CN116246703A - A quality assessment method for nucleic acid sequencing data - Google Patents

Abstract

The invention discloses a method for evaluating the quality of nucleic acid sequencing data, which uses the polymer in the nucleic acid sequence as the basic unit for quality evaluation instead of the base as the basic unit in the existing method, and is more suitable for 3' Sequences obtained by open-ended sequencing.

Description

A quality assessment method for nucleic acid sequencing data

technical field

The technology disclosed in the present invention relates to a method for evaluating the quality of nucleic acid sequencing data, which belongs to the field of gene sequencing.

Background technique

Gene sequencing technology can ascertain the sequence of genetic material, and is widely used in clinical tumor typing, microbial identification, and genetic disease diagnosis. In addition to producing the sequence of the tested nucleic acid sample, today's mainstream nucleic acid sequencing technology also assigns a quality value to each base to evaluate its accuracy. This quality value is generally expressed in the form of Phred:

q=-10log ₁₀ (1-a)

In the formula, a is the accuracy rate of the base, and q is the Phred value. For example, the Phred values corresponding to accuracy rates of 99%, 99.9%, and 99.99% are 20, 30, and 40, respectively.

In the bioinformatics analysis of nucleic acid sequencing data, the quality value plays a very important role. For example, when identifying a gene mutation, if a base on the measured sequence is different from the corresponding base on the reference sequence, when the quality value of the base is high, it will be judged as a gene mutation; and when When the quality value of the base is low, the sequence is considered to have a sequencing error and there is no genetic mutation.

For 454, Ion Torrent, and fluorescence sequencing, which are 3'-end open sequencing technologies, the reaction is carried out polymer by polymer, and each polymer is a whole in terms of sequencing chemistry. The existing method of assigning a quality value to each base is a remnant of the Sanger sequencing era, which is just suitable for Illumina's sequencing chemistry, so it is popular in the world, but it is actually not suitable for the 3' end open Sequencing technology has many shortcomings. For example, many sequencing technologies are prone to insertion or deletion errors when detecting longer homologous polymers, such as AAAAA is detected as AAAAA or AAA. This error occurs on homopolymers, and it is difficult to accurately evaluate the quality value of each base on the homopolymers. In some cases, some bases are not prone to substitution errors, so they should be retained when identifying single-base substitution mutations, but these bases often have low quality values because they are prone to insertion or deletion errors. Discarded when identifying mutations, resulting in false negatives. Therefore, there is a need for a quality assessment method that is more suitable for 3' open sequencing sequences.

Contents of the invention

The invention discloses a method for evaluating the quality of nucleic acid sequencing data, which uses the polymer in the nucleic acid sequence as the basic unit for quality evaluation instead of the base as the basic unit in the existing method, and is more suitable for 3' Sequences obtained by open-ended sequencing.

Specifically, the present invention provides a method for evaluating the quality of nucleic acid sequencing data, characterized in that it includes:

Provide the nucleic acid sequence to be tested, and use the polymer as the basic unit to calculate the sequencing signal characteristics of the polymer;

predicting a quality score for the polymer based on the sequencing signal characteristics using a training-calibrated quantization scheme;

The quantization scheme of the training calibration includes:

For the standard nucleic acid sequence provided, the polymer is used as the basic unit to calculate the sequencing signal characteristics of the polymer, and the polymer is marked as sequenced correctly or incorrectly according to the comparison result of the standard nucleic acid sequence; the classifier is trained to fit multiple The relationship between the sequencing signature of a polymer and its label.

According to a preferred embodiment, the polymer includes homopolymers, binary copolymers, terpolymers, and the like.

According to a preferred embodiment, the sequencing signal characteristics of a polymer refer to the characteristics of the signal generated when the polymer undergoes a sequencing chemical reaction during the sequencing process, including but not limited to, the type of bases that make up the polymer, The length of the polymer, the number of rounds of the sequencing chemical reaction, the signal strength, the degree to which the signal strength (and its adjacent signal strength) is close to an integer, the parameters of the sequencing signal (unit signal, background signal, lead coefficient, lag coefficient, decay coefficient ), the degree of dephasing when the polymer is measured, etc.

According to a preferred embodiment, fitting the relationship between the sequencing signal features of the polymer and its markers includes converting the fitting result of the classifier into a quality score.

According to a preferred embodiment, the standard nucleic acid sample refers to a nucleic acid sample whose source and sequence have been determined and is highly homozygous at almost all sites in the genome, including bacteriophage lambda DNA, Escherichia coli DNA, Saccharomyces cerevisiae DNA, and the like.

According to a preferred embodiment, the quality score refers to a numerical value characterizing the accuracy of polymer sequencing, which is selected from accuracy rate, error rate, Phred value and the like.

According to a preferred embodiment, the quality score is logarithmically based on the polymer call error probability, and wherein said quality score comprises Q10, Q15, Q20, Q25, Q30, Q35, Q40, Q45, Q50, Q55, Q60, etc.

According to a preferred embodiment, the classifier includes, but is not limited to, linear regression, polynomial regression, logistic regression, support vector machine, artificial neural network, random forest, Phred algorithm, ensemble learning and the like.

According to a preferred embodiment, the training of the classifier includes classifying the polymers into several classes according to the characteristics of the sequencing signals of the polymers, and counting the sequencing accuracy of each class of polymers.

According to a preferred embodiment, the training classifier is a probability distribution model based on maximum likelihood; probability distribution refers to a probability distribution with a unimodal shape feature, including but not limited to two-point distribution, binomial distribution, negative binomial distribution , Poisson distribution, geometric distribution, exponential distribution, normal distribution, Γ distribution, chi-square distribution, t distribution, F distribution, β distribution, lognormal distribution, and high-dimensional extensions of the above distributions, etc.

According to a preferred embodiment, the method further includes performing bioinformatics analysis on the nucleic acid sequence to be tested according to the quality score of the polymer.

According to a preferred embodiment, the bioinformatics analysis comprises, according to the assigned quality value, screening for high-quality nucleic acid sequences. Screening methods include, but are not limited to, screening nucleic acid sequences whose quality values are all higher or lower than a certain threshold, screening nucleic acid sequences whose average value of all quality values is higher than or lower than a certain threshold, and screening nucleic acid sequences whose quality values are all higher than or lower than a certain threshold. Regions above or below a certain threshold value, regions in which the average value of quality values in the nucleic acid sequence are both above or below a certain threshold value, and so on.

According to a preferred embodiment, the bioinformatics analysis comprises, according to the assigned quality value, the alignment of the nucleic acid sequence to a reference sequence.

According to a preferred embodiment, the bioinformatics analysis includes identifying genetic variations according to the alignment results and the quality values assigned to the aligned sequences.

According to a preferred embodiment, when identifying a gene variation, some characteristics of the comparison result can be used to remove potential false positive or false negative results.

According to a preferred embodiment, the bioinformatics analysis includes assembling the nucleic acid sequences into longer nucleic acid sequences according to the assigned quality values.

According to a preferred embodiment, the bioinformatics analysis includes performing at least two rounds of orthogonal degenerate sequencing, and after obtaining the mass value of the length of the degenerate polymer, using the mass value to perform error correction.

The present invention also provides a method for evaluating the quality of nucleic acid sequencing data, which is characterized in that it includes: performing fuzzy sequencing or missing sequencing on the nucleic acid sample to be tested to obtain input data, generating degenerate polymer length information of the input data, and calculating The sequencing signal characteristics of the length of the degenerate polymer;

predicting a quality score for the polymer based on the sequencing signal characteristics using a quantization scheme calibrated for training;

The quantization scheme of the training calibration includes:

Sequencing a standard nucleic acid sample to obtain a nucleic acid sequence, calculating the sequencing signal characteristics of the nucleic acid sequence, comparing the nucleic acid sequence to a reference sequence, and marking the polymer of the nucleic acid sequence as being sequenced correctly or incorrectly; training a classifier, fitting Relationship between sequencing signature features of polymers and their markers.

Benefits of the invention

Compared with the prior art, the method of the present invention has the following advantages:

This method uses polymers as the basic unit of sequencing quality assessment, and assigns a quality value to the polymers that make up the nucleic acid sequence instead of assigning a quality value to the bases. It is especially suitable for sequencing reactions that are open at the 3' end and helps Obtain more accurate bioinformatics analysis results, including identification of variants and more.

Description of drawings

The novel features of the invention are set forth with particularity in the appended claims. A better understanding of the features and advantages of the present invention will be obtained by reference to the following detailed description, which illustrates illustrative embodiments utilizing the principles of the invention, and the accompanying drawings, in which:

Figure 1. Illustrates an example of a sequencing signal signature.

Detailed ways

In order to further illustrate the core content of the present invention, the present invention is now illustrated with the following examples. The examples are for further explaining the summary of the invention, and do not limit the present invention.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. In order to better disclose the method and content of the present invention, the key terms in the present invention are explained in detail here.

Terminology Explanation

each

The term "each" is intended to identify a single item in the collection, but does not necessarily refer to every item in the collection. Exceptions may apply where expressly disclosed or where the context clearly dictates otherwise.

include

The term "comprising" is intended herein to be open-ended, encompassing not only the listed elements, but also any additional elements.

sample

The term "sample" in the present invention refers to a sample, usually derived from a biological fluid, cell, tissue, organ or organism, comprising a nucleic acid or a nucleic acid mixture comprising at least one nucleic acid to be sequenced and/or phased sequence. Such samples include, but are not limited to, blood, blood fractions, sputum/oral fluid, amniotic fluid, fine needle biopsy samples (eg, surgical biopsy, fine needle biopsy, etc.), urine, peritoneal fluid, pleural fluid, tissue explants, Organ cultures and any other tissue or cell preparations, or fractions or derivatives thereof, or fractions or derivatives isolated therefrom. While samples are typically taken from human subjects (eg, patients), samples can be taken from any organism that has a chromosome, including but not limited to bacteria, viruses, fungi, birds, mammals, and the like. Samples may be used as obtained from the biological source or after pretreatment to alter the properties of the sample. For example, such pretreatment may include preparation of plasma from blood, dilution of viscous fluids, and the like. Methods of pretreatment may also involve, but are not limited to, filtration, precipitation, dilution, distillation, mixing, centrifugation, freezing, lyophilization, concentration, amplification, nucleic acid fragmentation, inactivation of interfering components, addition of reagents, lysis, and the like.

standard nucleic acid sample

Refers to nucleic acid samples whose source and sequence have been determined and are highly homozygous at almost all loci in the genome. The advantage of using standard nucleic acid sample sequencing is that it can be more accurately determined whether the sequence in the sequencing result is different from the reference genome is a variation or a sequencing error. For example, the standard nucleic acid sample can be nucleic acid of Escherichia coli, Saccharomyces cerevisiae, etc., or lambda phage DNA produced by New England Biolabs, etc.

Error Correction Code (Error correcting Code, ECC) sequencing and error correction

In this application, error correction code sequencing has the following characteristics:

This sequencing method requires multiple rounds of sequencing, and the information obtained by each round of sequencing is incomplete, while the total information obtained by multiple rounds of sequencing is redundant; the information redundancy of multiple rounds of sequencing is used to detect and correct potential sequencing errors, and high sequence of accuracy. For example, taking 2+2 sequencing as an example, the sequencing reagents are divided into three groups with pairwise matching (for example, three groups of MK, RY, and WS) according to the paired bases, and the DNA sequence to be tested is subjected to three independent Sequencing, and then generate three degenerate sequence codes, these three codes can be mutually verified, and then not only can deduce the real base sequence information through decoding, but also have the ability to correct single-pass sequencing error sites. This correction process is error correction correction.

sequence

In the present invention, "sequence" is a nucleic acid sequence, including or representing nucleotide chains coupled to each other, which may be a definite nucleotide sequence or a degenerate base sequence, and may be based on DNA or RNA. It should be understood that a sequence may include multiple subsequences. For example, a single sequence (eg, the sequence of a PCR amplicon) can have 350 nucleotides. A sample read may include multiple subsequences within these 350 nucleotides. The sequence is not divided by a single base as the basic unit, but by a polymer as the basic unit, and the length of the polymer can be 1bp or longer.

degenerate base

According to the IUPAC symbol nomenclature (Nucleic acid notation), the letters in Table 1 below are used to represent degenerate bases, for example, the letter M represents A and/or C; the degenerate polymer MMKKK has a length of 5, that is, DPL is 5.

Table 1

letter base represented by m A/C K G/T R A/G Y C/T W A/T S C/G B C/G/T D. A/G/T h A/C/T V A/C/G

reference sequence

Reference sequence refers to any specific known genomic sequence, whether partial or complete, of any organism that can be used to reference an identified sequence from a subject. For example, reference genomes for human subjects, as well as many other organisms, can be found at NCBI, the National Center for Biotechnology Information at ncbi.nlm.nih.gov. "Genome" refers to the complete genetic information of an organism or virus expressed in nucleic acid sequences. The genome includes both genes and the non-coding sequence of DNA. A reference sequence can be larger than the reads it is aligned to. For example, the reference sequence can be at least about 100 times larger, or at least about 1000 times larger, or at least about 10,000 times larger, or at least about ¹⁰⁵ times larger, or at least about ¹⁰⁶ times larger, or at least about 10 ⁷ times larger. In one example, the reference genome sequence is the sequence of the full length human genome. In another example, the reference genome sequence is restricted to a specific human chromosome, such as chromosome 13. In some implementations, the reference chromosome is a chromosomal sequence from the human genome version hg19. Such sequences may be referred to as chromosomal reference sequences, but the term reference genome is intended to encompass such sequences. Other examples of reference sequences include genomes of other species, and chromosomes, subchromosomal regions (such as strands), etc. of any species. In various implementations, a reference genome is a consensus sequence or other combination derived from multiple individuals. However, in some applications, a reference sequence may be taken from a particular individual. In other embodiments, "genome" also encompasses so-called "graph genomes", which use specific storage formats and representations of genome sequences. In one implementation, Graph Genome stores data in linear files.

Comparison (align or alignment)

Alignment is a common concept in bioinformatics. In bioinformatics, alignment is often used to compare the similarity between different nucleic acids or between different proteins. Alignment in the present invention refers to the process of comparing a nucleic acid sequence with a reference sequence to determine whether the reference sequence contains a nucleic acid sequence. Commonly used sequence alignment algorithms and software include, but are not limited to, for example, Smith-Waterman algorithm, Bowtie, BWA, SOAP, Needleman-Wunch algorithm, Bowtie2, BLAST, ELAND, TMAP, MAQ, minimap2, SHRiMP, etc.

quality score

Or quality value, which refers to a numerical value that characterizes the accuracy of sequencing. Quality values can be expressed in different mathematical ways, such as accuracy rate, error rate, Phred value, etc. For example, the accuracy rates of 99%, 99.9%, and 99.99% correspond to error rates of 1%, 0.1%, and 0.01%, respectively, and the corresponding Phred values are 20, 30, and 40, respectively. In some implementations, in order to facilitate recording and storage, the Phred value will be converted into ASCII code after adding 33, for example, Phred value 20, 30, 40 will be converted into the characters '5', '? ’, ‘I’. The difference in the expression form of the quality value does not affect the essence of the present invention.

fuzzy sequencing

In the present invention, fuzzy sequencing refers to single-round (or single-round) 2+2 sequencing, or single-round 1+3 sequencing, or single-round 3×4 sequencing. Specifically, 2+2 sequencing refers to, in the sequencing, including two different sequencing reagents: the first sequencing reagent and the second sequencing reagent; the two sequencing reagents are added in a cycle; wherein the first sequencing reagent contains two different nucleomonomers; the second sequencing reagent comprises two different nucleomonomers with a detectable label, and said nucleomonomers are different from the nucleosides present in the first sequencing reagent acid monomer, and wherein the second sequencing reagent is provided after the first sequencing reagent is provided, after the nucleotide monomer is incorporated into the nucleic acid to be detected, the signal generated by the detectable label is detected, for example, a fluorescent signal; at 2 In +2 sequencing, a combination of fluorescently labeled nucleotides can be used to obtain the fluorescent signal value associated with the target DNA sequence. Examples of possible combinations are as follows: M/K mode: where odd-numbered rounds present dA4P and dC4P, where even-numbered rounds present dG4P and dT4P, or vice versa; R/Y mode: where odd-numbered rounds present dA4P and dG4P, where even-numbered rounds present dA4P and dG4P and the W/S pattern: where odd-numbered rounds present dA4P and dT4P, where even-numbered rounds present dC4P and dG4P; or vice versa. Using one of the above modes (for example, M/K mode) to perform multiple cycles of sequencing on the nucleic acid to be tested may be referred to as one round (or a single round). 1+3 sequencing is similar to 2+2 sequencing, which means that in sequencing, two different sequencing reagents are included: the first sequencing reagent and the second sequencing reagent; the two sequencing reagents are added in a cycle; the first sequencing reagent contains Three different nucleotide monomers with a detectable label; the second sequencing reagent contains one nucleotide monomer with a detectable label, and the nucleotide monomer is different from that present in the first sequencing reagent Nucleotide monomers, and wherein the second sequencing reagent is provided after the first sequencing reagent is provided, after the nucleotide monomers are incorporated into the nucleic acid to be detected, a signal generated by a detectable label is detected, for example, a fluorescent signal; 3x4 sequencing means that in sequencing, four different sequencing reagents are included: the first sequencing reagent, the second sequencing reagent, the third sequencing reagent and the fourth sequencing reagent; the four sequencing reagents are added in cycles; each of the sequencing reagents Each comprises three different nucleomonomers with detectable labels; and wherein the second sequencing reagent is provided subsequently to the first sequencing reagent, and detectable after the nucleomonomers are incorporated into the nucleic acid to be tested The signal generated by the label; the third sequencing reagent is provided after the second sequencing reagent is provided, and the signal generated by the detectable label is detected after the nucleotide monomer is incorporated into the nucleic acid to be tested; the fourth sequencing reagent is provided after the second sequencing reagent is provided The three sequencing reagents are then provided to detect the signal generated by the detectable label after the nucleotide monomer is incorporated into the nucleic acid to be tested; for example, the above four sequencing reagents are B, D, H, and V respectively.

The fuzzy sequence information is obtained according to the signal, and the fuzzy sequence information refers to the base sequence information that cannot be obtained from the sequence information to determine the nucleotide sequence. It is the encoded nucleic acid sequence information, or the nucleic acid sequence information encoded by A, G, U, C; wherein the base may be a methylated base. Fuzzy base sequence is a common concept in the field of scientific research, such as using the letter W to represent base A and/or T. There are also related definitions on WIKIPEDIA (https://en.wikipedia.org/wiki/Nucleotide).

Deletion Sequencing vs. Deletion Sequencing

Perform multiple sequencing chemical reaction cycles on the nucleic acid molecule to be tested on the gene sequencing chip; wherein, at least one cycle is pre-selected, and signal acquisition is only performed in the aforementioned cycle; only sequencing chemical reactions are performed in other cycles, and no signal acquisition is performed; The collected signal is coded as a sequence to obtain the missing nucleic acid sequence. This process is called deletion sequencing.

Missing sequence, that is, a sequence with missing information. The missing information here means that part of the sequence information in the obtained sequence is missing, for example: sequence ATTCGNNTTT, this sequence is the missing sequence, and N means that the sequence information is unknown, that is, the sequence information here is missing, the sequence is a missing sequence.

polybase sequencing

Multi-base sequencing, that is, a sequencing reaction with an open 3' end. In this sequencing reaction, the 3' end of the nucleotide molecule as a substrate is a hydroxyl group, which can be freely extended. In theory, one sequencing reaction can extend one or multiple nucleotide molecules. Common polynucleotide sequencing includes Ion Torrent's semiconductor sequencing, 454's pyrosequencing, and Senna's fuzzy sequencing, etc.

unit signal

The unit signal is the rising value of the signal detected by the sequencer when the DNA is extended by one base, and it is related to the number of DNA molecules that undergo the extension reaction, the exposure time of the camera, the intensity of the excitation light, and the light sensitivity of the camera. The unit signal refers to the signal intensity of one base extension at each sequencing site, which is a physical quantity proportional to the number of template DNA molecules at the sequencing site, and the intensity of the sequencing signal is equal to the unit signal multiplied by the number of bases in the extension reaction. Regardless of any sequencing technology, the measurement accuracy of unit 1 directly affects the accuracy of sequencing results.

background signal

The background signal refers to the reference signal detected by the sequencer when a base is extended, and is related to factors such as the chip material and the spontaneous hydrolysis of the sequencing reaction substrate. And the background signal can also change with the extension of the sequencing read length. Background signal belongs to the general definition.

Degenerate Polymer Length (DPL)

Degenerate sequencing is a kind of multi-base sequencing. Different from single-base sequencing, each round of reaction only extends one nucleotide molecule. Multi-base sequencing may extend multiple nucleotides in each round of reaction. The fluorescence released by the sequencing reaction The signal intensity is positively correlated with the number of released fluorophores. Under ideal conditions without attenuation and dephasing, the fluorescent signal of each round of reaction reflects the number of bases extended in this round, which is called the length of the degenerate polymer ( degenerate polymer length, DPL).

wheel (cycle)

A round of sequencing reaction, that is, a sequencing cycle, or a sequencing reaction round, a round of sequencing reaction refers to the provision of a sequencing reagent, which incorporates nucleotides with detectable labels in the sequencing reagent into the nucleic acid to be tested The process of detecting the signal generated by the detectable marker thereafter.

polymer

The polymers described in the present invention include homologous polymers, binary copolymers, and terpolymers, and the length of the polymers can be 1 bp or longer.

homopolymer

The homologous multimer or homologous multimer in the present invention refers to a polymer composed of multiple homologous nucleotide monomers, for example, AAAA or TTTTT, etc. all belong to homologous multimers .

binary copolymer

Binary copolymer is a polymer with two different monomer chains formed by the polymerization of two different monomers. In the present invention, it specifically refers to a polymer composed of two nucleotide monomers, such as , when performing MK degenerate sequencing, M is composed of two nucleotide monomers, A and C, K is composed of two nucleotide monomers, G and T, and both ACAC and GGTT are binary copolymers.

Terpolymer

Similar to binary copolymers, terpolymers are polymers composed of three different monomers at the same time. In the present invention, it specifically refers to polymers composed of three nucleotide monomers, such as , when performing AB degenerate sequencing, B may be composed of three nucleotide monomers C, G, and T, and both TCGG and GTGCC are terpolymers.

Classifier

Classifier Classification is a very important method of data mining. In machine learning, the role of a classifier is to judge the category of a new observation sample on the basis of the labeled training data.

The construction and implementation of classifier generally need to go through the following steps:

Select samples (including positive samples and negative samples), and divide all samples into two parts: training samples and test samples;

Execute the classifier algorithm on the training samples to generate a classification model;

Execute the classification model on the test samples to generate predictions.

Preferably, according to the prediction results, necessary evaluation indicators are calculated to evaluate the performance of the classification model.

genetic mutation

refers to a nucleic acid sequence that differs from a reference sequence. Typical variations include, but are not limited to, single nucleotide variation (SNV), short deletion and insertion polymorphism (Indel), copy number variation (CNV), epigenetic variation, microsatellite marker or short tandem repeat sequence and structure Mutations. Somatic variant calling is the effort to identify variants that are present at low frequency in a DNA sample. Somatic variant calling is of great interest in the context of cancer therapy. Cancer is caused by the accumulation of mutations in DNA. DNA samples from tumors are often heterogeneous, including some normal cells, some cells early in cancer progression (with fewer mutations), and some late-stage cells (with more mutations). Because of this heterogeneity, somatic mutations will often appear at low frequency when sequencing tumors (eg, from FFPE samples). For example, SNVs may be seen in only 10% of reads covering a given base. A variation to be classified as somatic or germline by a variation classifier is also referred to herein as a "genetic variation".

threshold

Refers herein to a numeric or non-numeric value used as a cutoff for characterizing a sample, a nucleic acid, or a portion thereof (eg, a read). Thresholds may vary based on empirical analysis. Threshold values can be compared to measured or calculated values to determine whether sources producing such values should be classified in a particular way. Thresholds can be identified empirically or analytically. The choice of threshold depends on the confidence level with which the user wishes to have to classify. Thresholds can be chosen for a particular purpose (eg, to balance sensitivity and selectivity). As used herein, the term "threshold" indicates a point at which an analysis process may be altered and/or an action may be triggered. The threshold need not be a predetermined amount. Instead, the threshold may be a function based on a number of factors, for example. The threshold can be adjusted according to the situation. Additionally, a threshold may indicate an upper limit, a lower limit, or a range between limits.

The general meanings of the terms involved in the present invention are indicated above. The above-mentioned terms all have conventional meanings in this field, and are explained again in order to avoid ambiguity. The above terms have no special meaning.

Detailed description of the invention

Quality assessment of nucleic acid sequencing data plays a very important role in subsequent bioinformatics analysis. For example, when identifying a gene mutation, if a base has a high-quality value, and the base is different from the corresponding base on the reference sequence, it will be judged as a gene mutation; and when the quality value of the base When it is low, the sequence is considered to have a sequencing error and there is no gene mutation. In the prior art, the method of assigning a quality value to each base with a single base as a basic unit has many defects. For example, many sequencing technologies are prone to insertion or deletion errors when detecting longer homologous polymers, such as measuring TTTT as TTT or TTTTT. Since this error occurs on the homopolymer, it is difficult to accurately evaluate the mass value of each base on the homopolymer. For another example, in some cases, some bases are not prone to substitution errors, so they should be retained when identifying single-base substitution mutations, but these bases often have low quality values because they are prone to insertion or deletion errors, Instead, it is easy to be discarded when identifying mutations, resulting in false negatives. In order to overcome the defects in the prior art, the present invention discloses a new method for evaluating the quality of sequencing data, which does not use bases as the basic unit, but polymers as the basic unit of scoring, especially suitable for 3' end opening The sequencing reaction helps to obtain more accurate bioinformatics analysis results.

Specifically, the present invention provides a method for evaluating the quality of nucleic acid sequencing data, characterized in that it includes:

Provide the nucleic acid sequence to be tested, and use the polymer as the basic unit to calculate the sequencing signal characteristics of the polymer;

Utilize the quantification scheme of training calibration, and based on the sequencing signal feature, predict the quality score of described polymer; The quantization scheme of described training calibration comprises:

For the standard nucleic acid sequence provided, the polymer is used as the basic unit to calculate the sequencing signal characteristics of the polymer, and the polymer is marked as sequenced correctly or incorrectly according to the comparison result of the standard nucleic acid sequence; the classifier is trained to fit multiple The relationship between the sequencing signature of a polymer and its label.

In the present invention, the sequencing methods for obtaining nucleic acid sequences include dideoxynucleotide termination method (Sanger sequencing method), chemical degradation method (Gilbert method), pyrosequencing method (pyrosequencing), semiconductor sequencing method (semiconductor sequencing), cycle reversible termination cyclic reversible terminator, fluorogenic sequencing, error-correction code sequencing, fuzzy sequencing, deletion sequencing (patent CN 2022101040373), combined probe anchoring Combinatorial probe-anchor ligation, combinatorial probe-anchor polymerization, sequencing by oligonucleotide ligation and detection, sequencing-by -binding), single-molecule fluorescence sequencing, single-molecule real-time sequencing, nanopore sequencing, etc.

According to a preferred embodiment, the sequencing method is preferably selected from a sequencing reaction with an open 3' end, including but not limited to, pyrosequencing, semiconductor sequencing, fluorescence sequencing, error-correcting code sequencing, fuzzy sequencing, deletion Sequencing method, nanopore sequencing method, etc. In the above-mentioned sequencing technologies, one or more nucleotide molecules can be extended in one sequencing chemical reaction cycle, that is, the reaction of each cycle occurs in units of polymers, and It does not occur in units of bases. Therefore, using polymers as the basic unit of quality assessment can undoubtedly more accurately reflect the real sequencing process, and the obtained sequencing quality assessment results are also more accurate.

In the present invention, nucleic acid samples include deoxyribonucleic acid (DNA), ribonucleic acid (RNA), peptide nucleic acid (PNA), xylose nucleic acid (XNA), locked nucleic acid (LNA) and the like.

In some embodiments, the nucleic acid sample to be tested and the standard nucleic acid sample can be labeled and then sequenced at the same time, using the same sequencing method, and corresponding nucleic acid sequences can be obtained respectively. In some embodiments, the standard nucleic acid sample can be sequenced first to obtain the corresponding nucleic acid sequence; then, the nucleic acid sample to be tested can be sequenced by the same sequencing method to obtain the corresponding nucleic acid sequence. In some embodiments, the nucleic acid sample to be tested can also be sequenced first to obtain the corresponding nucleic acid sequence; then, the standard nucleic acid sample can be sequenced to obtain the corresponding nucleic acid sequence. The sequencing order of the sample to be tested and the standard nucleic acid sample can be exchanged, and the important thing is that both need to use the same sequencing method for sequencing and base calling.

According to a preferred embodiment, the input data is image data derived from sequencing images generated by the sequencer during a sequencing run. For example, pyrosequencing, fluorogenic sequencing, error-correction code sequencing, cyclic reversible terminator, combined probe anchor ligation The input data generated by sequencing methods such as combinatorial probe-anchor ligation is image data.

In some embodiments, the input data is based on the pH change due to the release of hydrogen ions during the extension of the nucleotide substrate molecule, the pH change is detected and converted to a voltage proportional to the amount of incorporated nucleotide Variation, for example, is the case with the input data generated by Ion Torrent's semiconductor sequencing method.

In some embodiments, the input data is created from nanopore sensing, which uses a biosensor to measure the interruption of electrical current as an analyte passes through a nanopore or near its orifice, while determining the base's type. For example, Oxford Nanopore Technology (ONT) sequencing is based on the concept of passing single-stranded DNA (or RNA) through a membrane via a nanopore and applying a voltage difference across the membrane. The nucleotides present in the pore will affect the electrical resistance of the pore, so current measurements over time can indicate the sequence of DNA bases passing through the pore. This current signal (called a "squiggle" because of its appearance when plotted) is the raw data collected by the ONT sequencer. These measurements are stored as 16-bit integer data acquisition (DAC) values acquired at a frequency of, for example, 4kHz. At a DNA chain speed of about 450 base pairs/sec, this gave an average of about nine raw observations per base. This signal is then processed to identify breaks in the aperture signal that correspond to individual readings. These maximum utilizations of the raw signal are for base calling, the process of converting DAC values into DNA base sequences. In some implementations, the input data includes normalized or scaled DAC values.

In the present invention, polymers include homopolymers, binary copolymers, terpolymers, and the like. The measured nucleic acid sequence is divided into several polymers. When dividing, all polymers can be homologous polymers, binary copolymers, or terpolymers. A part may also be a homopolymer, a part may be a binary copolymer, and the other part may be a terpolymer, and the length of each polymer may be 1 bp or longer, for example, 5 bp, 10 bp or more than 10 bp.

The division of polymers should correspond to the sampling process in the sequencing process. For example, if the sequencing method is MK degenerate sequencing, the obtained nucleic acid sequence can be MMKKKMKMMM, which is a degenerate sequence, then when dividing polymers, you should Divide it according to the method corresponding to the actual extension of MK sequencing, that is: (MM)(KKK)(M)(K)(MMM); if the sequencing method is AB degenerate sequencing, the obtained sequence can be ABBBAAABB, then divide the multimer The body time should be divided according to the method corresponding to the actual extension of AB sequencing, that is: (A)(BBB)(AAA)(BB); if the sequencing method is 1x4, the obtained sequence can be ACCCTTGGATT, and the nucleic acid sequence is a definite nuclear sequence. nucleotide sequence, when dividing the polymer, it should be divided according to the method corresponding to the actual extension of 1x4 sequencing, with homopolymer as the basic unit, namely: (A)(CCC)(TT)(GG)(A)(TT).

In a preferred embodiment, the nucleic acid sequence is a definite nucleotide sequence, that is, the sequence represented by A, G, C, T, or the sequence represented by A, G, C, U.

In some embodiments, the nucleic acid sequence is a degenerate sequence, or ambiguous sequence, that is, the sequence contains uncertain information, represented by M, K, R, Y, W, S, B, D, H, V, etc. Degenerate bases, it can be understood that part of the sequence in the sequence may be determined.

In the present invention, a standard nucleic acid sample refers to a nucleic acid sample whose source and sequence have been determined and which are highly homozygous at almost all sites in the genome. For example, it may be lambda phage DNA produced by New England Biolabs. The reference sequence of the standard nucleic acid is known, the sequenced nucleic acid sequence can be compared to its corresponding reference nucleic acid sequence, and each polymer is marked as correct or incorrect, for example, a sequenced nucleic acid sequence For ATTGGCCAAAT, it is divided into 3 binary copolymers: (ATT)(GGCC)(AAAT), the reference sequence is ATTGGCCAAAA, then the first and second polymers are marked as correct, and the third polymer Marked as an error.

In a specific embodiment, the nucleic acid sequence obtained by sequencing is compared to its corresponding reference nucleic acid sequence to obtain the comparison result, and then the polymer is marked as sequenced correctly or sequenced incorrectly according to the comparison result; preferably, High-quality aligned sequences are further screened from the comparison results, and then the polymers in the high-quality aligned sequences are marked as sequenced correctly or sequenced incorrectly, and undetermined sequences are ignored (that is, sequences that cannot be successfully aligned to Bases on the reference sequence or bases with lower alignment quality). According to the alignment results, the polymers whose alignment result is "match" are marked as "sequencing correct", and the polymers whose alignment result is "mismatch", "insertion" or "deletion" are marked as "sequence error". ". The high-quality alignment described in the present invention needs to specifically select the range of quality values according to the comparison software or algorithm used; for example, when using BWA for sequence alignment, the sequence of high-quality alignment refers to the comparison For base sequences whose quality is greater than 0, or greater than or equal to 10, or greater than or equal to 20, or greater than or equal to 30, or greater than or equal to 40, or greater than or equal to 50, or greater than or equal to 60.

According to a preferred embodiment, the characteristics of the sequencing signal corresponding to the polymer refer to the characteristics of the signal generated when the polymer undergoes a sequencing chemical reaction during the sequencing process. Figure 1 shows examples of the characteristics of the sequencing signal, including but not Limited to: the type of the base, that is, which one the base belongs to A, G, C, T (or U); the position of the base in the sequence, that is, the position of the base in its nucleotide sequence Position, for example, for single-end sequencing, the sequencing quality value of the base at the front is usually higher than that of the base at the back; the length of the polymer where the base is located, that is, the homology of the base The number of bases in polymers or degenerate polymers, usually, the length of the polymer is short, and the sequencing quality value is high; the position of the base in the polymer, that is, the base and its position The distance from the nearest end of a homologous polymer or a degenerate polymer; the number of rounds of sequencing chemical reactions of the base, that is, the corresponding cycle number when the base is incorporated into the nucleotide chain, usually, its corresponding The number of cycles is small and the quality value is high; the signal strength can be the strength of the signal directly collected by the sequencer, including brightness, voltage level or current level, etc. It can be a normalized signal or a signal after phase loss correction; The degree to which the signal strength (and its adjacent signal strength) is close to an integer, that is, the difference between the normalized signal or the signal after phase loss correction or the signal after error correction correction and the nearest integer, usually, the difference is small, The accuracy is higher; the parameters of the sequencing signal, namely unit signal, background signal, lead coefficient, lag coefficient, attenuation coefficient, etc.; the degree of dephasing when the base is detected, usually, the degree of dephasing is low and the accuracy is higher ,etc.

According to a preferred embodiment, one or more sequencing signal features of the polymer are calculated. For example, only one sequencing signal feature may be selected for calculation, or two sequencing signal features may be selected, or more sequencing signal features may be selected.

Further, after labeling each polymer of the standard nucleic acid sequence, a classifier is trained to fit the relationship between the sequencing signal features of the polymer and its markers. Classifier is a conventional concept in the field of pattern recognition, including but not limited to, linear regression, polynomial regression, logistic regression, support vector machine, artificial neural network, random forest, Phred algorithm, ensemble learning, etc. With the development of the field of pattern recognition, a variety of novel classifier algorithms have been proposed in recent years. Using a novel classifier algorithm does not change the essence of the invention.

Specifically, a classifier is trained to fit the relationship between the sequencing signal features of each polymer and its markers; the classifier can divide the polymers into several categories according to the sequencing signal features of the polymers, and the statistics Accuracy for each class of polymers. For example, polymers with a length of 1, 2, 3, 4, 5, and 5 or more can be classified into one category, or the unit signal is in the range of 100-199, 200-299, 300-399, >400 The polymers within are divided into one class. When multiple sequencing signal features are used, orthogonal division can be performed. For example, polymers with a length of 1 and a unit signal within 100-199 are classified into one category, and polymers with a length of 1 and a unit signal within 200-299 objects into another category, and so on.

In a preferred embodiment, after the fitting is completed, the fitting result of the classifier is converted into a quality score. Extensive literature exists on how to convert classifier predictions into quality values. Taking the famous softmax algorithm as an example, let the output of a certain classifier be (a,b), where (1,0) means correct and (0,1) means error. Due to factors such as the accuracy of classifier training or calculation errors during prediction, the output of the classifier during prediction is not always exactly (1,0) or (0,1), but (0.9,0.05) or (0.1, 0.99) which is closer to (1,0) or (0,1) values. At this time, the softmax algorithm uses the following formula to convert the output (a, b) into a correct rate:

With the development of the field of pattern recognition, a variety of novel conversion algorithms have been proposed in recent years, for example, including Sparse-softmax, log-softmax, Taylor softmax, log-Taylor softmax, soft-marginsoftmax, SM-Taylor softmax, etc., using The novel transformation algorithm does not change the essence of the present invention.

For the nucleic acid sample to be tested, the above-mentioned method of converting the fitting result of the classifier into a quality score is used to predict the quality score of each polymer based on the sequencing signal characteristics of the nucleic acid to be tested.

In some embodiments, the quality score refers to a numerical value representing the accuracy of sequencing, which is selected from base calling accuracy, base calling error rate, Phred value and the like. It can be understood that the quality score is only an expression of the accuracy of base sequencing, and the expression itself is not important and does not affect the essence of the present invention. The quality score can also be directly expressed by the accuracy.

In some embodiments, the quality score is logarithmically based on base calling error probability, and wherein the plurality of quality scores includes Q10, Q15, Q20, Q25, Q30, Q35, Q40, Q45, Q50, Q55, Q60.

In some preferred embodiments, the quantization scheme for training and calibration can be completed in advance, and the trained classifier is stored in the system as a configuration file, which can be called when performing the quality scoring of the nucleic acid sequence to be tested.

In some preferred embodiments, the standard nucleic acid sample and the nucleic acid sample to be tested can be labeled with different molecular markers, and mixed together for simultaneous sequencing. After the sequencing is completed, the two samples are first split using molecular markers, and the quantification scheme for training and calibration is completed to obtain a trained classifier, which is then applied to the nucleic acid sample to be tested.

The present invention also discloses a method according to any of the preceding embodiments, wherein a reaction solution containing nucleotide substrate molecules is used for sequencing. Nucleotide substrate molecules refer to any one of A, G, C, T nucleotide substrate molecules, or two, or three; or A, G, C, U nucleotide substrate molecules Any one, or two, or three.

Disclosed herein is a sequencing method using fluorophore-labeled nucleotide substrate molecules according to any of the preceding embodiments, wherein each sequencing run uses a set of reaction solutions, each set of reaction solutions comprising at least two reaction solutions, each reaction solution comprising At least one of A, G, C, T nucleotide substrate molecules, or each reaction solution contains at least one of A, G, C, U nucleotide substrate molecules. In one aspect, the method includes fixing the nucleotide sequence fragment to be detected, passing a part of the reaction solution in a set of reaction solutions, and recording fluorescence information. In one aspect, the method includes passing one portion of the reaction solution at a time, and sequentially passing another portion of the same set of reaction solutions. In one aspect, there is at least one reaction solution in the reaction solution group, and the reaction solution contains two or three nucleotide molecules.

Disclosed herein is a sequencing method using fluorophore-labeled nucleotide substrate molecules, wherein each sequence uses a set of reaction solutions, each set of reaction solutions includes two reaction solutions, and each reaction solution contains two different bases. kind of nucleotides. On the one hand, the nucleotides in one of the reaction solutions can be complementary to the two bases on the nucleotide sequence to be tested, and the nucleotides in the other reaction solution can be complementary to the other two bases on the nucleic acid sequence to be tested. base complementarity. In one aspect, the method includes immobilizing the nucleotide sequence fragment to be detected, and passing through the first reaction solution in a set of reaction solutions. Then, a second reaction from the same set of reactions was added. The two reaction solutions can be added successively in an alternate manner, so as to obtain the coding information of the nucleotide substrate to be tested through the fluorescence information. When the two nucleotides in each reaction solution are labeled with different fluorescent labels (i.e., two-color 2+2 sequencing), perform two rounds of orthogonal two-color 2+2 sequencing; when the two nucleosides in each reaction solution When the acids have the same fluorescent label (ie, single-color 2+2 sequencing), preferably, three rounds of orthogonal single-color 2+2 sequencing are performed. Optionally, one part of a group of reaction solutions of the sequencing reaction contains 3 kinds of nucleotides, and the other part of the reaction solution contains another nucleotide different from it. At this time, four times of orthogonal Sequencing reactions (1+3 sequencing).

According to a preferred embodiment, the method of the present invention further includes performing bioinformatics analysis on the measured nucleic acid sequence according to the quality score of the polymer.

According to a preferred embodiment, the bioinformatics analysis refers to performing two or more times of orthogonal degenerate sequencing, after obtaining the quality value of each degenerate polymer length, using the quality value to correct Error code decoding (or error correction correction). For example, three times of orthogonal degenerate sequencing (MK, RY, WS) are performed on the nucleic acid sequence to be tested, and the obtained nucleic acid sequence is ACCGTTTGC. Taking the polymer CC as an example, it is sequenced in MK. The polymer is ACC, in the RY sequencing round, its degenerate polymer is CC, in the WS sequencing round, its degenerate polymer is CCG, and the polymer CC can be calculated separately in each degenerate polymer The characteristics of the object, such as the length of degenerate polymers, are 3, 2, and 3 respectively, and their quality scores are calculated respectively, and error correction correction (or error correction code decoding) is performed according to the quality scores, that is, the final nucleotide is determined sequence.

Take 2+2 degenerate sequencing as an example to illustrate the principle of ECC decoding. ECC sequencing consists of three rounds of degenerate sequencing of MK, RY, and WS. Each round of degenerate sequencing obtains a degenerate sequence containing half of the information of the DNA to be tested. By taking the intersection of three degenerate bases at the same position of the three degenerate sequences, the accurate base composition of the DNA to be tested can be obtained. There are a total of 8 intersection situations:

Among the 8 intersection situations, there are 4 legal situations, and four bases can be obtained respectively. In addition, there are 4 illegal situations, and the result of taking the intersection is an empty set. In an ideal situation without sequencing errors, the intersection of the three degenerate sequences should be all legal, and the sequence of the DNA to be tested can be obtained. Illegal situations occur near sequencing errors when the DPL calculated by the signal processing algorithm contains errors. Therefore, the illegality of the intersection of degenerate sequences indicates the presence of sequencing errors. In order to correct sequencing errors, firstly, the probability of occurrence of different sequencing error modes was obtained by using the sequencing data of the standard. Then, based on the maximum likelihood principle, the ECC decoding algorithm tries to correct the DPL calculated by the signal processing algorithm, and achieves two goals in the correction: 1. The corrected DPL can make the intersection of the three degenerate sequences all legal; 2. Under the constraints of the previous article, according to the probability of occurrence of different sequencing error modes, the probability of occurrence of this correction method is the highest. An optimization method can be used to achieve the above two goals.

In some embodiments, the bioinformatics analysis refers to screening high-quality nucleic acid sequences according to the assigned quality value. Screening methods include, but are not limited to, screening nucleic acid sequences whose quality values are all higher or lower than a certain threshold, screening nucleic acid sequences whose average value of all quality values is higher than or lower than a certain threshold, and screening nucleic acid sequences whose quality values are all higher than or lower than a certain threshold. Regions above or below a certain threshold value, regions in which the average value of quality values in the nucleic acid sequence are both above or below a certain threshold value, and so on. The high-quality nucleic acid sequence obtained after screening can be used to detect gene variation, detect gene expression, detect RNA alternative splicing status, detect gene modification status, identify species or individuals from nucleic acid sources, detect genome three-dimensional structure, detect nucleic acid and nucleic acid Interactions between nucleic acids, detection of interactions between nucleic acids and proteins, detection of chromatin accessibility, analysis of RNA structures, etc.

In some embodiments, the bioinformatics analysis refers to aligning the nucleic acid sequence to a reference sequence according to the assigned quality value. Alignment is a conventional concept in bioinformatics, which can be performed using BWA, Smith-Waterman algorithm, Bowtie, SOAP, Needleman-Wunch algorithm, Bowtie2, BLAST, ELAND, TMAP, MAQ, minimap2, SHRiMP software or algorithm. Methods for utilizing quality values in alignments include, but are not limited to:

1. Select the subsequence of high quality value as the seed of preliminary positioning sequence;

2. When there are multiple possible alignment methods, polymers/bases with low quality values are prioritized as unmatched parts in the alignment.

In some embodiments, the bioinformatics analysis refers to the identification of gene variation according to the alignment results and the quality values assigned to the aligned sequences. Genetic variation is a conventional concept in biology, including but not limited to single nucleotide polymorphism, copy number variation, epigenetic variation, large-scale structural variation, etc. Methods for utilizing quality values in identifying genetic variants include, but are not limited to:

1. For the genomic site of the variation to be identified, among all the bases of the sequence aligned to the site, the base with a higher quality value of the polymer/base is screened for identification.

2. Give the null hypothesis: there is no genetic variation at this site. According to the quality value and the comparison result, the probability of the null hypothesis being established is calculated. If the probability is greater than the given significance level, the null hypothesis is accepted; otherwise, the null hypothesis is rejected and the gene variation is considered to exist at this site.

In some embodiments, when identifying genetic variants, certain features of the alignment results can be used to remove potential false positive or false negative results. These are routine operations in bioinformatics, and their additions do not affect the essence of the present invention. Such characteristics include, but are not limited to:

1. The gene variation appears concentratedly in the sequence of forward or reverse alignment, but rarely occurs in the sequence of reverse or forward alignment;

2. The gene variation appears concentratedly at both ends of the sequence, but rarely occurs in the center of the sequence;

3. When pair-end sequencing is used, read1 detects that the site is mainly G to T, while read2 detects that the site is mainly C to A, or read1 detects that the site is mainly C Change to T, and the site detected by read2 is mainly G to A;

4. Other different gene mutations appear frequently near this gene variation.

In some embodiments, bioinformatics analysis refers to assembling nucleic acid sequences into longer nucleic acid sequences according to assigned quality values.

According to a preferred embodiment, training the classifier may be a probability distribution model based on maximum likelihood. Probability distribution refers to a probability distribution characterized by a unimodal shape, including but not limited to two-point distribution, binomial distribution, negative binomial distribution, Poisson distribution, geometric distribution, exponential distribution, normal distribution, Γ distribution, chi-square distribution, t distribution, F distribution, beta distribution, lognormal distribution, and high-dimensional extensions of the above distributions, etc. In the aforementioned probability distribution model, the expectation or peak of the probability distribution is related to the sequencing signal characteristics of the polymer, and the variance or skewness is related to the quality value of the polymer. It is known from basic statistics that different probability distributions can be completely determined by a set of parameters, for example, the normal distribution can be completely determined by two parameters of mean and standard deviation. After the sequence obtained by sequencing is compared to the reference genome, the corresponding relationship between the sequencing signal characteristics of each polymer and the length of the polymer (set as n) can be determined according to the comparison result. After being completely determined by the given parameters, the integrated area of the probability distribution between n-0.5 and n+0.5 represents the probability of the sequencing signal feature corresponding to the polymer length n. The likelihood function refers to the result of calculating the probability of a group of polymer sequencing signal features corresponding to the polymer length n after determining the probability distribution with given parameters, and then multiplying the probabilities together. The maximum likelihood refers to finding a set of parameters, so that after the probability distribution is determined with this set of parameters, the obtained likelihood function is the largest.

The maximum likelihood-based probability distribution model can divide the polymer sequencing signal features into several groups, and each group applies the maximum likelihood-based probability distribution model independently.

The likelihood function may be subjected to a certain mathematical transformation for the sake of simplicity of calculation. Mathematical transformations, for example, transform probabilistic multiplication into probabilistic addition by taking the logarithm.

The present invention also provides a method for evaluating the quality of nucleic acid sequencing data, which is characterized in that it includes: performing fuzzy sequencing or missing sequencing on the nucleic acid sample to be tested to obtain input data, generating degenerate polymer length information of the input data, and calculating The sequencing signal characteristics of the length of the degenerate polymer;

predicting a quality score for the polymer based on the sequencing signal characteristics using a quantization scheme calibrated for training;

The quantization scheme of the training calibration includes:

Sequencing a standard nucleic acid sample to obtain a nucleic acid sequence, calculating the sequencing signal characteristics of the nucleic acid sequence, comparing the nucleic acid sequence to a reference sequence, and marking the polymer of the nucleic acid sequence as being sequenced correctly or incorrectly; training a classifier, fitting Relationship between sequencing signature features of polymers and their markers.

According to a preferred embodiment, the sequencing signal characteristics of the length of the degenerate polymer include: the length of the polymer where the base is located, that is, the number of bases in the degenerate polymer where the base is located, usually, more The polymer length is short, the sequencing quality value is high; the position of the base in the polymer it is in, that is, the distance between the base and the nearest end of the degenerate polymer it is in, etc.

Specifically, train a classifier to fit the relationship between the sequencing signal features of each polymer and its markers; the classifier can divide the polymers into several categories according to the sequencing signal features of the polymers, and count each category Polymer Accuracy. For example, polymers with a length of 1, 2, 3, 4, 5, and more than 5 can be classified into one class respectively. Orthogonal partitioning can be performed when multiple sequencing signal features are used.

In a preferred embodiment, after the fitting is completed, the fitting result of the classifier is converted into a quality score. Extensive literature exists on how to convert classifier predictions into quality values.

The above is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed in the present invention can easily think of changes or Replacement should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be determined by the protection scope of the claims.

Claims (11)

1. A method for quality assessment of nucleic acid sequencing data, comprising:

providing a nucleic acid sequence to be detected, taking the polymer as a basic unit, and calculating sequencing signal characteristics of the polymer; predicting a mass score of the multimer based on sequencing signal features using a trained and calibrated quantization scheme;

the training calibrated quantization scheme includes:

for the provided standard nucleic acid sequence, taking the polymer as a basic unit, calculating the sequencing signal characteristics of the polymer, and marking the polymer as correct or incorrect sequencing according to the comparison result of the standard nucleic acid sequence; the classifier is trained to fit the relationship between the sequencing signal features of the multimer and its markers.

2. The method of claim 1, wherein the sequencing signal characteristic of the polymer refers to a characteristic of a signal generated when the polymer undergoes a sequencing chemistry during the sequencing process, including, but not limited to, the base type of the constituent polymer, the length of the polymer, the number of rounds of the sequencing chemistry, the signal strength, the degree to which the signal strength is near an integer, a parameter of the sequencing signal, the degree of loss of phase when the polymer is detected, etc.

3. The method according to claim 1 or 2, wherein the polymer comprises a homopolymer, a bipolymer, a terpolymer, or the like.

4. A method according to any one of claims 1-3, wherein the training classifier is based on a probability distribution model of maximum likelihood.

5. The method of claim 4, wherein training the classifier comprises classifying the polymers into a plurality of classes based on sequencing signal characteristics of the polymers, and counting sequencing accuracy for each class of polymers.

6. The method of claim 5, wherein the classifier includes, but is not limited to, linear regression, polynomial regression, logistic regression, support vector machines, artificial neural networks, random forests, phred algorithms, ensemble learning, and the like.

7. The method of any one of claims 1-6, wherein the standard nucleic acid sequence is a sequence obtained by sequencing a standard nucleic acid sample; the standard nucleic acid sample refers to a nucleic acid sample which has been determined in both source and sequence and is highly homozygous at almost all sites of the genome, and includes lambda phage DNA, E.coli DNA, saccharomyces cerevisiae DNA, etc.

8. The method of claim 7, further comprising performing a bioinformatic analysis of the nucleic acid sequence to be tested based on the mass score of the multimer.

9. The method of claim 8, wherein the bioinformatic analysis comprises identifying genetic variations based on the alignment and the quality value assigned to the aligned sequences.

10. The method of claim 9, wherein the bioinformatic analysis comprises performing at least two orthogonal degenerate sequencing runs to obtain a mass value for a degenerate polymer length, and correcting with the mass value.

11. A method for quality assessment of nucleic acid sequencing data, comprising:

performing fuzzy sequencing or deletion sequencing on a nucleic acid sample to be detected to obtain input data, generating degenerate polymer length information of the input data, and calculating sequencing signal characteristics of the degenerate polymer length; predicting a mass score of the multimer based on the sequencing signal features using a quantification protocol calibrated for training;

the training calibrated quantization scheme includes:

sequencing a standard nucleic acid sample to obtain a nucleic acid sequence, calculating sequencing signal characteristics of the nucleic acid sequence, comparing the nucleic acid sequence to a reference sequence, and marking a polymer of the nucleic acid sequence as sequencing correct or incorrect; the classifier is trained to fit the relationship between the sequencing signal features of the multimer and its markers.

Images