiments, but is not limited to any specific embodiment. It should be understood that the claimed subject matter can be embodied in various forms and encompasses many alternatives, modifications, and equivalents. Therefore, the specific details disclosed herein should not be construed as limiting, but rather serve as the basis for the claims and as a representative basis for teaching those skilled in the art to adopt the claimed subject matter in virtually any suitable detailed system, structure, or manner. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the invention. These details are provided for illustrative purposes only, and the claimed subject matter may be practiced according to the claims without some or all of these specific details. It should be understood that other embodiments may be used and structural changes may be made without departing from the scope of the claimed subject matter. It should be understood that the various features and functions described in one or more individual embodiments are not limited to their application to the specific embodiments in which they are described. Rather, they may be applied individually or in some combination to one or more other embodiments of this disclosure, whether or not such embodiments are described, and whether or not such features are presented as part of the described embodiments. For clarity, known technical materials in the relevant technical fields of the claimed subject matter have not been described in detail to avoid unnecessarily obscuring the claimed subject matter.

[0028] All technical terms, symbols, and other technical and scientific terms used herein are intended to have the same meaning as commonly understood by one of ordinary skill in the art to which the claimed subject matter pertains, unless otherwise defined. In some instances, for clarity and / or convenience of reference, terms with their commonly understood meanings are defined herein, and such definitions are incorporated herein, but this should not necessarily be construed as indicating a material difference from the meanings commonly understood in the art. Many of the techniques and procedures described or referenced herein are known to those skilled in the art and are commonly used when employing conventional methods.

[0029] All publications referenced in this application, including patent documents, scientific papers, and databases, are incorporated herein by reference in their entirety for all purposes, just as each individual publication is incorporated individually by reference. Where a definition set forth herein is contrary to or otherwise inconsistent with a definition set forth in a patent, patent application, publication, or other publication incorporated herein by reference, the definition set forth herein shall prevail. References to publications or documents are not intended as an admission that any of them is applicable prior art, nor do they constitute any admission of the content or dates of such publications or documents.

[0030] Unless otherwise stated, all headings are for the reader's convenience and should not be used to limit the meaning of the text following them.

[0031] Unless otherwise indicated, the practices provided will employ conventional techniques and descriptions of organic chemistry, polymer technology, molecular biology (including recombinant technologies), cell biology, biochemistry, and sequencing technologies, which are understood by one of ordinary skill in the art. Such conventional techniques include peptide and protein synthesis and modification, polynucleotide synthesis and modification, polymer array synthesis, polynucleotide hybridization and ligation, and detection of hybridization using labels. Specific descriptions of suitable techniques can be obtained by referring to the examples herein. However, other equivalent conventional procedures may of course be used. Such routine techniques and descriptions can be found in standard laboratory manuals, such as Green et al., Genome Analysis: A Laboratory Manual Series (Volumes I-IV) (1999); Weiner, Gabriel, and Stephens, Genetic Variation: A Laboratory Manual (2007); Dieffenbach and Dveksler, PCR Primer: A Laboratory Manual (2003); Bowtell and Sambrook, DNA Microarrays: A Molecular Cloning Manual (2003); Mount, Bioinformatics: Sequence and Genome Analysis (2004); Sambrook and Russell, Condensed Protocols from Molecular Cloning: A Laboratory Manual (2006); and Sambrook and Russell, Molecular Cloning: A Laboratory Manual (2002) (all from ColdSpring Harbor Laboratory Press); Ausubel et al., Current Protocols in Molecular Biology (1987); and T. Brown, Essential Molecular Biology. Biology (1991), IRL Press; Goeddel, ed., Gene Expression Technology (1991), Academic Press; A. Bothwell et al., ed., Methods for Cloning and Analysis of Eukaryotic Genes (1990), Bartlett Publ.; M.Kriegler, Gene Transfer and Expression (1990), Stockton Press; R. Wu et al., eds., Recombinant DNAMethodology (1989), Academic Press; M. McPherson et al., PCR: A Practical Approach (1991), IRL Press at Oxford University Press; Stryer, Biochemistry (4th Edition) (1995), WH Freeman, New York NY; Gait, Oligonucleotide Synthesis: A Practical Approach (2002), IRL Press, London; Nelson and Cox, Lehninger, Principles of Biochemistry (2000) 3rd ed., WHFreeman Pub., New York, NY; Berg, et al., Biochemistry (2002) 5th ed., WHFreeman Pub., New York, NY; D.Weir & C. Blackwell, ed., Handbook of Experimental Immunology(1996),Wiley-Blackwell; Cellular and Molecular Immunology (A. Abbas et al., WBSaunders Co. 1991, 1994); Current Protocols in Immunology (J. Coligan et al., eds., 1991). All references cited are incorporated herein by reference in their entirety for all purposes.

[0032] Throughout this disclosure, all aspects of the claimed subject matter are presented in a scope format. It should be understood that this scope format is for convenience and brevity only and should not be construed as a rigid limitation on the scope of the claimed subject matter. Therefore, the scope description should be considered to have specifically disclosed all possible sub-scopes, as well as individual numerical values ​​within that scope. For example, where a range of values ​​is provided, it should be understood that every intermediate value between the upper and lower limits of that range, as well as any other specified value or intermediate value within that specified range, is covered within the claimed subject matter. The upper and lower limits of these smaller ranges may be independently included within those smaller ranges and are also covered within the claimed subject matter, subject to any express exclusions within the specified scope. When the stated scope includes one or both of the stated limits, the scope extending beyond any or both of those included limits is also included within the claimed subject matter. This application is independent of the breadth of the scope. For example, a description of a range such as from 1 to 6 should be considered to specifically disclose subranges such as from 1 to 3, from 1 to 4, from 1 to 5, from 2 to 4, from 2 to 6, from 3 to 6, etc., as well as individual numerical values ​​within that range, such as 1, 2, 3, 4, 5, and 6.

[0033] I. Definition

[0034] Unless the context clearly indicates otherwise, as used herein, the singular forms “a / an” and “the” include plural indicators. For example, “a / an” means “at least one” or “one or more”. It should be understood that aspects and variations described herein include “consisting of” and / or “substantially composed of” aspects and variations.

[0035] As used herein, the term "about" refers to a common range of error for a corresponding value that is readily known to those skilled in the art. References to "about" a value or parameter herein include (and describe) implementations for said value or parameter itself. For example, a description of "about X" includes a description of "X" itself.

[0036] The terms “polynucleotide,” “oligonucleotide,” “nucleic acid,” and “nucleic acid molecule” are used interchangeably herein to refer to polymeric forms of nucleotides of any length, and include ribonucleotides, deoxyribonucleotides, and their analogues or mixtures. The terms include triple-stranded, double-stranded, and single-stranded deoxyribonucleic acid (“DNA”), and triple-stranded, double-stranded, and single-stranded ribonucleic acid (“RNA”). They also include polynucleotides in their unmodified forms, as well as those modified, for example by alkylation and / or by end-capping. More specifically, the terms “polynucleotide,” “oligonucleotide,” “nucleic acid,” and “nucleic acid molecule” include polydeoxyribonucleotides (containing 2-deoxy-D-ribose), polyribonucleotides (containing D-ribose), including tRNA, rRNA, hRNA, and mRNA (whether spliced ​​or unspliced), any other type of polynucleotide that is an N- or C-glycoside of a purine or pyrimidine base, and other polymers containing a non-nucleotide backbone, such as polyamides (e.g., peptide nucleic acids (“PNA”)) and polymorpholino polymers (commercially available from Anti-Virals, Inc., Corvallis, OR, as with Neugene), as well as other synthetic sequence-specific nucleic acid polymers, provided that the polymer contains nucleic acid bases in a structure that allows base pairing and base stacking, such as in the construction of DNA and RNA. Therefore, these terms include, for example, 3'-deoxy-2',5'-DNA, oligodeoxyribonucleotides N3' to P5' phosphoramidites, 2'-O-alkyl-substituted RNA, hybrids between DNA and RNA or between PNA and DNA or RNA; they also include modifications of known types, such as labeling, alkylation; "capping"; substitution of one or more nucleotides with analogs; internucleotide modifications, such as modifications with uncharged bonds (e.g., methylphosphonates, phosphate triesters, phosphoramidites, carbamates, etc.); and modifications with negatively charged bonds (e.g., thiophosphates, dithiophosphates, etc.). Modifications include: phosphodiester bonds; modifications with positively charged linkages (e.g., aminoalkylphosphamide esters, aminoalkylphosphotriesters); modifications containing side-linked portions, such as proteins (including enzymes (e.g., nucleases), toxins, antibodies, signal peptides, poly-L-lysine, etc.); modifications with intercalating agents (e.g., acridine, psoralen, etc.); modifications containing chelates (e.g., chelates of metals, radioactive metals, boron, metal oxides, etc.); modifications containing alkylating agents; modifications with modified linkages (e.g., α-anomeric nucleic acids, etc.); and unmodified forms of polynucleotides or oligonucleotides. Nucleic acids typically contain phosphodiester bonds, but in some cases may include nucleic acid analogs with an alternative backbone, such as phosphoramide, dithiophosphate, or methylphosphorimide linkages; or peptide nucleic acid backbones and linkages.Other nucleic acid analogs include those with a bicyclic structure, including locked nucleic acids, positively charged backbones, nonionic backbones, and nonribose backbones. Molecular stability can be increased by modifying the ribose-phosphate backbone; for example, PNA:DNA hybrids exhibit greater stability in certain environments. The terms "polynucleotide," "oligonucleotide," "nucleic acid," and "nucleic acid molecule" can include any suitable length, such as at least 5, 6, 7, 8, 9, 10, 20, 30, 40, 50, 100, 200, 300, 400, 500, 1,000, or more nucleotides.

[0037] It should be understood that the terms “nucleoside” and “nucleotide” as used herein include not only known purine and pyrimidine bases but also other modified heterocyclic bases. Such modifications include methylated purines or pyrimidines, acylated purines or pyrimidines, or other heterocycles. Modified nucleosides or nucleotides may also include modifications to the glycosyl moiety, for example, where one or more hydroxyl groups are substituted with halogens, aliphatic groups, or functionalized to ethers, amines, etc. The term “nucleotide unit” is intended to encompass both nucleosides and nucleotides.

[0038] The terms "complementary" and "substantially complementary" encompass hybridization or base pairing, or the formation of a double helix between nucleotides or nucleic acids (e.g., between the two strands of a double-stranded DNA molecule or between primer binding sites on oligonucleotide primers and single-stranded nucleic acids). Complementary nucleotides are typically A and T (or A and U) or C and G. Two single-stranded RNA or DNA molecules can be described as substantially complementary when the nucleotides of one strand (optimally arranged and contrasted, with appropriate nucleotide insertions or deletions) pair with at least about 80% of the other strands, typically at least about 90% to about 95% of the other strands, or even about 98% to about 100% of the other strands. On the one hand, the two complementary sequences of nucleotides are capable of hybridization with opposing nucleotides, preferably with less than 25% mismatch, more preferably less than 15% mismatch, even more preferably less than 5% mismatch, and most preferably no mismatch. Preferably, the two molecules will hybridize under highly stringent conditions.

[0039] As used herein, “hybridization” can refer to the process by which two single-stranded polynucleotides non-covalently combine to form a stable double-stranded polynucleotide. The resulting double-stranded polynucleotide can be a “hybrid” or a “double strand.” Typical “hybridization conditions” include salt concentrations of approximately less than 1 M, typically less than about 500 mM, and possibly less than about 200 mM. A “hybridization buffer” contains a buffered salt solution, such as 5% SSPE or other such buffers known in the art. Hybridization temperatures can be as low as 5 °C, but are typically above 22 °C, more typically above about 30 °C, and usually above 37 °C. Hybridization is often performed under stringent conditions, which are conditions under which the sequence will hybridize with its target sequence but not with other non-complementary sequences. Stringent conditions are sequence-dependent and vary under different conditions. For example, for specific hybridization, longer fragments may require higher hybridization temperatures than shorter fragments. Since other factors, including the base composition and length of the complementary strand, the presence of organic solvents, and the degree of base mismatch, can affect the stringency of hybridization, the combination of parameters is more important than any single parameter as an absolute measure. Typically, stringent conditions are chosen to be approximately 5 °C lower than the Tm of a specific sequence at defined ionic strengths and pH. The melting temperature Tm can be the temperature at which a group of double-stranded nucleic acid molecules begins to partially dissociate into single strands. Several equations for calculating the Tm of nucleic acids are known in the art. As shown in standard references, a simple estimate of the Tm value can be calculated using the equation Tm = 81.5 + 0.41 (%G + C) when the nucleic acid is in aqueous solution at 1 M NaCl (see, for example, Anderson and Young, Quantitative Filter Hybridization, in Nucleic Acid Hybridization (1985)). Other references (e.g., Allai and Santa Lucia, Jr., Biochemistry, 36:10581-94 (1997)) include alternative methods for calculation, where structural and environmental factors, as well as sequence characteristics, are considered in the calculation of Tm.

[0040] Typically, the stability of heterozygotes is a function of ion concentration and temperature. Hybridization reactions are usually carried out under lower stringency conditions, followed by washing with different but higher stringency conditions. Exemplary stringency conditions include a pH of about 7.0 to about 8.3 and a temperature of at least 25°C, with a salt concentration of at least 0.01 M to no more than 1 M sodium ion (or other salt). For example, 5×SSPE conditions (750 mM NaCl, 50 mM sodium phosphate, 5 mM EDTA at pH 7.4) and a temperature of about 30°C are suitable for allele-specific hybridization, although suitable temperatures are related to the length of the hybridization region and / or GC content. On one hand, the “stringency of hybridization” in the mismatch percentage can be determined as follows: 1) High stringency: 0.1×SSPE, 0.1% SDS, 65°C; 2) Moderate stringency: 0.2×SSPE, 0.1% SDS, 50°C (also known as moderate stringency); and 3) Low stringency: 1.0×SSPE, 0.1% SDS, 50°C. It should be understood that equivalent stringency can be achieved using alternative buffers, salts, and temperatures. For example, moderately stringent hybridization can refer to conditions that allow nucleic acid molecules, such as probes, to bind to complementary nucleic acid molecules. Hybridized nucleic acid molecules generally have at least 60% identity, including, for example, any of at least 70%, 75%, 80%, 85%, 90%, or 95% identity. Moderately stringent conditions can be equivalent to hybridization at 42°C in 50% formamide, 5×Denhardt's solution, 5×SSPE, and 0.2% SDS, followed by washing at 42°C in 0.2×SSPE and 0.2% SDS. Highly stringent conditions, for example, can be provided as follows: hybridization at 42°C in 50% formamide, 5×Denhardt's solution, 5×SSPE, and 0.2% SDS, followed by washing at 65°C in 0.1×SSPE and 0.1% SDS. Low-strictness hybridization can refer to conditions equivalent to the following: hybridization at 22°C in 10% formamide, 5× Dunhardt solution, 6× SSPE, and 0.2% SDS, followed by washing at 37°C in 1× SSPE and 0.2% SDS. The Dunhardt solution contains 1% Ficoll, 1% polyvinylpyrrolidone, and 1% bovine serum albumin (BSA). 20× SSPE (sodium chloride, sodium phosphate, EDTA) contains 3M sodium chloride, 0.2M sodium phosphate, and 0.025M EDTA.Other suitable moderately stringent and highly stringent hybridization buffers and conditions are known to those skilled in the art and described, for example, in Sambrook et al., Molecular Cloning: A Laboratory Manual, 2nd ed., Cold Spring Harbor Press, Plainview, NY (1989); and Ausubel et al., Short Protocols in Molecular Biology, 4th ed., John Wiley & Sons (1999).

[0041] Alternatively, essential complementarity exists when the RNA or DNA strand hybridizes with its complement under selective hybridization conditions. Typically, selective hybridization occurs when there is at least about 65% complementarity, preferably at least about 75%, and more preferably at least about 90% complementarity, across a sequence segment of at least 14 to 25 nucleotides. See M. Kanehisa, Nucleic Acids Res. 12:203 (1984).

[0042] The "primers" used in this article can be natural or synthetic oligonucleotides that can act as initiation sites for nucleic acid synthesis after forming a double helix with a polynucleotide template, and can extend along the template from its 3' end to form an extended double helix. The sequence of nucleotides added during the extension process is determined by the sequence of the template polynucleotide. Primers are typically amplified using polymerases, such as DNA polymerase.

[0043] The “substantially non-fluorescent” portion refers to a portion that emits approximately or substantially no detectable fluorescence. For example, at approximately the same concentration of the fluorescent and substantially non-fluorescent portions, the ratio of detectable absolute fluorescence emission from the fluorescent portion to detectable absolute fluorescence emission from the substantially non-fluorescent portion is typically greater than or equal to about 500:1, more typically greater than or equal to about 1000:1, and even more typically greater than or equal to about 1500:1 (e.g., about 2000:1, about 2500:1, about 3000:1, about 3500:1, about 4000:1, about 4500:1, about 5000:1, about 10...). 4 1. Approximately 10 5 1. Approximately 10 6 1. Approximately 10 7 :1 or approximately 10 8 :1).

[0044] "Sequencing," such as nucleotide sequencing methods, involves the determination of information related to the nucleotide base sequence of nucleic acids. This information can include the confirmation or determination of partial and complete sequence information of the nucleic acid. Sequence information can be determined with varying degrees of statistical reliability or confidence. On one hand, the term includes determining the identity and order of multiple adjacent nucleotides in a nucleic acid. "High-throughput sequencing" or "next-generation sequencing" includes sequencing using methods that determine many (typically thousands to billions) nucleic acid sequences in an inherently parallel manner, i.e., where the DNA template is prepared for batch processing rather than sequencing one at a time, and preferably for parallel readout of many sequences, or using ultra-high-throughput serial processing that can be parallelized itself. Such methods include, but are not limited to, pyrosequencing (e.g., as commercialized by 454 Life Sciences, Inc., Branford, CT); ligation sequencing (e.g., as in SOLiD); and sequencing by linker (e.g., as in SOLiD). TM Technology, Life Technologies, Inc., Carlsbad, CA (commercialized); sequencing through the use of modified nucleotide synthesis (such as in TruSeq) TM and HiSeq TM The technology is commercialized by Illumina, Inc., San Diego, CA; in HeliScope TM Commercialized in China by Helicos Biosciences Corporation, Cambridge, MA; and in PacBio RS by Pacific Biosciences of California, Inc., Menlo Park, CA, using ion detection technologies such as Ion Torrent TM Sequencing technologies include: sequencing of DNA nanospheres (Complete Genomics, Inc., Mountain View, CA); sequencing technologies based on nanopores (e.g., developed by Oxford Nanopore Technologies, LTD, Oxford, UK); and highly parallelized sequencing methods, for example.

[0045] In any of the embodiments disclosed herein, the method for obtaining sequence information of the target polynucleotide can be performed in a multiplexing assay. As used herein, “multiplexing” or “multiplex assay” can refer to an assay or other analytical method in which the presence and / or amount of multiple targets (e.g., multiple nucleic acid sequences) can be determined simultaneously, wherein each target has at least one distinct detection property, such as fluorescence properties (e.g., excitation wavelength, emission wavelength, emission intensity, FWHM (full width at half maximum peak) or fluorescence lifetime) or unique nucleic acid or protein sequence properties.

[0046] In any of the embodiments disclosed herein, the sequencing reaction of the target polynucleotide can be performed on an array, such as a microchip. The array may include multiple reaction volumes, for example, created by multiple reaction chambers disposed on the array. The target nucleotide sequence or fragment thereof may be immobilized or immobilized in the reaction volume, such as by adsorption or specific binding to a trap molecule on a solid support in each reaction volume. After the reaction solution is provided in the reaction mixture and delivered to each reaction volume, each reaction volume may be sealed and / or separated from other reaction volumes on the array. Signals such as fluorescence information can then be detected and / or recorded by each reaction volume.

[0047] In any of the embodiments disclosed herein, the array may be addressable. On one hand, addressability includes the ability of the microchip to guide substances such as nucleic acids and enzymes, as well as other amplification components, from one location on the microchip to another (capture sites on the chip). On the other hand, addressability includes the ability to spatially encode sequencing reactions and / or their sequencing products on each arrayspot, such that after sequence readout, the sequencing reactions and / or their sequencing products can be mapped back to the specific arrayspot and associated with additional identification information from that specific arrayspot. For example, spatially encoded tags may be conjugated to target polynucleotides such that when the conjugated target polynucleotides are sequenced, the tagged sequence reveals the location of the array target.

[0048] In some respects, this article provides a fluorescent nucleotide polymerase reaction. On one hand, the fluorescent nucleotide polymerase reaction uses a fluorescent nucleotide, a nucleic acid polymerase (DNA polymerase), a phosphatase, and a nucleic acid substrate. In some embodiments, firstly, the DNA polymerase polymerizes the fluorescent nucleotide into the nucleic acid substrate to release a phosphorylated fluorophore, which is then further hydrolyzed by the phosphatase to remove the phosphate and release a fluorophore with a altered fluorescence state.

[0049] In some respects, this article provides fluorescence-generating sequencing methods. One aspect of these methods is the use of fluorescence-generating nucleotide polymerase reactions to test changes in fluorescence (intensity and spectrum) of the fluorophore, thereby obtaining information about the polymerase reaction. In some respects, this article provides fluorescence-generating sequencing reaction solutions that may contain fluorescence-generating nucleotides, nucleic acid polymerase (DNA polymerase), and phosphatase.

[0050] As described herein, "fluorescent nucleotide" may comprise one or more fluorescent nucleotides. As described herein, "nucleotide" may comprise one or more nucleotides. In some embodiments, multiple nucleotides may be labeled with the same or different fluorescent substrates. In some aspects, this document provides a set of fluorescent sequencing reaction solutions that may contain two or more fluorescent sequencing reaction solutions, such as specific concentrations of A, C, G, and T reaction solutions, or specific concentrations of AC and GT reaction solutions.

[0051] In some aspects, this document provides a fluorescence sequencing reaction cycle, which may include performing a single fluorescence-generating polymerase reaction and testing the fluorescence signal using a sequencing reaction solution. In some aspects, this document provides a round of fluorescence-generating sequencing reactions, which may include using members of a set of fluorescence-generating sequencing reaction solutions in a predetermined order for sequencing reaction cycles. In some aspects, this document provides a set of fluorescence-generating sequencing reactions, which may include one or more rounds of fluorescence-generating sequencing.

[0052] In some respects, this paper provides single-base-resolved sequencing reactions. One approach is a (2+2 single-color two-set) reaction, where the first reaction solution is a mixture of two bases (e.g., AC), and the second reaction solution is a mixture of two other bases (GT), with the two solutions used alternately for sequencing. In this case, the number of extended bases increases in each cycle. After N rounds of sequencing, the number of extended bases is 2N nt, carrying 2N bits of information. There are three combinations that complete the above sequencing: AC / GT, AG / CT, and AT / CG; or, according to standard degenerate base (degenerate nucleotide) notation, written as M / K, R / Y, and W / S. These three combinations can be sequenced separately, or a new set of sequencing can be completed before re-sequencing. The i-th base determined on the DNA sequence will always pair and release a signal in a unique cycle of either of the two sequencing sets. In each sequencing set, the determined base sampling injection cycles include two types, so there are a total of 2×2=4 possible cases, corresponding to exactly four bases. The order of sequencing combinations does not affect base inference.

[0053] Table 1

[0054]

[0055] Table 2

[0056]

[0057] Table 3

[0058]

[0059] In a further implementation, the method also includes sequencing using a third, different combination of reaction solutions after completing two different sequencing runs. The i-th base determined on the DNA sequence must pair and release a signal in a unique cycle among the three sequencing runs. In each sequencing run, the determined base sampling injection cycle includes two types, resulting in 2 × 2 × 2 = 8 possible cases, only four of which are valid, and the other four are invalid. Insertion or deletion errors are likely to occur in fluorescence switching sequencing. If a sequencing error occurs in one of the three sequencing runs for a specific base, the sequence cannot be correctly deduced, and it can be concluded that one or more of the three sequencing runs contain a sequencing error at that point.

[0060] Table 4

[0061]

[0062] This type of error can be corrected because when sequencing errors in a single dataset are corrected, a large number of subsequent errors will also be corrected.

[0063] Another specific implementation is a 2+2 two-color two-round mode. The first reaction solution is made from a mixture of two bases and carries different fluorescent labels (such as AX / CY), and the second reaction solution is made from a mixture of two other bases (GX / TY). In this case, the number of bases extended per cycle increases, averaging 2nt. The information carried is 2N bits.

[0064] II. Methods for detecting and / or correcting sequencing errors

[0065] On the one hand, this article relates to methods for detecting and / or correcting errors in one or more sequence data in sequencing results, which falls under the field of nucleic acid sequencing.

[0066] On the one hand, this paper provides a method for detecting and / or correcting sequence data errors in sequencing results. On the other hand, the sequencing reaction solution contains at least two types of nucleotide substrate molecules with different bases. On the other hand, degenerate gene coding information can be obtained. By comparing two or more degenerate coding information, it is possible to determine whether conflicting sequence information is present in one or more nucleotide residues. Using this method to correct sequence information, even small improvements that reduce the sequencing error rate in the original sequencing data can lead to a more significant reduction in the error rate of the corrected sequence information.

[0067] On the one hand, this paper discloses a method for detecting and / or correcting sequence data errors in sequencing results. On the one hand, the method includes sequencing a nucleic acid sequence to obtain sequence data of three or more orthogonal nucleotide degenerate sequences. On the other hand, the method further includes detecting errors in the sequence by comparing the three or more orthogonal nucleotide degenerate sequences. On the other hand, at the location where the error is found, a corrected sequence is obtained by modifying at least one sequence.

[0068] This document also discloses methods for detecting and / or correcting sequence data errors in sequencing results, wherein the methods include performing a sequencing reaction on nucleotide sequences to obtain three or more degenerate sequences represented by the letters M, K, R, Y, W, S, B, D, H, and V. On the one hand, according to the IUPAC nucleic acid symbols, the degenerate bases in this invention are represented by the letters in Table 5. For example, M represents A and / or C bases.

[0069] Table 5: Letters representing degenerate bases

[0070] letter The bases represented M AC K GT R AG Y CT W AT S CG B CGT D AGT H ACT V ACG

[0071] In any of the foregoing embodiments, sequence errors can be detected by comparing three or more degenerate sequences. In any of the foregoing embodiments, the erroneous nucleotide position is identified during comparison, and a corrected sequence can be obtained by modifying at least one sequence; in any of the foregoing embodiments, the erroneous position identified during comparison can be the actual location where the sequencing error occurred.

[0072] On the other hand, this document discloses methods for detecting and / or correcting sequence data errors in sequencing results, wherein the methods include sequencing the same nucleic acid sequence to obtain two or more degenerate sequences represented by the letters M, K, R, Y, W, S, B, D, H, and V, to obtain sequence information represented by nucleic acid residues A, G, T, and C or nucleic acid residues A, G, U, and C. On another aspect, the methods also include detecting sequence errors by using light or electrical signals generated by one or more functional groups coupled to different bases in the sequencing reaction. For example, light or electrical signals from different fluorescent groups coupled to different bases in the sequencing reaction can be used as “redundant” information that distinguishes one base from another at a specific position in the sequence. In any of the foregoing embodiments, if an erroneous nucleotide position is found during comparison, a corrected sequence can be obtained by modifying at least one sequence; in any of the foregoing embodiments, the position of the error identified during comparison can be the actual location where the sequencing error occurred.

[0073] On the other hand, this paper discloses a method for detecting and / or correcting sequence errors in sequencing results using the memory property of nucleic acid sequences. One aspect of the method involves sequencing the same nucleic acid sequence to obtain data of three or more orthogonal degenerate nucleic acid sequences. Another aspect of the method includes comprehensively comparing degenerate sequences and utilizing the memory property of nucleic acid sequences to detect sequence errors. One aspect is that, at the location where an error occurs, a corrected sequence can be obtained by modifying at least one sequence. In some embodiments, each degenerate sequence represents only a portion of the sequence information of the actual polynucleotide template, and nucleotide identity at a position in one degenerate sequence does not indicate, or does not necessarily indicate, nucleotide identity at the same position in another degenerate sequence.

[0074] On the one hand, this document discloses a method for detecting and / or correcting sequence data errors in sequencing results. This method includes immobilizing a nucleic acid fragment to be sequenced onto a vector and providing a reaction solution to initiate a sequencing reaction, from which a degenerate nucleic acid sequence is obtained. The sequencing reaction can be repeated multiple times, such that a degenerate nucleic acid sequence is obtained from each sequencing round. After N sequencing rounds, N degenerate nucleic acid sequences are obtained. On the other hand, by comprehensively comparing the N degenerate sequences, the location of the sequence error can be detected. Furthermore, the method may also include, by comparing the location of the error, obtaining a corrected sequence by modifying at least one sequence. In any of the foregoing embodiments, the reaction solution may contain two or more types of nucleotide substrate molecules with different bases. In any of the foregoing embodiments, N may be a positive integer equal to or greater than 2.

[0075] In any of the foregoing embodiments, the method may include comparing N-1 degenerate nucleic acid sequences out of N to obtain nucleic acid sequence information encoded by A, G, T, and C, or nucleic acid sequence information encoded by A, G, U, and C. In one aspect, the method further includes comparing N degenerate nucleic acid sequences. In any of the foregoing embodiments, N may be a positive integer equal to or greater than 3.

[0076] In any of the foregoing embodiments, the method may include comparing N degenerate nucleic acid sequences to obtain nucleic acid sequence information encoded by A, G, T, and C, or nucleic acid sequence information encoded by A, G, U, and C. On one hand, the method further includes detecting the location of an error by using optical and / or electromagnetic information provided by two or more functional groups coupled to nucleotide residues. In any of the foregoing embodiments, N may be a positive integer equal to or greater than 2.

[0077] On the other hand, this document discloses a method for detecting and / or correcting sequence data errors in sequencing results, wherein the method includes immobilizing the nucleic acid fragment to be tested onto a vector. The method further includes providing a reaction solution to initiate a sequencing reaction, wherein the reaction solution contains nucleotide substrate molecules for sequencing and is divided into three groups according to different bases, each group comprising two different reaction solutions, and each sample of the reaction solution containing nucleotide substrate molecules with different bases. There is no intersection between the bases of the nucleotides in the two samples of the same reaction solution. Each round of sequencing uses one set of reaction solutions, providing two samples of each set to react sequentially with the nucleic acid template in any suitable order. Three rounds of sequencing are performed using the three sets of reaction solutions to obtain three degenerate sequences. The location of the sequence error can be detected by comprehensively comparing the three degenerate sequences. In one embodiment, by modifying at least one sequence at the location of the error, a corrected sequence can be obtained.

[0078] In any of the foregoing embodiments, the sequencing reaction can be carried out using a nucleotide substrate molecule modified with a fluorophore having fluorescence switching properties (such as dNTPs or ddNTPs), wherein the modification is on the 5'-terminal polyphosphate group of the nucleotide substrate molecule. On one hand, fluorescence switching properties can refer to a significant change in the fluorescence signal after sequencing compared to before the sequencing reaction. On the other hand, fluorescence switching occurs after the nucleotide substrate is polymerase-catalyzed and incorporated into an extension primer. On one hand, the nucleotide sequence fragment to be tested is immobilized on a vector, and then a reaction solution containing a nucleotide substrate molecule is provided to react with the template nucleotide sequence fragment. On the other hand, an enzyme is then used to release a fluorophore from the nucleotide substrate incorporated into the extension primer (and the duplex polymerase extension product), resulting in fluorescence switching.

[0079] On the one hand, after each sequencing reaction, the fluorescence signal may be significantly enhanced or weakened compared to before the sequencing reaction, or the frequency of emitted light may change significantly.

[0080] In any of the foregoing embodiments, sequence errors may include insertions and / or deletions. In any of the foregoing embodiments, a sequence data error can be considered to have occurred at a position when at least two degenerate nucleic acid sequences do not share a common base.

[0081] In any of the foregoing embodiments, correcting sequence errors may include correcting at least one nucleotide residue in a sequence such that the corrected sequence has the correct nucleotide residue at at least one position following the corrected nucleotide residue. On the one hand, the nucleotide residue is correct if the nucleic acid sequence information of any two rounds of sequencing determined at the same nucleotide residue position is not inconsistent with the nucleic acid sequence information of another round of sequencing.

[0082] In any of the foregoing embodiments, correcting sequence errors may include correcting errors in at least one sequence such that common nucleotide residues at at least one position in the sequence can be obtained by comparing sequence information from multiple rounds of sequencing.

[0083] In any of the foregoing embodiments, correcting sequence errors may include extending (e.g., by inserting nucleic acid residues at the location where an error is believed to have occurred) and / or shortening (e.g., by deleting nucleic acid residues at the location where an error is believed to have occurred) sequences representing nucleic acid sequence information from multiple rounds of sequencing. On one hand, by extending and / or shortening at least one sequence from multiple rounds of sequencing, the corrected sequence will be consistent with sequences from other rounds at at least one nucleotide residue location.

[0084] In any of the foregoing embodiments, the memory of a nucleic acid sequence can refer to the fact that, in the sequencing results, the nucleic acid sequence information at a specific position involves not only the nucleotide residues in the corresponding nucleic acid in the template, but also the sequence information preceding that sequence information.

[0085] In any of the foregoing embodiments, using sequencing signals from the other two rounds of sequencing, the sequence in the sequencing signal can be extended (e.g., by inserting nucleic acid residues at locations where errors are believed to have occurred) by a certain length to obtain a corrected nucleic acid sequence. In any of the foregoing embodiments, using sequencing signals from the other two rounds of sequencing, the sequence in the sequencing signal can be shortened (e.g., by deleting nucleic acid residues at locations where errors are believed to have occurred) by a certain length to obtain a corrected nucleic acid sequence.

[0086] In any of the foregoing embodiments, the reaction solution can be divided into three groups based on different bases, wherein the bases include A, G, C, and T bases or A, G, C, and U bases. In any of the foregoing embodiments, the bases can be methylated, hydroxymethylated, or modified with aldehyde or carboxyl groups, or be unmethylated, unhydroxymethylated, or not modified with aldehyde or carboxyl groups.

[0087] In any of the foregoing embodiments, the nucleotide substrate reaction solution may contain different bases and may be divided into two reaction solutions according to the different bases, for example, one reaction solution contains A+G and the other contains C+T; one reaction solution contains A+C and the other contains G+T; or one reaction solution contains A+T and the other contains C+G.

[0088] In any of the foregoing embodiments, the reaction solution may include multiple reaction solutions, one of which may be used for the sequencing reaction. On one hand, one or more reaction solutions are used per round of sequencing. On the other hand, at least one reaction solution contains two or more types of nucleotide substrate molecules with different bases. In any of the foregoing embodiments, the reaction solutions used in different rounds of sequencing contain different combinations of nucleotide substrate molecules.

[0089] In any of the foregoing embodiments, the nucleotide substrate molecule can be labeled with fluorescence. On one hand, a fluorescent group (or a functional group with fluorescence-switching properties acquired through a chemical reaction) is coupled to a base of the nucleotide residue. On the other hand, one of the fluorophores or functional groups can be used to modify the nucleotide substrate molecule, or multiple fluorophores or functional groups can be used to modify the nucleotide substrate molecule with different bases.

[0090] With the deepening understanding of genes in recent years, gene sequencing has brought about tremendous changes in pharmacy and biology. Conventional sequencing methods include Sanger DNA sequencing, restriction fragment length polymorphism (RFLP), single-strand conformation polymorphism (SCM), and allele-specific oligonucleotide hybridization sequencing based on gene chips. Due to various influencing factors during the sequencing process, such as inaccurate CCD luminescence, fluid movement, ambient light, contaminating DNA, signal correction system errors, or impurities in the sequencing reaction solution, errors are inevitable in the sequencing results. As genetic material, DNA stores the genetic information of an organism, a characteristic that makes it suitable as a storage medium for basic information. When DNA is used to store information, this information needs to be encoded into the DNA sequence, and then the information is read using gene sequencing methods. To avoid encoding and / or reading errors, redundant information is often introduced into the encoding process and used for signal correction during reading. For example, George Church et al., “Next-Generation Digital Information Storage in DNA,” Science, 2012, used Reed Solomon codes to encode information into DNA sequences and used the Illumina sequencing platform to read the information from the DNA sequences. DNA encoding-reading techniques are also used in combinatorial chemistry and other fields. In previous DNA encoding techniques, the type of each base was typically independent of bases at other positions (memoryless encoding) or only related to its neighboring bases. This paper presents a memory-based, distributed, orthogonal DNA encoding method where the type of each base is related to all bases in the preceding positions. Furthermore, the method can effectively improve the encoding-reading process and ultimately the decoding accuracy by comprehensively comparing multiple sets of orthogonal codes.

[0091] On one hand, the present invention provides a method for detecting and / or correcting coding errors in sequencing results, wherein the method includes sequencing the same nucleic acid sequence to obtain three or more orthogonal nucleotide degenerate sequence data, wherein errors in the sequence can be detected by comparing the three or more orthogonal nucleic acid degenerate sequences, and wherein a corrected sequence can be obtained by modifying at least one sequence at the location where the error was found during the comparison.

[0092] On the one hand, this invention provides a method for detecting and / or correcting code errors in sequencing results. The method involves sequencing the same nucleic acid sequence to obtain three or more degenerate sequence data represented by the letters M, K, R, Y, W, S, B, D, H, and V. Errors in the sequence can be detected by comparing the three or more degenerate sequences, and a corrected sequence can be obtained by modifying at least one sequence at the location where the error was found during the comparison. On the one hand, the method is applicable to routine sequencing. On the other hand, with proper sequencing substrate design, three or more coding results can be obtained through multiple rounds of sequencing, where redundancy of information can be used to detect and / or correct error codes.

[0093] On one hand, the present invention provides a method for detecting and / or correcting code errors using the memory of genetic code, wherein the method includes sequencing the same nucleic acid sequence to obtain two or more degenerate sequences represented by the letters M, K, R, Y, W, S, B, D, H, V, or obtaining nucleic acid sequence information encoded by A, G, T, C, or nucleic acid sequence information encoded by A, G, U, C, wherein the light or electrical signals caused by different functional groups linked on different bases in the sequencing reaction are used as redundant information to detect sequence errors, wherein a corrected sequence can be obtained by modifying at least one sequence at the location where the error is found during comparison.

[0094] On one hand, the present invention provides a method for detecting and / or correcting code errors using the memory of the gene code, wherein the method includes sequencing the same nucleic acid sequence to obtain three or more orthogonal degenerate nucleotide sequence data, and comprehensively comparing the degenerate sequences and using the memory of the nucleic acid sequences to detect sequence errors, wherein a corrected sequence can be obtained by modifying at least one sequence at the location where the error is found during the comparison, wherein in the degenerate sequences, each sequence signal represents a portion of gene sequence information, and the signal at the same position on another degenerate sequence cannot be inferred from the signal on one such degenerate sequence.

[0095] In any of the foregoing embodiments, the method may include immobilizing the nucleic acid fragment to be tested on a vector, providing a reaction solution to initiate a sequencing reaction, such that each round of sequencing yields a degenerate nucleic acid sequence; after at least N rounds of sequencing, N degenerate nucleic acid sequences are obtained, wherein by comprehensively comparing the N degenerate sequences, the location of an error in the sequence can be detected, wherein by modifying at least one sequence at the location of the error found during comparison, a corrected sequence can be obtained, wherein the reaction solution may contain two or more types of nucleotide substrate molecules with different bases, and wherein N is a positive integer equal to or greater than 2.

[0096] On the one hand, by comparing N-1 degenerate nucleic acid sequences, nucleic acid sequence information encoded by A, G, T, C, or nucleic acid sequence information encoded by A, G, U, C, can be obtained. On the other hand, by comparing N degenerate sequences, the location of sequence errors can be detected. N can be a positive integer equal to or greater than 3.

[0097] On the one hand, by comparing N degenerate nucleic acid sequences, nucleic acid sequence information encoded by A, G, T, and C, or nucleic acid sequence information encoded by A, G, U, and C, can be obtained. Furthermore, by comparing N degenerate sequences, the location of sequence errors can be detected. On the other hand, the location of errors can be detected using luminescent information provided by two or more functional groups linked to the bases, where N is a positive integer equal to or greater than 2. On the other hand, the method includes using changes in the bases themselves as redundant information for correction during the sequencing reaction that generates information from molecules such as phosphate and hydrogen ions released during the reaction.

[0098] On one hand, the present invention provides a method for detecting and / or correcting code errors in sequencing results, wherein the method includes immobilizing the nucleic acid fragment to be tested, providing a reaction solution to initiate a sequencing reaction, wherein the reaction solution for the nucleotide substrate molecules used for sequencing is divided into three groups according to different bases, each group comprising two different reaction solutions, each reaction solution containing nucleotide substrate molecules with different bases. On one hand, there is no overlap between the bases of the nucleotides in the two reaction solutions. On the other hand, one set of reaction solutions is used for each round of sequencing, with the two reaction solutions in each group provided alternately. On the other hand, the method includes performing three rounds of sequencing using the three sets of reaction solutions to obtain three degenerate sequences, the location of the error can be detected by comprehensively comparing the three degenerate sequences, and a corrected sequence can be obtained by modifying at least one sequence at the location of the error during comparison.

[0099] On the one hand, the reaction solution containing two different bases can be divided into two reaction solutions; other steps of the method can be adjusted accordingly.

[0100] On the one hand, the reaction solution may contain multiple reaction solutions, one for each sequencing run, wherein one or more reaction solutions are used per sequencing run, wherein at least one reaction solution contains two or more types of nucleotide substrate molecules with different bases, and wherein the reaction solutions used for different sequencing runs contain different combinations of nucleotide substrate molecules.

[0101] On one hand, the sequencing of the present invention includes sequencing by using a nucleotide substrate molecule with a fluorophore modified with a 5'-terminal polyphosphate to have fluorescence switching properties, wherein fluorescence switching properties refer to a significant change in the fluorescence signal after sequencing compared to the state before the sequencing reaction. In this process, the nucleotide sequence fragment to be tested is first immobilized on a vector, then a reaction solution containing nucleotide substrate molecules is provided, and then an enzyme is used to release the fluorophore on the nucleotide substrate, thereby causing fluorescence switching.

[0102] On the one hand, "significant changes in fluorescence signal after sequencing compared to before sequencing reaction" means that after each sequencing reaction, the fluorescence signal is significantly enhanced or significantly weakened compared to before the sequencing reaction, or the emission light frequency range is significantly changed.

[0103] On the one hand, sequence errors refer to insertion or deletion errors. On the other hand, sequence data errors refer to the situation where at least two nucleic acid sequence pieces do not represent the same base at the same position, which is considered an error. In yet another aspect, the method includes correcting errors in at least one sequence such that subsequent sequences at at least one position are correct, wherein a correct sequence means that the nucleic acid sequence information determined by any two rounds of sequences at the same position does not contradict the nucleic acid sequence information of another round of sequences, or in other words, the nucleic acid sequence information represented by any two rounds of sequences at the same position does not contradict the luminescence information provided by the functional group linked to the base or information from another sequencing process.

[0104] On one hand, the method includes correcting sequences by correcting errors in at least one sequence such that common bases can be obtained by comprehensively comparing sequences at at least one position.

[0105] On the one hand, by modifying at least one sequence, a corrected sequence can be obtained at the location where the error occurs by extending or shortening the sequence representing nucleic acid sequence information. Extension or shortening refers to increasing or decreasing the length of the same detection sequence. When encoding causes shortening or extension at that position, the sequence information represented by the code remains unchanged, resulting in the same code. For example, when the signal intensity of the degenerate code M is 2 (MM), it can be extended to 3 (MMM).

[0106] On the one hand, the memory property of nucleic acid sequences means that the nucleic acid sequence information at a certain position in the sequencing results is related not only to the sequence on the corresponding nucleic acid to be tested, but also to the sequence information preceding it.

[0107] On the one hand, by extending or shortening some sequencing signals at a certain position, the gene sequence represented by that position is extended or shortened in order to obtain a corrected nucleic acid sequence using two other rounds of sequencing signals. Extending the sequencing signals includes adding or inserting a specific length of the gene sequence represented by that position, while shortening some sequencing signals includes shortening or deleting a specific length of the gene sequence represented by that position, and obtaining a corrected nucleic acid sequence using two other rounds of sequencing signals.

[0108] On the one hand, the reaction solution is divided into three groups according to the different bases, where the base refers to A, G, C, T bases or A, G, C, U bases, and the base can be methylated, hydroxymethylated, or have an aldehyde or carboxyl group, or it can be unmethylated, unhydroxymethylated, or have no aldehyde or carboxyl group.

[0109] On the one hand, the reaction solution of nucleotide substrate containing two different bases can be divided into two reaction solutions according to the difference in bases.

[0110] On one hand, nucleotide substrate molecules can be labeled with fluorescence. On the other hand, the method includes modifying a fluorophore or functional group with fluorescence switching by a chemical reaction on the bases of the nucleotide substrate molecule. On the other hand, one of the fluorophores or functional groups can be used to modify the nucleotide substrate molecule, or multiple fluorophores or functional groups can be used to modify the nucleotide substrate molecule with different bases.

[0111] On the one hand, each round of sequencing yields a set of degenerate gene sequence information. On the other hand, degenerate gene sequence information refers to information containing possible gene sequences. For example, when the reaction solution contains nucleotide substrate molecules with A and G bases, the degenerate gene sequence information obtained from sequencing includes the gene sequence information of C and / or T bases in the nucleotide sequence to be tested. When the reaction solution contains nucleotide substrate molecules with T and G bases, the gene sequence information obtained by sequencing includes the gene sequence information of C and / or A bases in the nucleotide sequence to be tested.

[0112] On the one hand, in the comprehensive comparison of information from three rounds of sequencing, if the gene sequence information represented by the signal from one round of sequencing is an oversized erroneous sequencing signal, the gene sequence information represented by that sequence signal can be shortened so that the comparison result of at least one subsequent sequencing signal is correct.

[0113] On the one hand, in the comprehensive comparison of information from three rounds of sequencing, if the gene sequence information represented by the signal from one round of sequencing is a small erroneous sequence signal, then a gap can be added or the sequence can be extended to correct the comparison result of at least one subsequent sequencing signal. For example, when the signal intensity of degenerate code M is 2, i.e., MM, it can be extended to 3, i.e., MMM.

[0114] On the one hand, this paper provides methods for detecting and / or correcting errors in gene sequencing results, particularly sequencing methods using one or more reaction solutions containing nucleotide substrate molecules with two or more bases. In a specific aspect, this method is applicable to SBS (sequencing by synthesis) methods used for sequencing.

[0115] On the one hand, the degenerate gene sequence information in this paper includes possible gene sequence information for a given target (or template) sequence. For example, when the reaction solution contains nucleotide substrate molecules with A and G bases, the degenerate gene sequence information obtained by sequencing includes the gene sequence information of C and / or T bases in the nucleotide sequence to be tested. Assuming that the intensity information obtained from the sequencing reaction is 3, it means that the gene to be tested may contain three Cs and / or Ts, such as three Cs or three Ts, or one C and two Ts, or one T and two Cs, and the exact relative positions of Ts and / or Cs cannot be distinguished based on the degenerate sequence. Degenerate gene sequence information and degenerate codes are terms commonly used in this art.

[0116] On the one hand, the method described in this paper can detect and / or correct errors in sequencing, but it cannot completely eliminate sequence errors. It is possible that a specific location in the sequence signal that has been modified is not the actual location of a sequencing error, but the probability is extremely low. Ultimately, accuracy can be further improved. For example, if the modified signals of MK, RY, and WS are combined, and two modifications are found in N consecutive signals, it is considered highly likely that an error has occurred, and the corresponding sequence should be discarded. Here, N can be a positive integer equal to or greater than 2. The larger the value of N, the higher the probability that the sequence should be discarded, and the higher the final decoding ratio. In this paper, the optimized value of N is 3.

[0117] DNA sequences are copolymers; for example, DNA regions may include two different deoxyribonucleotides, such as AAC and GGTG.

[0118] On the one hand, methods for detecting and / or correcting sequence data errors can detect the location of errors and / or correct sequence errors.

[0119] On the one hand, in the actual sequencing process, the method includes first obtaining the relative intensity values ​​of optical or other signals through a cyclic sequencing reaction. These intensity values ​​can be represented in a specific form. For example, M represents positional information and the number of bases at that position (multiple bases are acceptable), or it can represent the degenerate gene coding result. By decoding the relative intensity values ​​of a sufficient amount of information, the gene sequence information to be tested can be obtained.

[0120] On one hand, delivering or providing reagents or reaction solutions means adding reagents or reaction solutions, such as the reaction mixture for a sequencing reaction, into a container. On the other hand, three or more rounds of sequencing can be used. Alternatively, two or more rounds of sequencing can be used. On the other hand, sequencing signals are counted by number of times. The signal intensity information for each sequencing run can be recorded, and in some embodiments, the intensity information is perfectly identical to the length of the corresponding copolymer.

[0121] Sequencing signals can be counted either by level or by the number of times a specific nucleotide is detected. For example, if the signal intensity is n and the number of nucleotides added to the reaction solution is X, then the sequencing result is represented as XXX...X, where the sequence length is n nucleotides. Figure 1 When the sequencing signal is counted by number, it can be converted into a level-counted sequencing signal MMMKKKKKMKKKMMK or written as (A / C, A / C, A / C, G / T, G / T, G / T, G / T, A / C, G / T, G / T, A / C, A / C and G / T).

[0122] For example, sequencing reaction solutions containing dA4P and dC4P (nucleotides with four phosphate groups and terminal phosphates labeled with fluorescent groups) can be used in odd-numbered cycles, while sequencing reaction solutions containing dG4P and dT4P can be used in even-numbered cycles. A set of fluorescence signal values ​​after multiple reactions can be found in Table 6 below.

[0123] Combinations of other fluorescently labeled nucleotides can be used to obtain fluorescence signal values ​​associated with the target DNA sequence. Examples of possible combinations are shown below:

[0124] M / K pattern: dA4P and dC4P are presented an odd number of times, and dG4P and dT4P are presented an even number of times; or vice versa.

[0125] R / Y mode: dA4P and dG4P are presented an odd number of times, and dC4P and dT4P are presented an even number of times; or vice versa; and

[0126] W / S mode: dA4P and dT4P are presented an odd number of times, and dC4P and dG4P are presented an even number of times; or vice versa.

[0127] Table 6

[0128]

[0129] Sequencing data obtained from three different nucleotide combinations can be combined as a horizontally counted signal. For each position, the next step is to analyze the intersection of the nucleotide types represented by the three horizontally counted sequencing signals at that position to obtain the target DNA sequence. This is, in part, the basic principle of signal decoding. For example, if the horizontally counted sequencing signals correspond to combinations of M / K, R / Y, and W / S such as (3, 5, 1, 3, 2, 1), (2, 4, 3, 2, 1, 3), and (2, 1, 3, 2, 3, 3, 1), the sequence can be summarized as AACTTTGGATTGCCT (SEQ ID NO:1).

[0130] On the one hand, the comprehensive comparison of the results of the three rounds of sequencing involves converting chemiluminescent signals or other forms of intensity signals into gene sequence information, and then comparing the results of the three rounds of sequencing at the same base position. If the results obtained from the three rounds of sequencing are consistent, the sequencing at that position is considered correct; if the gene sequence information represented by the results obtained from the three rounds of sequencing is inconsistent, the sequencing result at that base position is considered incorrect.

[0131] On the one hand, if factors such as inaccurate CCD luminescence, fluid movement, ambient light, contaminating DNA, errors in the signal correction system, or impurities in the sequencing reaction solution cause the sequence signal at a specific time point to be larger or smaller when counted in terms of number of iterations, this will result in an empty intersection of the nucleotide types represented by the corresponding or subsequent positions in the sequencing signal when counted horizontally, making it impossible to resolve the nucleotide types. Clearly, errors in the number-counted sequencing signal can cause the overall level-counted sequencing signal to shift from the position where the error occurred. Therefore, the level-counted sequencing signal is a signal with memory. Based on the memory characteristic of the level-counted sequencing signal, errors in the sequencing signal can be corrected.

[0132] On the one hand, this invention provides a method for detecting and / or correcting sequence data errors in sequencing results. The sequencing reaction solution contains at least two types of nucleotide substrate molecules with different bases; degenerate gene coding information can be obtained. Those skilled in the art can determine whether a conflict occurs in the code at a certain position by comparing two or more degenerate coding information. Compared to the same test substrate, using different primers or directly testing multiple rounds is easier, and the test can be completed with a single test design. On the other hand, the method provided herein is completely different from the method of testing the same gene in multiple rounds. In some aspects, the method provided herein lacks a correction basis if there are only two mutually orthogonal degenerate gene coding results (excluding cases where redundant information such as color is added). On the other hand, this invention first assumes that the detection and correction of errors in three or more mutually orthogonal degenerate codings leads to this sequencing type.

[0133] On the one hand, this article provides a method for detecting and / or correcting sequence data errors in sequencing results. Specifically, sequencing is performed on nucleotide substrate molecules with 5' polyphosphate-modified fluorophores exhibiting fluorescence switching properties; this method is also known as fluorescence switching sequencing. When using fluorescence switching sequencing in conjunction with a 2+2 sequencing method, the sequencing method itself offers many advantages, such as long readouts of 300 bp and sequencing accuracy up to 99.99%; all of these cannot be achieved by using either the 2+2 sequencing method or fluorescence switching sequencing alone. Furthermore, the combined method offers other advantages, such as higher throughput, simpler reaction, lower error rate, and no need for real-time information acquisition. Similarly, sequencing on other nucleotide substrate molecules with fluorescence switching properties exhibits the same characteristics. For example, fluorescence switching sequencing and 2+2 sequencing methods provide redundant information (luminescence or other detectable information) in addition to color information during three rounds of sequencing, which can be used for correction. This can also extend the effective reads without changing accuracy. The correction result depends on the accuracy of the sequencing method, and it can significantly improve the overall accuracy of the effective reads while keeping the sequencer accuracy constant. For example, the accuracy of sequencing a 400 bp nucleic acid fragment is up to 97.36%. The corrected accuracy reaches 99.17%. Therefore, if a sequencer employs this error detection and correction method, the effective reads can be extended accordingly. When using the methods presented in this paper for correction, a clear rule emerges: any small improvement in the sequencing method that can reduce the error rate can significantly reduce the error rate of the modified coding data.

[0134] Example

[0135] Example 1: Sequencing using the "2+2 monochromatic" method

[0136] To further describe this disclosure, specific embodiments are provided below. Unless otherwise specified, specific parameters, steps, etc., are conventional in the art. These specific embodiments are not intended to limit the scope of the invention.

[0137] For sequencing using the "2+2 monochromatic" method, three sets of reaction solutions were prepared. Each set consisted of two vials, each containing two bases labeled with the same fluorescent group X. For each set, the two vials contained exactly all four bases required for the sequencing reaction. The six vials (two vials per set) were all unique.

[0138] Table 7: Reaction solution in the "2+2 monochromatic" method

[0139] First bottle Second bottle First set AX+CX GX+TX Second set AX+GX CX+TX The third set AX+TX CX+GX

[0140] The complete sequencing process consists of three rounds of sequencing, which can be performed sequentially in any suitable order. Each round of sequencing uses one of the three sets of reaction solutions listed in Table 7. For example, the order of the three rounds could be set 1 → set 2 → set 3, or set 2 → set 3 → set 1, etc. Each round of sequencing uses the three sets of reaction solutions mentioned above, and all other conditions are exactly the same (e.g., the same sequencing primers and reaction conditions are used in all three rounds). Two bottles from the same set of reaction solutions can also be used in any suitable order; for example, the first bottle can be used before or after the second bottle.

[0141] Each round of sequencing includes:

[0142] 1. Hybridize the sequencing primers onto the prepared DNA array.

[0143] 2. Begin the sequencing reaction. Steps 2.1-2.4 can be repeated multiple times.

[0144] 2.1. Add the first bottle of reaction solution (e.g., the first or second bottle of the first set) to the sequencing reaction mixture (e.g., in a flow cell) to allow the reaction to proceed and collect the fluorescence signal from the fluorescent group X.

[0145] 2.2. Clean all residual reaction solution and fluorescent molecules from the flow cell.

[0146] 2.3. Add a second bottle of reaction solution (e.g., the second bottle or the first bottle of the first set) to the sequencing reaction mixture, allow the reaction to proceed, and collect the fluorescence signal.

[0147] 2.4. Clean all residual reaction solution and fluorescent molecules from the flow cell.

[0148] 3. Unwind the extended sequencing primers.

[0149] At this point, a new round of sequencing reactions can begin.

[0150] The solution used in this example can be prepared as follows. The washing solution for the sequencing reaction solution contains: 20 mM Tris-HCl pH 8.8; 10 mM (NH4)2SO4; 50 mM KCl; 2 mM MgSO4; and 0.1% 20. The master solution for the sequencing reaction contains: 20 mM Tris-HCl pH 8.8; 10 mM (NH4)2SO4; 50 mM KCl; 2 mM MgSO4; 0.1% 8000 units / mL Bst polymerase; and 100 units / mL CIP (alkaline phosphatase, bovine intestine).

[0151] The three sequencing reaction solutions were prepared as follows:

[0152] Set 1 (bottles 1A and 1B):

[0153] Bottle 1A: Main solution + 20 μM dA4P-TG + 20 μM dC4P-TG

[0154] Bottle 1B: Main solution + 20 μM dG4P-TG + 20 μM dT4P-TG

[0155] Set 2 (bottles 2A and 2B):

[0156] Bottle 2A: Main solution + 20 μM dA4P-TG + 20 μM dG4P-TG

[0157] Bottle 2B: Main solution + 20 μM dC4P-TG + 20 μM dT4P-TG

[0158] Set 3 (bottles 3A and 3B):

[0159] Bottle 3A: Main solution + 20 μM dA4P-TG + 20 μM dT4P-TG

[0160] Bottle 3B: Main solution + 20 μM dC4P-TG + 20 μM dG4P-TG

[0161] Place the prepared reaction solution and main solution in a 4°C refrigerator or on ice until needed.

[0162] To prepare the sequencing primers, the sequencing primer solution (10 μM primers in 1×SSC buffer) was injected into the sequencing chip, then heated to 90°C, and then cooled to 40°C at a rate of 5°C / min. The sequencing primer solution was then washed away with washing buffer.

[0163] To perform the sequencing reaction, place the sequencing chip on the sequencer. To perform sequencing using the first set of reaction solutions, follow these steps:

[0164] 1. Add 10 mL of cleaning solution to rinse the chip.

[0165] 2. Cool the chip down to 4°C.

[0166] 3. Add 100 μL of reaction solution 1A.

[0167] 4. Heat the chip to 65°C.

[0168] 5. Wait 1 minute.

[0169] 6. Take fluorescence images at an excitation laser wavelength of 473 nm.

[0170] 7. Add 10 mL of cleaning solution to rinse the chip.

[0171] 8. Cool the chip down to 4°C.

[0172] 9. Add 100 μL of reaction solution 1B.

[0173] 10. Heat the chip to 65°C.

[0174] 11. Wait 1 minute.

[0175] 12. Take fluorescence images at an excitation laser wavelength of 473 nm.

[0176] 13. Repeat steps 1-12 50 times to obtain 100 fluorescence signals.

[0177] The second round of sequencing can be performed as follows. First, cool the chip to room temperature. Then add 200 μL of 0.1 M NaOH solution to denature the extended DNA double strands from the first round of sequencing. Then add 10 ml of washing buffer to wash away residual NaOH and denatured DNA single strands.

[0178] Then, the sequencing primers were rehybridized to the DNA array, as described above. The sequencing reaction using the second set of reaction solutions was performed as follows:

[0179] 1. Add 10 mL of cleaning solution to rinse the chip.

[0180] 2. Cool the chip down to 4°C.

[0181] 3. Add 100 μL of reaction solution 2A.

[0182] 4. Heat the chip to 65°C.

[0183] 5. Wait 1 minute.

[0184] 6. Take fluorescence images at an excitation laser wavelength of 473 nm.

[0185] 7. Add 10 mL of cleaning solution to rinse the chip.

[0186] 8. Cool the chip down to 4°C.

[0187] 9. Add 100 μL of reaction solution 2B.

[0188] 10. Heat the chip to 65°C.

[0189] 11. Wait 1 minute.

[0190] 12. Take fluorescence images at an excitation laser wavelength of 473 nm.

[0191] 13. Repeat steps 1-12 50 times to obtain 100 fluorescence signals.

[0192] The third round of sequencing can be performed as follows. First, cool the chip to room temperature. Then add 200 μL of 0.1 M NaOH solution to denature the extended DNA double strands from the second round of sequencing. Then add 10 ml of washing buffer to wash away residual NaOH and denatured DNA single strands.

[0193] Then, the sequencing primers were rehybridized to the DNA array, as described above. The sequencing reaction using the third set of reaction solutions was performed as follows:

[0194] 1. Add 10 mL of cleaning solution to rinse the chip.

[0195] 2. Cool the chip down to 4°C.

[0196] 3. Add 100 μL of reaction solution 3A.

[0197] 4. Heat the chip to 65°C.

[0198] 5. Wait 1 minute.

[0199] 6. Take fluorescence images at an excitation laser wavelength of 473 nm.

[0200] 7. Add 10 mL of cleaning solution to rinse the chip.

[0201] 8. Cool the chip down to 4°C.

[0202] 9. Add 100 μL of reaction solution 3B.

[0203] 10. Heat the chip to 65°C.

[0204] 11. Wait 1 minute.

[0205] 12. Take fluorescence images at an excitation laser wavelength of 473 nm.

[0206] 13. Repeat steps 1-12 50 times to obtain 100 fluorescence signals.

[0207] At this point, the three rounds of sequencing were completed.

[0208] Example 2: Sequencing using the "2+2 two-color" method

[0209] In this embodiment, three sets of reaction solutions were prepared. Each set contained two vials, and each vial contained two types of nucleotide bases. The two nucleotide bases in each vial were labeled with two different fluorophores (so that their emission wavelengths were different) to distinguish the signals from the two nucleotide bases.

[0210] In this embodiment, the two types of fluorophores are X and Y. For each set, the two vials contain exactly all four bases of the sequencing reaction. The six vials (two vials per set) contain no duplicate solutions.

[0211] Table 8: Reaction solution in the "2+2 two-color" method

[0212] First bottle Second bottle First set AX+CY GX+TY Second set AX+GY CX+TY The third set AX+TY CX+GY

[0213] The complete sequencing process consists of three rounds of sequencing, which can be performed sequentially in any suitable order. Each round of sequencing uses one of the three sets of reaction solutions listed in Table 8. For example, the order of the three rounds could be set 1 → set 2 → set 3, or set 2 → set 3 → set 1, etc. Each round of sequencing uses the three sets of reaction solutions listed above, and all other conditions are exactly the same (e.g., the same sequencing primers and reaction conditions are used in all three rounds). Two bottles from the same set of reaction solutions can also be used in any suitable order; for example, the first bottle can be used before or after the second bottle.

[0214] Each round of sequencing includes:

[0215] 1. Hybridize the sequencing primers onto the prepared DNA array.

[0216] 2. Begin the sequencing reaction. Steps 2.1-2.4 can be repeated multiple times.

[0217] 2.1. Add a first bottle of reaction solution (e.g., the first or second bottle of the first set) to the sequencing reaction mixture (e.g., in a flow cell) to allow the reaction to proceed, and collect fluorescence signals from fluorescent group X and fluorescent group Y, respectively.

[0218] 2.2. Clean all residual reaction solution and fluorescent molecules from the flow cell.

[0219] 2.3. Add a second bottle of reaction solution (e.g., the second bottle or the first bottle of the first set) to the sequencing reaction mixture to allow the reaction to proceed, and collect fluorescence signals from fluorescent group X and fluorescent group Y, respectively.

[0220] 2.4. Clean all residual reaction solution and fluorescent molecules from the flow cell.

[0221] 3. Unwind the extended sequencing primers.

[0222] At this point, a new round of sequencing reactions can begin.

[0223] Example 3: Detecting and / or correcting sequencing errors

[0224] In this embodiment, the single-stranded DNA molecule to be sequenced is immobilized on a solid surface. The immobilization method can be chemical cross-linking, molecular adsorption, etc. The 3' or 5' end of the DNA can be immobilized on the surface. The DNA to be tested contains a fixed fragment with a known sequence that can hybridize complementaryly with sequencing primers. The sequence from the 3' end of this fragment to the 3' end of the DNA to be tested is the region of the sequence to be tested. In this embodiment, the sequence to be tested is 5'-TGAACTTTAGCCACGGAGTA-3' (SEQ ID NO:2).

[0225] First, sequencing primers are hybridized to fragments containing known sequences of the target DNA. Each nucleotide substrate molecule has a fluorescently switching functional group attached to its bases; the number of phosphate molecules is four.

[0226] dG4P and dT4P, along with the corresponding reaction buffer, enzyme, and metal ions, are added to the reaction system to initiate a sequencing reaction that generates a fluorescent signal. The signal is acquired using a CCD (charge-coupled device). The values ​​of these fluorescent signals are recorded. This reaction is recorded as the first reaction.

[0227] The residual dG4P and dT4P from the reaction are washed away. Then, dA4P and dC4P are added to the reaction system to initiate the same sequencing reaction as above, and the fluorescence signal value is recorded. This reaction should be recorded as the second reaction. This method is also known as the monochromatic 2+2 sequencing method.

[0228] Repeat the above process. For odd-numbered reactions, add dG4P and dT4P; for even-numbered reactions, add dA4P and dC4P to obtain a set of sequencing signal values: x = (2, 3, 3, 1, 1, 3, 2, 1, 2, 1).

[0229] For example, the newly synthesized DNA strands from the sequencing reaction can be unwound and washed away using high temperatures or strongly hydrophilic substances (such as urea and formamide). Sequencing primers are then re-hybridized to the template DNA. For odd-numbered reactions, dC4P and dT4P are added; for even-numbered reactions, dA4P and dG4P are added to obtain a set of sequencing signal values: y = (1, 4, 4, 2, 2, 1, 1, 4, 1, 1).

[0230] For example, the newly synthesized DNA strands from the sequencing reaction can be unwound and washed away using high temperatures or strongly hydrophilic substances (such as urea and formamide). Sequencing primers are then re-hybridized to the template DNA. For odd-numbered reactions, dA4P and dT4P are added; for even-numbered reactions, dC4P and dG4P are added to obtain a set of sequencing signal values: z = (1, 1, 2, 1, 4, 3, 1, 3, 1, 1, 2).

[0231] The sequencing signal values ​​were then analyzed based on the type of nucleotide base represented by the signal to obtain sequencing information. For each residue of the target DNA, the common base in the three signals was identified and listed in the table below as the nucleotide residue at that position.

[0232] Table 9: Sequencing results before correction

[0233] signal x K K M M M K K K M K M M M K K M K K M Signal y Y R R R R Y Y Y Y R R Y Y R Y R R R R Y R signal z W S W W S W W W W S S S W S S S W S W W common base T G A A ？ T T T ？ G ？ C ？ G ？ ？ ？ G A ？ ？

[0234] When analyzing the common bases at each position of the three sets of signals, several positions showed no common bases. This indicates that an error has occurred in the sequence. In this embodiment, changing the second value of signal Y from 4 to 3 and changing the sixth value of signal X from 3 to 4 will result in the signals shown in the table below.

[0235] Table 10: Corrected sequencing results

[0236]

[0237] In the table above, "changing the second value of signal y from 4 to 3" is represented by a strikethrough R, and "changing the sixth value of signal X from 3 to 4" is represented by adding an M (in italics and underline). After these two modifications, all positions of the three sets of signals have common bases, and the sequence composed of these common bases is the DNA sequence to be tested. This result shows that by "encoding" DNA with degenerate indicators (e.g., M, K, R, Y, W, S, B, and D), the method can effectively detect errors that occur during sequencing, while the method of "decoding" the sequence can effectively correct these errors. The short sequence of this embodiment can effectively explain the error correction method provided in this disclosure. The modification method used in this embodiment is the method with the least variation and is also the simplest method to achieve subsequent sequence matching. In practical applications, mathematical models can be constructed to achieve such variations. In practically feasible algorithms, all possible variations are statistically based on probability. After correction of probability parameters, the above variation is the most likely correct variation. On the one hand, this calculation is a simple application of the maximum likelihood method based on Bayesian probability. On the other hand, this calculation method is generally a conventional mathematical method.

[0238] By encoding and decoding DNA sequences, this method can effectively improve sequencing accuracy when applied to DNA sequencing signals. For decoding, the sequencing signal is represented as a weighted graph, such as... Figure 1 As shown. A weighted graph is denoted as G(V, E, W), where V are the nodes of the graph, E are the edges of the graph, and W is the weight of each edge (e.g., a real number). The encoding and decoding process is explained below, assuming the sequencing signal counted by number i is a. i .

[0239] 1) For each signal a i If the nucleotide provided in the i-th sequencing reaction is X, then plot node a of h. i Each node represents an X base.

[0240] 2) This a i The nodes are connected sequentially and in an orderly manner, meaning that the first point in this node points to the second point, the second point points to the third point, and so on.

[0241] 3) The last node of this node has a cycle pointing to itself.

[0242] 4) All nodes in the i-th iteration point to the first node representing the (i+1)-th iteration.

[0243] 5) Assign weights to all edges based on the statistical results of a large amount of sequencing data.

[0244] If a DNA sequence is sequenced once using M / K, R / Y, and W / S combinations respectively, three sequencing signals are obtained. These three sequencing signals can be represented graphically using the methods described above, as shown below. Figure 1 As shown.

[0245] The three sets of signals of sequence 5'-TGAACTTTAGCCACGGAGTA-3 (SEQ ID NO:2) are as follows (including errors):

[0246] M / K: 2, 3, 3, 1, 1, 3, 2, 1, 2, 1

[0247] R / Y: 1, 4, 4, 2, 2, 1, 1, 4, 1, 1

[0248] W / S: 1, 1, 2, 1, 4, 3, 1, 3, 1, 1, 2

[0249] A path in a directed weighted graph is defined as a set of nodes in the directed weighted graph, i.e., v1v2...v... n This set of nodes can be completely different, or some nodes can be the same (for example, v1 and v2 represent the same node). Furthermore, for any two adjacent nodes v in this set... i and v i+1In this graph, there exists a directed edge from v. i Point to v i+1 The weight of a path is defined as the sum of the weights of all edges in that path. If each sequencing signal is represented as a weighted graph, then each path in the graph represents a possible DNA sequence. Signal decoding involves finding the maximum common path among all graphs. Specific implementation methods include exhaustive search, greedy search, dynamic programming, and heuristic search.

[0250] Example 4: Detecting and / or correcting sequencing errors

[0251] Following the sequencing method described in Example 3, 5000 DNA sequences of 400 bp each were decoded; all DNA was divided into 5 groups of 1000 DNA sequences each. The coding accuracy and decoding accuracy are summarized in the table below, based on the sequencing correction method described in Example 3.

[0252] Table 11: Sequencing Accuracy

[0253] Group Code accuracy Decoding accuracy 1 0.9736 0.9917 2 0.9813 0.9951 3 0.9878 0.9977 4 0.9953 0.9997 5 0.9973 0.9999

[0254] Clearly, the encoder-decoder method presented in this paper can effectively improve sequencing accuracy. For example, when the error rate is 0.0364 (in other words, the accuracy is 0.9736), the corrected error rate becomes 0.0083 (in other words, the accuracy becomes 0.9917). When the error rate is 0.0047, the corrected error rate becomes 0.0003. By comparison, when the error rate before correction decreases by 7.74 times (0.0364 divided by 0.0047), the corrected error rate will decrease by 27.6 times (0.0083 divided by 0.0003). The overall data shows a clear trend: reducing the sequencing error rate leads to a further reduction in the inferred error rate after correction. In other words, using the correction method disclosed in this paper, any small improvement to the sequencing method that can reduce the error rate can lead to a more significant reduction in the error rate of the corrected sequencing data.

[0255] The encoding accuracy and decoding accuracy of each group are calculated separately and represented using violin plots and box plots, such as... Figure 2 As shown.

[0256] Based on the characteristics of the modified signals during encoding, sequences with a higher probability of correct decoding can be selected, further improving decoding accuracy. The frequency distribution histogram of the number of modified signals in each sequence during decoding in the above data is shown below. Figure 3 As shown, this frequency distribution histogram has the following characteristics: there is a sharp peak on the left side of the image, while the frequencies on the right side of the peak exhibit a long tail distribution. If the sequences in the long-tailed distribution region of the image are discarded, and only the sequences in the peak region are used for analysis, the decoding accuracy can be further improved by 2-10 times.

[0257] Figure 4 This represents the relationship between the number of signals that occurred incorrectly during encoding and the number of signals that were incorrectly modified during decoding. The horizontal axis represents the number of signals that occurred incorrectly during encoding, and the vertical axis represents the correlation between the number of signals that were incorrectly modified during decoding. The grayscale value of the color indicates the proportion of times that point was counted out of all sequences. Figure 3 In most cases, even if an error occurs during decoding, the modified signal and the signal that actually erred are very close to each other. Therefore, this characteristic can be used to judge the quality of decoding. If a signal and its neighboring signals are not modified during decoding, the base type represented by this signal has extremely high reliability.

[0258] Example 5: Detecting and / or correcting sequencing errors

[0259] In this embodiment, the single-stranded DNA molecule to be sequenced is immobilized on a solid surface. The immobilization method can be chemical cross-linking, molecular adsorption, etc. The 3' or 5' end of the DNA can be immobilized on the surface. The DNA to be tested contains a fixed fragment with a known sequence that can hybridize complementaryly with sequencing primers. The sequence from the 3' end of this fragment to the 3' end of the DNA to be tested is the region of the sequence to be tested. In this embodiment, the sequence to be tested is 5'-TGAACTTTAGCCACGGAGTA-3' (SEQ ID NO:2).

[0260] First, sequencing primers are hybridized to a fragment containing a known sequence of the target DNA. Four types of dNTPs, along with corresponding reaction buffers, enzymes, and metal ions, are added to the reaction system. The 3' end of each type of dNTP is blocked by a chemical group. Furthermore, dGTP and dTTP are each labeled with a fluorescent group of the same color, while dATP and dCTP are each labeled with a fluorescent group of a different type of dNTP of the same color. During the reaction, dNTPs complementary to the bases at the extension site on the template DNA are incorporated into the nascent DNA strand by DNA polymerase. After the reaction, residual dNTPs are washed away, and the fluorescence signal is recorded using a CCD. The above reaction is repeated to obtain a set of sequencing signal values: x = KKMMMKKKMKMMMKKMKKM.

[0261] For example, the newly synthesized DNA strands from the sequencing reaction can be unwinded and washed away using high temperatures or strongly hydrophilic substances (such as urea and formamide). Sequencing primers are then re-hybridized to the DNA template, and the sequencing process is repeated, but dCTP and dTTP are labeled with the same color fluorescent group, while dATP and dGTP are labeled with a different color fluorescent group. The sequenced signal value is obtained as: y = YRRRRYYYYRRYYRYRRRRYR.

[0262] For example, the newly synthesized DNA strands from the sequencing reaction can be unwound and washed away using high temperatures or strongly hydrophilic substances (such as urea and formamide). Sequencing primers are then re-hybridized to the DNA template, and the sequencing process is repeated, but dATP and dTTP are labeled with the same color fluorescent group, and dCTP and dGTP are labeled with a different color fluorescent group. The value of the sequencing signal is obtained: z = WSWWSWWWWSSSWSSSWSWW.

[0263] The sequencing signal values ​​were then analyzed based on the type of nucleotide base represented by the signal to obtain sequencing information. For each residue of the target DNA, the common base in the three signals was identified and listed in the table below as the nucleotide residue at that position.

[0264] Table 12: Sequencing results before correction

[0265] signal x K K M M M K K K M K M M M K K M K K M Signal y Y R R R R Y Y Y Y R R Y Y R Y R R R R Y R signal z W S W W S W W W W S S S W S S S W S W W common base T G A A ？ T T T ？ G ？ C ？ G ？ ？ ？ G A ？ ？

[0266] When analyzing the common bases at each position of the three sets of signals, several positions showed no common bases. This indicates that an error has occurred in the sequence. In this embodiment, changing the second value of signal Y from 4 to 3 and changing the sixth value of signal X from 3 to 4 will result in the signals shown in the table below.

[0267] Table 13: Corrected sequencing results

[0268]

[0269] In the table above, "changing the second value of signal y from 4 to 3" is represented by a strikethrough R, and "changing the sixth value of signal X from 3 to 4" is represented by adding an M (in italics and underline). After these two modifications, all positions in the three signal groups have common bases, and the sequence composed of these common bases is the DNA sequence to be tested. This result indicates that by "encoding" DNA with degenerate indicators (e.g., M, K, R, Y, W, S, B, and D), the method can effectively detect errors that occur during sequencing, while the method of "decoding" the sequence can effectively correct these errors.

[0270] Example 6: Detecting and / or correcting sequencing errors

[0271] In this embodiment, the DNA to be tested contains a fixed fragment with a known sequence that can hybridize complementaryly with the sequencing primers. The region from the 3' end of this fragment to the 3' end of the DNA to be tested is the region of the test sequence. In this embodiment, the test sequence is 5'-TGAACTTTAGCCACGGAGTA-3' (SEQ ID NO:2).

[0272] First, sequencing primers are hybridized to fragments containing known sequences of the target DNA. The reaction volume containing the template DNA molecule with the hybridization sequencing primers is divided into three portions, which can be measured in parallel or sequentially. Each portion contains four types of dNTPs, certain types of ddNTPs, and the enzymes and buffers necessary for DNA synthesis. In some respects, the added dNTPs are natural dNTPs, and the added ddNTPs have detectable labels (e.g., labels that can be detected by instruments), including but not limited to radioactive isotope labels, chemifluorescent group labels, etc. In the first portion, ddGTP and ddTTP have the same label, while ddATP and ddCTP have another identical label. In the second portion, ddCTP and ddTTP have the same label, and ddATP and ddGTP have another identical label. In the third portion, ddATP and ddTTP have the same label, and ddCTP and ddGTP have another identical label.

[0273] All three samples were reacted under suitable conditions for a period of time, during which DNA synthesis occurred. After the reaction was complete, the reaction products could be optionally washed or purified. Then, DNA electrophoresis was performed on the three reaction products. Based on the electrophoretic bands, three sequencing signals could be obtained:

[0274] x=KKMMMKKKMKMMMKKMKKM

[0275] y=YRRRRYYYYRRYYRYRRRRYR

[0276] z = WSWWSWWWWSSSWSSSWSWW

[0277] The sequencing signal values ​​are then analyzed based on the type of nucleotide base represented by the signal to obtain sequencing information. For each residue of the target DNA, the common base in the three signals is identified and listed in the table below as the nucleotide residue at that position.

[0278] Table 14: Sequencing results before correction

[0279] signal x K K M M M K K K M K M M M K K M K K M Signal y Y R R R R Y Y Y Y R R Y Y R Y R R R R Y R signal z W S W W S W W W W S S S W S S S W S W W common base T G A A ？ T T T ？ G ？ C ？ G ？ ？ ？ G A ？ ？

[0280] When analyzing the common bases at each position of the three sets of signals, several positions showed no common bases. This indicates that an error has occurred in the sequence. In this embodiment, changing the second value of signal Y from 4 to 3 and changing the sixth value of signal X from 3 to 4 will result in the signals shown in the table below.

[0281] Table 15: Corrected sequencing results

[0282]

[0283] In the table above, "changing the second value of signal y from 4 to 3" is represented by a strikethrough R, and "changing the sixth value of signal X from 3 to 4" is represented by adding an M (in italics and underline). After these two modifications, all positions of the three signal groups have common bases, and the sequence composed of these common bases is the DNA sequence to be tested. This result indicates that by "encoding" DNA with degenerate indicators (e.g., M, K, R, Y, W, S, B, and D), the method can effectively detect errors that occur during sequencing, while the method of "decoding" the sequence can effectively correct these errors.

[0284] Example 7: Sequencing was performed using a "2+2 two-color three-round" method.

[0285] In this embodiment, the single-stranded DNA molecule to be sequenced is immobilized on a solid surface. The immobilization method can be chemical cross-linking, molecular adsorption, etc. The 3' or 5' end of the DNA can be immobilized on the surface. The DNA to be tested contains a fixed fragment with a known sequence that can hybridize complementaryly with sequencing primers. The sequence from the 3' end of this fragment to the 3' end of the DNA to be tested is the region of the sequence to be tested. In this embodiment, the sequence to be tested is 5'-TGAACTTTAGCCACGGAGTA-3' (SEQ ID NO:2).

[0286] First, sequencing primers are hybridized to fragments containing known sequences of the target DNA. dG4P and dT4P (each labeled with a different fluorescent group, e.g., fluorescent group X and group Y), along with the appropriate reaction buffer, enzyme, and metal ions, are added to the reaction system to initiate a sequencing reaction that produces a fluorescent signal. The signal is acquired using a CCD. The values ​​of these fluorescence signals are recorded. This reaction is recorded as the first reaction.

[0287] Then, the residual dG4P and dT4P from the reaction are washed away. Next, dA4P and dC4P (each labeled with a fluorescent group emitting a different color, such as fluorescent group X and group Y) are added to the reaction system to initiate the same sequencing reaction as above, and the fluorescence signal values ​​are recorded. This reaction should be recorded as the second reaction.

[0288] Repeat the above process. For odd-numbered reactions, add dG4P and dT4P; for even-numbered reactions, add dA4P and dC4P. Each reaction adds two types of dN4P labeled with different colored fluorescent groups. A set of signal values ​​can be obtained: x = (1G+1T, 2A+1C, 0G+3T, 1A+0C, 1G+0T, 1A+2C, 2G+0T, 1A+0C, 1G+1T, 1A+0C).

[0289] For example, the newly synthesized DNA strands from the sequencing reaction can be unwound and washed away using high temperatures or strongly hydrophilic substances (such as urea and formamide). Sequencing primers are then re-hybridized to the template DNA. Two types of dN4P labeled with different colored fluorescent groups are added for each reaction. A set of signal values ​​can be obtained: y = (0C+1T, 3A+1G, 1C+3T, 1A+1G, 2C+0T, 1A+0G, 1C+0T, 1A+3G, 0C+1T, 1A+0G).

[0290] For example, the newly synthesized DNA strands from the sequencing reaction can be unwound and washed away using high temperatures or strongly hydrophilic substances (such as urea and formamide). Sequencing primers are then re-hybridized to the template DNA. For odd-numbered reactions, dG4P and dT4P are added; for even-numbered reactions, dA4P and dC4P are added. Each reaction uses two types of dN4P labeled with different colored fluorescent groups. This yields a sequencing signal: z = (0A+1T, 0C+1G, 2A+0T, 1C+0G, 1A+3T, 2C+1G, 1A+0T, 0C+1G, 1A+1T).

[0291] This method is called the "2+2 two-color" sequencing method. Sequence information can be obtained from the sequencing data of any two rounds of sequencing. It can be considered as orthogonal sequencing results.

[0292] The sequencing signal values ​​were then analyzed based on the type of nucleotide base represented by the signal to obtain sequencing information. For each residue of the target DNA, the common base in the three signals was identified and listed in the table below as the nucleotide residue at that position.

[0293] Table 16: Sequencing results before correction

[0294] xA A A A A A A xC C C C xG G G G G G xT T T T T T yA A A A A A A A yC C C C C yG G G G G G yT T T T T T zA A A A A A A zC C C C C zG G G G G G zT T T T T T common base T G A A ？ ？ T T ？ ？ ？ C ？ ？ ？ ？ ？ ？ ？ ？ ？

[0295] When analyzing the common bases at each position of the three sets of signals, no common bases were found at several positions, indicating an error in the sequence. Changing the second value of signal y (3A+1G) to (2A+1G) and the sixth value of signal X (1A+2C) to (1A+3C) will result in the signals shown in the table below.

[0296] Table 17: Corrected sequencing results

[0297]

[0298]

[0299] In the table above, "the second value of signal y (3A+1G) changed to (2A+1G)" is represented by a strikethrough A, and "the sixth value of signal x (1A+2C) changed to (1A+3C)" is represented by adding a C (in italics and underlined). After these two modifications, all positions of the three signal groups have common bases, and the sequence composed of these common bases is the DNA sequence to be tested. This result indicates that by "encoding" DNA with degenerate indicators (e.g., M, K, R, Y, W, S, B, and D), the method can effectively detect errors that occur during sequencing, while the method of "decoding" the sequence can effectively correct these errors.

[0300] Example 8: Sequencing using the "2+2 two-color two-round" method

[0301] In this embodiment, the single-stranded DNA molecule to be sequenced is immobilized on a solid surface. The immobilization method can be chemical cross-linking, molecular adsorption, etc. The 3' or 5' end of the DNA can be immobilized on the surface. The DNA to be tested contains a fixed fragment with a known sequence that can hybridize complementaryly with sequencing primers. The sequence from the 3' end of this fragment to the 3' end of the DNA to be tested is the region of the sequence to be tested. In this embodiment, the sequence to be tested is 5'-TGAACTTTAGCCACGGAGTA-3' (SEQ ID NO:2).

[0302] First, sequencing primers are hybridized to fragments containing known sequences of the target DNA. dG4P and dT4P (each labeled with a different fluorescent group, e.g., fluorescent group X and group Y), along with the appropriate reaction buffer, enzyme, and metal ions, are added to the reaction system to initiate a sequencing reaction that produces a fluorescent signal. The signal is acquired using a CCD. The values ​​of these fluorescence signals are recorded. This reaction is recorded as the first reaction.

[0303] Then, the residual dG4P and dT4P from the reaction are washed away. Next, dA4P and dC4P (each labeled with a fluorescent group emitting a different color, such as fluorescent group X and group Y) are added to the reaction system to initiate the same sequencing reaction as above, and the fluorescence signal values ​​are recorded. This reaction should be recorded as the second reaction.

[0304] Repeat the above process. For odd-numbered reactions, add dG4P and dT4P; for even-numbered reactions, add dA4P and dC4P. Each reaction adds two types of dN4P labeled with different colored fluorescent groups. A set of signal values ​​can be obtained: x = (1G+1T, 2A+1C, 0G+3T, 1A+0C, 1G+0T, 1A+2C, 2G+0T, 1A+0C, 1G+1T, 1A+0C).

[0305] For example, the newly synthesized DNA strands from the sequencing reaction can be unwound and washed away using high temperatures or strongly hydrophilic substances (such as urea and formamide). Sequencing primers are then re-hybridized to the template DNA. Two types of dN4P labeled with different colored fluorescent groups are added for each reaction. A set of signal values ​​can be obtained: y = (0C+1T, 3A+1G, 1C+3T, 1A+1G, 2C+0T, 1A+0G, 1C+0T, 1A+3G, 0C+1T, 1A+0G).

[0306] The sequencing signal values ​​were then analyzed based on the type of nucleotide base represented by the signal to obtain sequencing information. For each residue of the target DNA, the common base in the three signals was identified and listed in the table below as the nucleotide residue at that position.

[0307] Table 18: Sequencing results before correction

[0308] xA A A A A A A xC C C C xG G G G G G xT T T T T T yA A A A A A A A yC C C C C yG G G G G G yT T T T T T common base T G A A ？ ？ T T ？ ？ ？ C ？ ？ ？ ？ ？ ？ ？ ？ ？

[0309] When analyzing the common bases at each position of the two sets of signals, no common bases were found at several positions, indicating an error in the sequence. Changing the second value of signal y (3A+1G) to (2A+1G) and the sixth value of signal X (1A+2C) to (1A+3C) will result in the signals shown in the table below.

[0310] Table 19: Corrected sequencing results

[0311]

[0312] In the table above, "the second value of signal y (3A+1G) changed to (2A+1G)" is represented by a strikethrough A, and "the sixth value of signal x (1A+2C) changed to (1A+3C)" is represented by adding a C (in italics and underlined). After these two modifications, all positions of the three signal groups have common bases, and the sequence composed of these common bases is the DNA sequence to be tested. This result indicates that by "encoding" DNA with degenerate indicators (e.g., M, K, R, Y, W, S, B, and D), the method can effectively detect errors that occur during sequencing, while the method of "decoding" the sequence can effectively correct these errors.

[0313] Example 9: Sequencing using the "1+3, monochromatic" method

[0314] In this embodiment, the single-stranded DNA molecule to be sequenced is immobilized on a solid surface. The immobilization method can be chemical cross-linking, molecular adsorption, etc. The 3' or 5' end of the DNA can be immobilized on the surface. The DNA to be tested contains a fixed fragment with a known sequence that can hybridize complementaryly with sequencing primers. The sequence from the 3' end of this fragment to the 3' end of the DNA to be tested is the region of the sequence to be tested. In this embodiment, the sequence to be tested is 5'-TGAACTTTAGCCACGGAGTA-3' (SEQ ID NO:2).

[0315] First, sequencing primers are hybridized to fragments containing known sequences of the target DNA. dC4P, dG4P, and dT4P, along with the corresponding reaction buffers, enzymes, and metal ions, are added to the reaction system to initiate a sequencing reaction that generates a fluorescent signal. The signal is acquired using a CCD. The values ​​of these fluorescence signals are recorded. This reaction is recorded as the first reaction.

[0316] Then, the residual dC4P, dG4P, and dT4P from the reaction are washed away. Next, dA4P is added to the reaction system to initiate the same sequencing reaction as described above, and the fluorescence signal value is recorded. This reaction should be recorded as the second reaction.

[0317] Repeat the above process. For odd-numbered reactions, add dC4P, dG4P, and dT4P; for even-numbered reactions, add dA4P. Obtain a set of signal values: x = (2, 2, 4, 1, 3, 1, 3, 1, 2, 1).

[0318] For example, the newly synthesized DNA strands from the sequencing reaction are unwound and washed away using high temperature or strongly hydrophilic substances (such as urea and formamide). Sequencing primers are then re-hybridized to the template DNA. dA4P, dG4P, and dT4P are added for odd-numbered reactions, and dC4P is added for even-numbered reactions. A set of signal values ​​is obtained: y = (4, 1, 6, 2, 1, 1, 6).

[0319] For example, the newly synthesized DNA strands from the sequencing reaction are unwound and washed away using high temperature or strongly hydrophilic substances (such as urea and formamide). Sequencing primers are then re-hybridized to the template DNA. dA4P, dC4P, and dT4P are added for odd-numbered reactions, and dG4P is added for even-numbered reactions. A set of signal values ​​is obtained: z = (1, 1, 7, 1, 4, 2, 1, 1, 2).

[0320] For example, the newly synthesized DNA strands from the sequencing reaction can be unwound and washed away using high temperature or a strongly hydrophilic substance (such as urea and formamide). Sequencing primers are then re-hybridized to the template DNA. dT4P is added for odd-numbered reactions, and dA4P, dC4P, and dG4P are added for even-numbered reactions. A set of signal values ​​is obtained: w = (1, 4, 3, 9, 1, 1).

[0321] The sequencing signal values ​​were then analyzed based on the type of nucleotide base represented by the signal to obtain sequencing information. For each residue of the target DNA, the common base in the three signals was identified and listed in the table below as the nucleotide residue at that position.

[0322] Table 20: Sequencing results before correction

[0323]

[0324]

[0325] When analyzing the common bases at each position of the two sets of signals, no common bases were found at several positions, indicating an error in the sequence. Changing the third value of signal y from 6 to 5 and the fourth value of signal w from 9 to 10 will alter the signals as shown in the table below.

[0326] Table 21: Corrected sequencing results

[0327]

[0328] In the table above, "changing the third value of signal y from 6 to 5" is represented by a strikethrough D, and "changing the fourth value of signal w from 9 to 10" is represented by adding a V (in italics and underline). After these two modifications, all positions in the four signal groups have common bases, and the sequence composed of these common bases is the target DNA sequence. This result indicates that by "encoding" DNA with degenerate indicators (e.g., M, K, R, Y, W, S, B, and D), the method can effectively detect errors occurring during sequencing, while the method of "decoding" the sequence can effectively correct these errors.

[0329] Example 10: Methods for detecting and / or correcting sequencing errors

[0330] Section 1: Decoding

[0331] 1.1. Characteristics of different sequencing flow graphs

[0332] 1.1.1 Information Entropy of DNA

[0333] For a sufficiently long DNA molecule of length d, if the types of each base are independent and the probability of each type of base appearing is equal, that is,

[0334]

[0335] So, the Shannon entropy of this DNA molecule is

[0336]

[0337] 1.1.2 Information Entropy of Pyrosequencing

[0338] In this embodiment, the term "degenerate sequence of a DNA molecule" is used to describe sequences that have the same nucleotide type but whose homopolymer length is equal to 1. For example, the degenerate sequence of 'ATTCCCG' is 'ATCG'.

[0339] In this embodiment, the term "dark cycle" is used to describe the reaction cycle in 1×4 sequencing, where the signal strength of the sequencing is 0.

[0340] Consider a 1×4 sequencing process using a flow graph (T, C, A, G, T, C, A, G, ...). Without loss of generality, assume cycle 1 is not a dark cycle and the delivered nucleotide is T. The second homopolymer will be of type C, A, or G, with a 1 / 3 equal probability. If the second homopolymer is C, then cycle 2 will not be a dark cycle. If the second homopolymer is A, then cycle 2 is a dark cycle, and cycle 3 is not a dark cycle. If the second homopolymer is G, then cycles 2 and 3 are both dark cycles, and cycle 4 is not a dark cycle. Therefore, there are two non-dark cycles N. dark The probability distribution of the number of dark cycles is as follows:

[0341]

[0342] Therefore N dark The expected value is That is, the ratio of the number of non-dark cycles to the number of dark cycles is 1:1.

[0343] The probability of signal strength x in a non-dark cycle is:

[0344]

[0345] The average signal strength (expected) for non-dark cycles is:

[0346]

[0347] Since the ratio of non-dark cycles to dark cycles is 1:1, the average signal intensity (expected value) for any 1x4 sequencing cycle is:

[0348]

[0349] The expected cycle for sequencing a DNA molecule of length d is:

[0350]

[0351] The Shannon entropy of a single signal from 1x4 sequencing is:

[0352]

[0353] Note that if we consider a single signal from a 1x4 sequencing run without any prior knowledge of its previous cycle, the probability of the signal intensity is:

[0354]

[0355] The average signal strength remains constant:

[0356]

[0357] However, the Shannon entropy result is:

[0358]

[0359] and:

[0360] H′ 1×4 ×N 1×4 =2.31d>H DNA

[0361] This counterintuitive phenomenon is due to the fact that the signal intensity of each cycle in the same 1x4 sequencing run is not independent, so its Shannon entropy cannot be simply added together.

[0362] 1.1.3 Information entropy of ECC sequencing (monochrome)

[0363] The probability of signal intensity x in 2+2 sequencing is:

[0364]

[0365] Therefore, the average signal intensity of 2+2 sequencing is:

[0366]

[0367] The Shannon entropy of a single monochrome 2+2 sequencing signal is:

[0368]

[0369] Therefore, a single-color sequencing provides some information.

[0370] 1.1.4 Information entropy of ECC sequencing (two colors)

[0371] The probability of signal intensity (x, y) in 2+2 sequencing is:

[0372]

[0373] The Shannon entropy of a single two-color 2+2 sequencing signal is:

[0374]

[0375] Given that the average signal intensity of continuous sequencing is 2, d / 2 cycles are required to complete the sequencing. Therefore, one round of two-color sequencing provides a certain amount of information.

[0376] 1.1.5 Differences in sequencing reactions required for three rounds of ECC sequencing

[0377] When sequencing the same molecule, different base combinations require different numbers of cycles. For example, for the sequence 'ACACA', 5 cycles are needed to extend the entire molecule with R / Y, but only 1 cycle is needed for M / K. Calculate the sequencing cycles required for the three base combinations M / K, R / Y, and W / S by randomly generating 10,000 different DNA sequences of 100 bp in length. Figure 11 The range distribution of the three base combinations is shown. The average value of the range is 8.43, as indicated by the red vertical line.

[0378] 1.2 ECC Decoding Algorithm

[0379] 1.2.1 Signal Diagram

[0380] The homopolymer length of the DNA molecules to be sequenced is h, and the signal in ECC sequencing (after phase correction) is s = (s1, s2, ..., s...). n Assuming that in loop i, h i A given signal s i The probability is P(h) i |s i Therefore, the signal can be represented as shown in the diagram below.

[0381] Each signal s in s i All in s i Each node represents a signal s. i Given nodes, draw a graph from the j-th node to the (j+1)-th node, where j = 1, 2, ... s i The directed edges. Draw the edge representing the signal s. i The s i The directed edges from each node (the last node) to itself. Draw the directed edges from the node representing signal s. i Each node to the representation s i+1 The directed edge of the first node.

[0382] Each node in the diagram can be labeled as 1 or 0 depending on the type of nucleotide delivered in that cycle.

[0383] Next, we define the weights of the paths in the graph representing the sequencing signal. The paths are limited to a series of nodes v1v2...vK For each adjacent node v k and v k+1 There exists from v k to v k+1 A directed edge. Allow v k and v k+1 They are the same node, and in this case, both are the last nodes of the signal representing a specific cycle.

[0384] If in path v1v2...v K middle, Both represent signal s i If the nodes are of type t, then these nodes are all assigned a weight P(t). i |s i ). Path v1v2...v K The weight of a path is defined as the product of the weights of all its nodes. For ease of calculation, the weight of a node can also be specified as the logarithm of its probability, and the weight of the path can be adjusted to the sum of the weights of all its nodes.

[0385] The path in the graph represents a possible DPL (degenerate polymer length) from the sequencing results, such as... Figure 1 As shown. Specifically, from the representation s i The edge from the last node to itself represents an insertion, and the edge from the node representing s represents an insertion. i The nodes (except the last one) to the representation s i+1 The edge of the first node represents deletion.

[0386] For a DNA molecule, sequencing is performed using base combinations M / K, R / Y, and W / S to obtain three signals. Each of the three signals can be represented as a graph as described above. Assuming... and These are paths from three graphs, each with the same length K: if If the parity check is true for all k = 1, 2, ..., then these three paths are called common paths of the three graphs. Clearly, the decoding problem is actually about finding the common path of the three graphs with the maximum weight (Maximum Common Path, MCP).

[0387] 1.2.2 ECC Decoding via Dynamic Programming

[0388] Terms used in this section:

[0389] Codeword space and nodes: The element index [i,j,k] (i,j,k∈N) called a node represents the 3D discrete space of the codeword, including the 1st round of the i-th bit, the 2nd round of the j-th bit, and the 3rd round of the k-th bit. Compared to BS (degenerate sequence), the codeword space records every possible codeword alignment in an intuitive way.

[0390] Jump: A one-way link between bits.

[0391] Connection: A directed link between nodes. A connection consists of three hops in different rounds.

[0392] Parity of a node: the three-bit XOR value of the node.

[0393] Preparation of necessary variables

[0394] Assuming the maximum length of the three binary strings is N, the binary strings (degenerate nucleotide sequences) are preprocessed into a lookup table.

[0395] BS (Binary String) is a 3*N Boolean matrix, representing the binary version of the sequencing data. Values ​​0 (or 1) represent degenerate bases. For example:

[0396] [0,0,1,1,1,0,0,0,1,0,1,1,1,0,0,1,0,0,1,0,0;

[0397] 0,1,1,1,1,0,0,0,0,1,1,0,0,1,0,1,1,1,1,0,1;

[0398] 1,0,1,1,0,1,1,1,1,0,0,0,1,0,0,0,1,0,1,1,0;]

[0399] CNS (Cyclic Number Sequence) is a 3*N integer matrix that records the cycle number of binary bits (degenerate bases) read from it.

[0400] [1,1,2,2,2,3,3,3,4,5,6,6,6,7,7,8,9,9,10,11,11;

[0401] 1,2,2,2,2,3,3,3,3,4,4,5,5,6,7,8,8,8,8,9,10;

[0402] 1,2,3,3,4,5,5,5,5,6,6,6,7,8,8,8,9,10,11,11,12;]

[0403] DPL (Degenerate Polymer Length) is a 3*N integer matrix that records the DPL of the read loop.

[0404] [2,2,3,3,3,3,3,3,1,1,3,3,3,2,2,1,2,2,1,2,2;

[0405] 1,4,4,4,4,4,4,4,4,2,2,2,2,1,1,4,4,4,4,1,1;

[0406] 1,1,2,2,1,4,4,4,4,3,3,3,1,3,3,3,1,1,2,2,1;]

[0407] These tables make it easy to look up DPL information for cycles and individual bits. Different cycle numbers indicate that two different bits come from different cycles. DPL is the input to the scoring function of dynamic programming. For example, the cycle number for the 11th bit in the first round is CNS(1,11) = 6, and the DPL (the individual bit in that cycle) is DPL(1,11) = 3.

[0408] Initialize the comparison variables

[0409] SCORE = a number matrix of size N*N*N, defaulting to NaN.

[0410] CONNECTION = A node matrix of size N*N*N

[0411] ROUTABLE = a Boolean matrix of size N*N*N, defaults to false, but Routable(1,1,1) = true.

[0412] If the query ROUTABLE(node) is true, it means that the node has a connection that returns node (1,1,1).

[0413] STEP = a tuple matrix of 3 elements, with a size of N*N*N. The default value is (0,0,0), but Step(1,1,1) = (1,1,1).

[0414] The 3-element tuple of the STEP (node) records the number of bits counted in the three sequencing rounds within that cycle. The STEP value is incremented by 1 for each link in the cycle and reset to 1 in a new cycle. When jumping between cycles, the STEP value is used as the corrected DPL value.

[0415] Pseudocode for the comparison process

[0416]

[0417] Parity (Node) Checks the parity of a node. The binary value of node (i,j,k) is taken from BS(1,i), BS(2,j), and BS(3,k). Calculate the three-bit XOR.

[0418] Layers and Traversal. The Lth layer of the codeword space contains all nodes within the box [L,L,L], but not the nodes within the box [L-1,L-1,L-1]. Take the 5th layer as an example. This layer has 61 nodes. The traversal order must ensure that each node (index [i,j,k]) can be connected to a traversed node (index [ii,jj,kk], ii <= i, jj <= j, and kk <= k). Figure 12 This shows an example of a possible traversal order—the layer and node traversal order of a rating matrix structure.

[0419] 1.2.3 Hidden Markov Model for ECC Decoding

[0420] Similar to its application in sequence alignment, Hidden Markov Models (HMMs) can also be applied to signal decoding problems. Three symbols are introduced to describe the state: match (m), asterisk (*), and interval (-). A signal of intensity *a* is represented as a match (*a). If the ideal signal intensity is *b* and *b* > *a*, then an interval (*ba*) is added immediately after the match (*a*). If *b* < *a*, then an interval (*ab*) is added at the two orthogonal positions representing the nearest match. For example, the ideal DPL for the sequence TGAACTTTAGCCACGGAGTA in a three-base combination is:

[0421] M / K: 0, 2, 3, 3, 1, 1, 4, 2, 1, 2, 1;

[0422] R / Y: 0, 1, 3, 4, 2, 2, 1, 1, 4, 1, 1;

[0423] W / S: 1, 1, 2, 1, 4, 3, 1, 3, 1, 1, 2.

[0424] The DPL measured in the experiment is: (The bold and underlined numbers indicate errors)

[0425] M / K: 0, 2, 3, 3, 1, 1, 3 2, 1, 2, 1;

[0426] R / Y: 0, 1, 4 , 4, 2, 2, 1, 1, 4, 1, 1;

[0427] W / S: 1, 1, 2, 1, 4, 3, 1, 3, 1, 1, 2.

[0428] The signal decoded and corrected using the above representation method is:

[0429] M / K: mmmm-mmmmmmmmmm*mmmmmm

[0430] R / Y: mmmmmmmmmmmmmmmmmmmm

[0431] W / S: mmmm-mmmmmmmmmmmmmmmm

[0432] Clearly, the alignment of signals M / K, R / Y, and W / S can be viewed as a transition process of the following alignment states: (mmm), (mmm), (mmm), (mmm), (-m-), (mmm), (mmm), (mmm), (mmm), (mmm), (mmm), (mmm), (mmm), (mmm), (mmm), (*mm), (mmm), (mmm), (mmm), (mmm), (mmm), (mmm), (mmm). This inspires us to use Hidden Markov Models to describe signal alignment.

[0433] Typically, the overall hidden states of the model are: (mmm), (m--), (-m-), (--m), (*mm), (m*m), (mm*), (m**), (*m*), and (**m). Each state except (m--), (-m-), and (--m) will emit a nucleotide, the type of which is determined by the corresponding sequencing signal type. States (m--), (-m-), and (--m) will not emit any nucleotides. Simulate 1 million DNA reads to count the probability of state transitions (...). Figure 13 The Vitebi algorithm of this Hidden Markov Model will be an alternative implementation of the ECC decoding algorithm. Figure 13 The state transition network of the Hidden Markov Model for ECC decoding is shown. The width of the edge represents the order of magnitude of the transition probability.

[0434] 1.3 Other ECC Decoding Results

[0435] Example decoding results are shown in Figure 15 .

[0436] 1.4 Simulation of Decoding at Different Raw Accuracies

[0437] To further investigate the ability of ECC decoding to enhance accuracy, simulations were performed at five different levels of raw precision, each with 10,000 DNA sequences. Two parameters, γ and δ, were used to generate the probability matrix P, with entries P... ij The probability of a DPL of length i being sequenced as length j is indicated by the following steps:

[0438] 1. For each entry P in P ij ,set up

[0439] 2. N(x; μ, σ) 2 The probability density function is set to a normal distribution, i.e. but

[0440] 3. Normalize P, and the sum of each row will be equal to one.

[0441] In the simulation, γ was set to 1.6, 1.7, 1.8, 2.0, and 2.1, and δ was set to 0.1. The overall raw accuracies under these parameter settings were 97.42%, 98.34%, 98.97%, 99.64%, and 99.80%, respectively. Using the same 10,000 random 400bp DNA sequences, their theoretical DPLs were calculated, randomly modified to new values ​​according to the generated probability matrix P, and corrected using a decoding algorithm. The scoring function used in the decoding algorithm is subjected to its respective probability matrix P. If two out of three consecutive DPLs of a DNA sequence are modified by the decoding algorithm, the DNA sequence is discarded due to the high probability of incorrect decoding. The sequencing accuracy is defined as follows: if a DPL of length i is sequenced (or decoded) to length j, then the accuracy of those i bases in that DPL is [missing value]. The accuracy distribution of the first 300 bp of the DNA sequence was calculated before and after decoding, and a significant accuracy shift was found after decoding. Figure 14 This indicates the capability of the decoding algorithm. Figure 14 The simulated distribution of accuracy before and after decoding is shown.

[0442] Example 11: Error Correction Code Fluorescent DNA Sequencing

[0443] The principle of degenerate base fluorescence sequencing:

[0444] In this embodiment, a series of fluorescent sequencing substrates (using the high-performance fluorophore Tokyo Green (TG)) were developed with terminally labeled tetraphosphate nucleotides (dN4P or dN, see [link]). Figure 6 A). TG offers higher fluorescence quantum yield (0.82 at 490 nm), higher absorption coefficient, higher on / off ratio, and better photostability than previously reported fluorescent dyes. In fluorescence sequencing-by-synthesis (SBS) processes, single-stranded DNA templates are grafted onto the surface of a glass flow cell using solid-phase PCR. Figure 5Each template is then annealed with a sequencing primer, the 3' end of which serves as the starting point for the SBS reaction. In each sequencing cycle, the reaction mixture (Bst polymerase, alkaline phosphatase, and fluorescent nucleotides) reacts with those immobilized primer DNA templates. As the polymerase incorporates the corrected nucleotides into the primer ends, a non-fluorescent "dark" state dye—a triphosphate—is simultaneously released, which is then immediately switched to a highly fluorescent "bright" state via dephosphorylation. This fluorescent SBS reaction produces a native DNA duplex, leaving the 3' end of the synthetic strand unterminated (still elongable, awaiting further elongation). Substrates that form a correct Watson-Crick pair at the primer ends will elongate continuously until the first mismatch is encountered.

[0445] This feature has been used for sequencing 30-40 bases via a single-base flow graph, where one of four substrates is introduced into the reaction in each cycle. In this embodiment, a two-base flow graph is used. For example, in the first cycle of sequencing ( Figure 6 The B, K (dG & dT) reaction mixture is brought to the initiating DNA template with the starting sequence ACTTGAAA. DNA polymerase incorporates one dT and one dG to pair with the first two bases AC and produce two fluorophores, then stops at the third base T due to a mismatch. In the next M cycle, two dA and one dC are paired with the next three bases TTG, producing three fluorophores. The conjugation mixture M and K are introduced alternately to react with the initiating DNA template. Figure 6 C). The amount of fluorophore produced in each cycle is equal to the number of extended bases.

[0446] After polymerase extension is complete, the fluorescence signal is measured. The normalized fluorescence signal, representing the number of extended bases in each cycle, rather than the actual composition and sequence, is called the degenerate polymer length (DPL). Figure 6 In sequence C, the DPL array (0, 2, 3, 3, 1, ...) can be transformed into a degenerate sequence (KKMMMKKKM...), where M = A or C, and K = G or T. In addition to this MK dibasic flow graph, there are two additional dibasic flow graphs, R(A,G)-Y(C,T) and W(A,T)-S(C,G), which allow the same template to be represented as different degenerate sequences (YRRRYYYYRRYY…) and (WSWWSWWWW…). To obtain these three orthogonal degenerate sequences, a reset operation is required between sequencing rounds to denature the nascent strands and anneal the sequencing primers. Each actual base can be inferred from the three sequences by calculating the intersection of the degenerate bases. This sequencing method is called error-correcting code (ECC) sequencing, which can detect and correct sequencing errors.

[0447] Degenerate base recognition

[0448] In this embodiment, a laboratory prototype was established for fluorescence sequencing using two-base flow graphs. Similar to other SBS sequencing methods, fluorescence intensity decay is unavoidable. This decay, primarily due to incomplete reactions and template or primer loss, has posed a significant challenge to base recognition. Figure 7 A). In a typical fluorescence degenerate sequencing run, the decrease in fluorescence intensity can be normalized by an exponential decay function, with a signal drop of approximately 1% between reaction cycles. The normalized fluorescence signal in each cycle should have been rounded to DPL (Difference of Fluorescence Intensity). Figure 7 B). However, the consistency between intensity and DPL can only be maintained for about the first 30 cycles, after which phase loss cannot be ignored, meaning that the signal in each cycle becomes significantly affected by the adjacent cycles.

[0449] Phase desynchronization, or asynchrony of primer ensembles, has two main components: "lag" and "lead." Lag is primarily caused by incomplete extension, while in two-base sequencing, lead is mainly due to unintended extension caused by contaminated bases. In a given cycle, asynchronous primer ensembles produce fluorescent signals different from their corresponding DPLs. Accumulated phase desynchronization gradually reduces the correlation between the sequencing signal and the DPL array.

[0450] However, it has been shown that the cumulative effects of signal phase loss and attenuation can be estimated well based on the first-order response mechanism, with residuals between the estimates and measurements below 0.2. Furthermore, a sequence-independent iterative phase loss correction algorithm was developed to infer the DPL array for each sequencing round. Through phase loss correction, the low error range for the DPL array length can be significantly extended from the first 50 cycles (approximately 100 nt) to over 150 cycles (approximately 300 nt). Beyond this point, the phase loss algorithm cannot accurately correct crowded errors. Figure 7 C). For the same template, using RY and WS flow graphs, this correction method can also be applied to two other orthogonal degenerate sequences ( Figure 7 DE). Each of the three degenerate sequences contains a rare error (<1%) that is unlikely to be located at the same base position.

[0451] ECC sequencing information communication model

[0452] From an information theory perspective, this paper analyzes the information redundancy in dual-base sequencing. On one hand, the DPL array acquired from a single round of dual-base sequencing cannot provide a definitive DNA sequence. When there are no sequencing errors, the DNA information entropy of a random sequence of length L-nt is 2L bits, while the information entropy of its DPL array is only L bits. The orthogonality property ensures that the mutual information entropy of two DPL arrays acquired from different flow graphs is 0 bits, and the node information entropy is 2L bits. Therefore, two degenerate sequences provide both sufficient and necessary information for a definitive DNA sequence (L + L - 0 = 2L). The definitive DNA sequence can be inferred using the intersection of degenerate bases from two DPL arrays from different flow graphs. For example, if a base in the MK DPL array is sequenced as M (A / C) and in the RY DPL array as R (A / G), then it can be inferred that it is base A ({A, C} ∩ {AC} = {A}).

[0453] However, due to experimental sequencing errors, the entropy of the DPL array (referred to as l) is less than L bits. The node information provided by two DPL arrays containing this error is insufficient to infer the DNA sequence (l + l - 0 < 2L). At our current experimental error rate, additional DPL arrays are introduced to provide mutual information / redundancy (2L < 3l < 3L), which can be used to detect errors and infer definitive sequences.

[0454] An information communication model, as well as a model containing encoders, decoders, and communication channels, were also established to depict two-base sequencing with inherent error detection and correction features. Figure 8 A). Three orthogonal dibasic flow maps encode DNA sequences and information sources into three original DPL arrays (n). Analysis of the DPL distribution in the genomes of humans, yeast, and E. coli revealed that it approximates P(n) = 1 / 2. n This refers to the theoretical distribution of DPL derived from random DNA sequences. Figure 8 B also found that only 0.39% of DPLs were greater than 8.0.

[0455] The sequencing reaction is considered a communication channel through which sequencing errors are inevitably introduced into the received information. For example, in cycle 3 of the RY round, the original DPL n=3 is incorrectly measured as m=4 (3 to -4 insertion errors). Figure 8 A). The concordance between raw and measured DPLs was analyzed in 42 rounds of dual-base sequencing data. 5503 / 5609 (98.1%) of the raw DPLs (n≤9) were accurately transmitted. Figure 8 C).

[0456] The measured DPL array was rewritten as a degenerate base sequence by defining codewords as 3-tuples derived from degenerate bases at the same positions in the degenerate base sequence in the order MK, RY, and WS. Figure 8 In case A, the first few codewords are (KYW), (KRS), (MRW), etc. These codewords can be further compiled into binary format. M, R, and W are assigned logic 1, and K, Y, and S are assigned logic 0. Each degenerate sequence in any single flow graph becomes a bit string (BS). The parity of a codeword is limited to the result of a three-bit XOR operation. Figure 8 D). A codeword contains only one common base among its degenerate bases if and only if the parity is logic 1. This common base is considered the decoding result. Specifically, 111 (MRW) is decoded as base A, 100 (MYS) is decoded as C, 010 (KRS) is decoded as G, and 001 (KYW) is decoded as T. The Hamming distance between these four valid codewords is 2. On the other hand, the remaining four invalid codewords with a parity logic of 0 (no common base) indicate sequencing errors. Figure 8 In the scenario shown in A, the DNA sequence is decoded from BS, and a 3-to-4 error (MRS / 110) at the 5th codeword is captured by the decoder during parity checking. Typically, memoryless codewords with a Hamming distance of 2 are only error-detectable, not correctable. However, it has been found that BS formats generated by dual-base sequencing are not memoryless but context-dependent, providing additional information for error correction beyond error detection.

[0457] Sequence decoding using dynamic programming

[0458] Error correction decoding is performed using a dynamic programming-based algorithm. Dibase sequencing errors, or DPLs (Diagrams with different base pairs), can be easily identified in codewords through parity checks. These unique errors are simply bit insertions or deletions in the base sequence (BS), not bit alterations. When an error is detected, it is possible to correct it by changing the corresponding DPL based on the BS environment. Errors must be corrected sequentially from the first error, as DPL changes corresponding to BS move operations affect downstream codewords.

[0459] Typical embodiments are shown in Figure 9A. Detecting the first illegal codeword at codeword 5 reveals three possible sources of error: (1) an insertion error in loop 2 of the MK round, where the original DPL (n=2) is incorrectly measured as 3; (2) an insertion error in loop 2 of the RY round, where the original DPL (n=3) is measured as 4; and (3) a deletion error in loop 3 of the WS round, where the original DPL (n=3) is measured as 2. The insertion error in loop 2 of the RY round is corrected by shifting left by BS2 from the 6th bit. After this shift operation, many of the following illegal codewords pass the parity check simultaneously. Then, the second error is detected at base 14. This deletion error, along with the remaining illegal codewords, is resolved by shifting right by BS1 from the 14th bit. In this case, nine illegal codewords are legalized with only two correction operations, resulting in an error-free decoded DNA sequence.

[0460] In fact, there are many possible combinations of operations to decode the sequence. Furthermore, the number of combinations increases exponentially with the read length, making it impossible in practice to obtain the optimal sequence by counting all possible combinations.

[0461] Therefore, dynamic programming is used to determine the globally optimal decoding sequence. The codeword space is constructed as a 3D matrix, with three Bitwise Bars (BSs) as its axes. Each node (i, j, k) represents a codeword composed of the i-th bit of BS1, the j-th bit of BS2, and the k-th bit of BS3. It can be classified into two types based on parity checks: valid or invalid. Figure 9 C). Any path starting from node (1, 1, 1) and passing only through qualified nodes represents a possible decoded DNA sequence. The probability of a given path in the codeword space can be calculated using Bayes' theorem. The prior probability of a DPL of length n is 1 / 2. n ( Figure 8 B), the probability P(m|n) of a DPL of length n sequenced at length m can be obtained from the reference sequence, and the data can be compared with the theoretical value. Figure 8 C). Then for the r-th round (r is MK, RY, or WS), its length is m. i The DPL of the i-th measurement is composed of length n i The posterior probability P generated by DPL r (n i |m i The following can be given:

[0462]

[0463] The probability P of the measured DPL array being generated by a specific DNA r It is P r (n i |m iThe cumulative results of the three rounds of ECC sequencing. Under the assumption that the three rounds of ECC sequencing are independent of each other, the probability of a given path is...

[0464] P Path =P MK ·P RY ·P WS

[0465] The probability of each path in the codeword space can be calculated in the same way. Figure 9 C). Use dynamic programming to find the path with the highest probability.

[0466] Decoding improves ECC sequencing accuracy

[0467] ECC decoding can effectively correct errors in long sequencing reads. Fourteen relatively long, three-round ECC experiments were performed with sequences from three different templates derived from λ phage. Before ECC decoding, occasional small errors were present in the sequencing signal. After decoding, these errors were completely eliminated up to 200 bp and significantly reduced in the 200-250 bp range. Figure 10 AC). For example in Figure 10 In A, although the first sequencing error occurred at base 39 in round RY, this error, along with several other sequencing errors in round WS, was successfully corrected after ECC decoding. The first error after ECC decoding was delayed by more than 270 bp.

[0468] The ECC decoding algorithm can accurately identify complex error patterns. Compared to scattered sequencing errors, adjacent errors in the same or different rounds are more challenging to correct because they require more numerous and finer-grained correction operations in the decoding algorithm. When parity checks between three rounds of sequencing signals fail, the algorithm calculates the probabilities of different operations.

[0469] In one scenario, two sequencing errors occur within the three cycles of round RY (a base deletion in cycle 22 and a base insertion in cycle 24). At least two alternative correction pathways, each involving two correction operations, can correct these errors. Figure 10 B). The first method performs 1-to-2 insertion correction and 2-to-1 deletion correction (p(2|1)*p(1|2) = 0.00015), while the second method contains 1-to-2 insertion correction and 3-to-2 deletion correction (p(2|1)*p(3|2) = 0.00022). Therefore, the second method is preferred due to its higher probability.

[0470] In another case, two adjacent long DPL sequencing errors occur in rounds MK and RY, respectively. Clearly, a left shift of one base in round WS can also restore parity validity. Figure 10C). However, because longer DPLs are more error-prone, the algorithm prefers to correct two longer DPLs by comparing the probabilities of different methods, rather than the shorter one.

[0471] Fluorescent degenerate sequencing is inherently highly accurate. Error frequencies at different DPLs per 50 nt were analyzed along the sequencing reads. Figure 10 D). Without ECC correction, 106 errors were found in 11,062 bases. Similar to other sequencing methods, these errors are more likely to occur at longer DPLs and later positions. See Forgetta et al. (2013) Journal of Biomolecular Techniques, 24(1), 3949; and Loman et al. (2012) Nature Biotechnology, 30(5), 4349. The raw precision was 99.82% in the first 100 nt and 99.45% in the first 200 nt. Read lengths exceeding 250 nt were achieved at the 99% accuracy cutoff.

[0472] ECC decoding eliminates most sequencing errors. The high raw accuracy of the fluorescence degenerate sequencing method forms the basis for ECC correction to completely eliminate all errors in the first 200 nt (including errors in DPL up to 9 nt), with an estimated upper bound error rate as low as 0.034%. Furthermore, ECC decoding effectively reduces the cumulative error rate over 250 nt from 0.96% to 0.33%.

Claims

1. A method for obtaining sequence information of a target polynucleotide, the method comprising: a) Providing a first sequencing reagent to a target polynucleotide in the presence of a first polynucleotide replication catalyst, wherein the first sequencing reagent comprises two different nucleotide monomers each conjugated to a first label, and the nucleotide monomer / first label conjugate is non-fluorescent until the nucleotide monomer is incorporated into the target polynucleotide according to its complementarity with the target polynucleotide; and b) A second sequencing reagent is provided to the target polynucleotide in the presence of a second polynucleotide replication catalyst, wherein the second sequencing reagent comprises two different nucleotide monomers each conjugated to a second label, and the nucleotide monomer / second label conjugate is non-fluorescent until the nucleotide monomer is incorporated into the target polynucleotide according to its complementarity with the target polynucleotide; and wherein the second sequencing reagent is provided subsequently after the first sequencing reagent has been provided. c) By detecting the fluorescence emission caused by the first and second labels after incorporating the nucleotide monomer into the polynucleotide in steps a) and b), the sequence information of at least a portion of the target polynucleotide is obtained. The nucleotide monomers in the first sequencing reagent and the second sequencing reagent are selected from the group consisting of the following combinations: 1) Adenine (A) nucleotide monomer and thymine (T) / uracil (U) nucleotide monomer in one sequencing reagent and cytosine (C) nucleotide monomer and guanine (G) nucleotide monomer in another sequencing reagent; 2) Adenine (A) nucleotide monomer and guanine (G) nucleotide monomer in one sequencing reagent and cytosine (C) nucleotide monomer and thymine (T) / uracil (U) nucleotide monomer in another sequencing reagent; and 3) Adenine (A) nucleotide monomer and cytosine (C) nucleotide monomer in one sequencing reagent and guanine (G) nucleotide monomer and thymine (T) / uracil (U) nucleotide monomer in another sequencing reagent.

2. The method according to claim 1, wherein the first label of the two different nucleotide monomers in step a) and the second label of the two different nucleotide monomers in step b) are the same label.

3. The method of claim 1, wherein the first mark comprises two different marks, and wherein one of the first marks is identical to one of the second marks, and the other of the first marks is identical to the other of the second marks.

4. The method according to any one of claims 1-3, wherein multiple rounds of steps a) and b) are performed, each round using a combination selected from the combinations 1)-3).

5. The method according to claim 4, wherein at least two or three sets of sequence information are obtained in step c), the method comprising: Combination 1) involves multiple rounds of steps a) and b) in the first sequencing reaction volume to obtain the first set of sequence information. Using combination 2), multiple rounds of steps a) and b) are performed in the second sequencing reaction volume to obtain a second set of sequence information, and / or Using combination 3), multiple rounds of steps a) and b) are performed in the third sequencing reaction volume to obtain a third set of sequence information.

6. The method of claim 5, wherein the first, second, and third sets of sequence information are obtained in parallel from separate sequencing reaction volumes.

7. The method according to claim 5, wherein the first, second and third sets of sequence information are obtained successively from the same sequencing reaction volume, wherein the product of the previous sequencing reaction is removed before starting the next sequencing reaction.

8. The method according to any one of claims 5-7, further comprising comparing the at least two or three sets of sequence information to reduce or eliminate the sequence errors.

9. The method of claim 8, wherein when the at least two or three sets of sequence information are consistent with each other, the comparison indicates that there are no errors in the obtained target polynucleotide sequence.

10. The method of claim 8, wherein the comparison indicates an error in the obtained target polynucleotide sequence when the at least two or three sets of sequence information contain differences at at least one nucleotide residue of the target polynucleotide sequence.

11. The method of claim 10, further comprising correcting at least one nucleotide residue in the obtained target polynucleotide sequence such that, after correction, the at least two or three sets of sequence information are consistent with each other.

12. A method for obtaining sequence information of a target polynucleotide, the method comprising: a) Providing a first sequencing reagent to a target polynucleotide in the presence of a first polynucleotide replication catalyst, wherein the first sequencing reagent comprises three different nucleotide monomers each conjugated to a first label, and the nucleotide monomer / first label conjugate is non-fluorescent until the nucleotide monomer is incorporated into the target polynucleotide according to its complementarity with the target polynucleotide; and b) Providing a second sequencing reagent to the target polynucleotide in the presence of a second polynucleotide replication catalyst, wherein the second sequencing reagent comprises a nucleotide monomer conjugated to a second label, and the nucleotide monomer / second label conjugate is non-fluorescent until after the nucleotide monomer is incorporated into the target polynucleotide according to its complementarity with the target polynucleotide; and wherein the second sequencing reagent is provided before or after the provision of the first sequencing reagent, and c) By detecting the fluorescence emission caused by the first and second labels after incorporating the nucleotide monomer into the polynucleotide in steps a) and b), the sequence information of at least a portion of the target polynucleotide is obtained. The nucleotide monomers in the first sequencing reagent and the second sequencing reagent are selected from the group consisting of the following combinations: 1) A cytosine (C) nucleotide monomer, a guanine (G) nucleotide monomer and a thymine (T) / uracil (U) nucleotide monomer in one sequencing reagent, and an adenine (A) nucleotide monomer in another sequencing reagent; 2) A sequencing reagent containing adenine (A) nucleotide monomers, guanine (G) nucleotide monomers, and thymine (T) / uracil (U) nucleotide monomers, and a sequencing reagent containing cytosine (C) nucleotide monomers; and 3) Adenine (A) nucleotide monomer, cytosine (C) nucleotide monomer, and thymine (T) / uracil (U) nucleotide monomer in one sequencing reagent, and guanine (G) nucleotide monomer in another sequencing reagent; and 4) A sequencing reagent containing adenine (A) nucleotide monomers, cytosine (C) nucleotide monomers and guanine (G) nucleotide monomers, and another sequencing reagent containing thymine (T) / uracil (U) nucleotide monomers.

13. The method according to claim 12, wherein the first label of the three different nucleotide monomers in step a) and the second label of the one nucleotide monomer in step b) are the same label.

14. The method according to claim 12 or 13, wherein multiple rounds of steps a) and b) are performed, each round using a combination selected from combinations 1)-4).

15. The method of claim 14, wherein at least two, three, or four sets of sequence information are obtained in step c), the method comprising: Combination 1) involves multiple rounds of steps a) and b) in the first sequencing reaction volume to obtain the first set of sequence information. Combination 2) involves performing multiple rounds of steps a) and b) in the second sequencing reaction volume to obtain a second set of sequence information. Using combination 3), multiple rounds of steps a) and b) are performed in the third sequencing reaction volume to obtain a third set of sequence information, and / or Using combination 4), multiple rounds of steps a) and b) are performed in the fourth sequencing reaction volume to obtain a fourth set of sequence information.

16. The method of claim 15, wherein the first, second, third, and fourth sets of sequence information are obtained in parallel from separate sequencing reaction volumes.

17. The method of claim 15, wherein the first, second, third and fourth sets of sequence information are obtained successively from the same sequencing reaction volume, wherein the product of the previous sequencing reaction is removed before starting the next sequencing reaction.

18. The method according to any one of claims 15-17, further comprising comparing the at least two, three, or four sets of sequence information to reduce or eliminate sequence errors.

19. The method of claim 18, wherein when the at least two, three, or four sets of sequence information are consistent with each other, the comparison indicates that there are no errors in the obtained target polynucleotide sequence.

20. The method of claim 18, wherein the comparison indicates an error in the obtained target polynucleotide sequence when the at least two, three, or four sets of sequence information contain differences at at least one nucleotide residue of the target polynucleotide sequence.

21. The method of claim 20, further comprising correcting at least one nucleotide residue in the obtained target polynucleotide sequence such that, after correction, the at least two, three, or four sets of sequence information are consistent with each other, wherein the at least one nucleotide residue in one set of sequence information is corrected by insertion or deletion at the position where the error occurred, wherein the insertion extends the sequence by at least one nucleotide and the sequence information from another set or more sets of sequence information is compared with the extended sequence to achieve a corrected sequence, wherein the deletion shortens the sequence by at least one nucleotide and the sequence information from one or more sets of sequence information is compared with the shortened sequence to achieve a corrected sequence.

Images