A sequencing quality assessment method, device, equipment, medium and product

By evaluating sequence characteristics using insertion, mismatch and missing base quality models, and calculating the quality of target parameters, the problem of traditional base quality value evaluation deviation is solved, and the accuracy of nucleic acid sequence quality evaluation is improved.

CN118471340BActive Publication Date: 2024-12-03GENEMIND BIOSCIENCES CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410910765.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-09
Publication Date
2024-12-03
Estimated Expiration
2044-07-09

AI Technical Summary

Technical Problem

Traditional base mass values ​​are affected by factors such as insertion errors and deletion errors when evaluating the quality of nucleic acid sequencing, resulting in large deviations in quality assessment.

Method used

The quality of target parameters is calculated by determining the characteristic data of the sequence to be evaluated and using pre-trained insertion base quality model, mismatch base quality model and missing base quality model to improve the accuracy of nucleic acid sequence quality evaluation.

Benefits of technology

The particle size and accuracy of nucleic acid sequence quality evaluation have been improved, and the problem of unclear base mass value factors have been solved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118471340B_ABST
    Figure CN118471340B_ABST
Patent Text Reader

Abstract

The present invention discloses a sequencing quality assessment method, device, equipment, medium and product. The method comprises: determining sequence feature data to be assessed according to the sequencing data to be assessed; determining the target parameter quality of the sequencing data to be assessed according to at least one sequence feature to be assessed in the sequence feature data to be assessed and a sequence parameter quality model that has been pre-trained; wherein the sequence parameter quality model is at least selected from at least one of an insertion base quality model, a mismatched base quality model and a missing base quality model, and when the sequence parameter quality model is an insertion base quality model, the target parameter quality is the target insertion base quality, when the sequence parameter quality model is a mismatched base quality model, the target parameter quality is the target mismatched base quality, and when the sequence parameter quality model is a missing base quality model, the target parameter quality is the target missing base quality, thereby improving the granularity and accuracy of the quality assessment of the nucleic acid sequence.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of gene sequencing technology, and in particular to a sequencing quality assessment method, device, equipment, medium and product. Background Art

[0002] Nucleic acid sequencing technology is a method used to determine the molecular sequence of DNA (Deoxyribonucleic Acid) or RNA (Ribonucleic Acid), and is widely used in technical fields such as genomics, medical diagnosis and drug development.

[0003] Currently, sequencing quality assessment indicators rely on base quality values ​​determined based on base accuracy. However, base accuracy is affected by multiple factors such as insertion errors and deletion errors. Therefore, in some base sequences with insertion errors and deletion errors, the base quality values ​​may have large deviations when reflecting the true quality of nucleic acid sequencing sequences. Summary of the Invention

[0004] The embodiments of the present invention provide a sequencing quality assessment method, apparatus, device, medium and product to solve the problem of unclear quality factors of traditional base quality values ​​and improve the granularity and accuracy of nucleic acid sequence quality assessment.

[0005] According to one embodiment of the present invention, a sequencing quality assessment method is provided, the method comprising:

[0006] Determining sequence feature data to be evaluated based on the sequencing data to be evaluated; wherein the sequence feature data to be evaluated includes at least one sequence feature to be evaluated;

[0007] Determining the target parameter quality of the sequencing data to be evaluated based on the characteristics of each sequence to be evaluated and a pre-trained sequence parameter quality model;

[0008] Wherein, the sequence parameter quality model is at least selected from at least one of an inserted base quality model, a mismatched base quality model and a deleted base quality model, and when the sequence parameter quality model is an inserted base quality model, the target parameter quality is the target inserted base quality, when the sequence parameter quality model is a mismatched base quality model, the target parameter quality is the target mismatched base quality, and when the sequence parameter quality model is a deleted base quality model, the target parameter quality is the target deleted base quality.

[0009] According to another embodiment of the present invention, a sequencing quality assessment device is provided, the device comprising:

[0010] A module for determining sequence feature data to be evaluated, configured to determine sequence feature data to be evaluated based on sequencing data to be evaluated; wherein the sequence feature data to be evaluated includes at least one sequence feature to be evaluated;

[0011] a target parameter quality determination module, configured to determine the target parameter quality of the sequencing data to be evaluated based on the characteristics of each sequence to be evaluated and a pre-trained sequence parameter quality model;

[0012] Wherein, the sequence parameter quality model is at least selected from at least one of an inserted base quality model, a mismatched base quality model and a deleted base quality model, and when the sequence parameter quality model is an inserted base quality model, the target parameter quality is the target inserted base quality, when the sequence parameter quality model is a mismatched base quality model, the target parameter quality is the target mismatched base quality, and when the sequence parameter quality model is a deleted base quality model, the target parameter quality is the target deleted base quality.

[0013] According to another embodiment of the present invention, an electronic device is provided, the electronic device including:

[0014] at least one processor; and

[0015] a memory communicatively connected to the at least one processor; wherein,

[0016] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can perform the sequencing quality assessment method described in any embodiment of the present invention.

[0017] According to another embodiment of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the sequencing quality assessment method described in any embodiment of the present invention when executed.

[0018] According to another embodiment of the present invention, a computer program product is provided, comprising a computer program, wherein when the computer program is executed by a processor, the method for assessing sequencing quality according to any embodiment of the present invention is implemented.

[0019] The technical solution of the embodiment of the present invention determines the sequence feature data to be evaluated based on the sequencing data to be evaluated, and determines the target parameter quality of the sequencing data to be evaluated based on at least one sequence feature to be evaluated in the sequence feature data to be evaluated and a pre-trained sequence parameter quality model, wherein the sequence parameter quality model is at least selected from at least one of an inserted base quality model, a mismatched base quality model, and a deleted base quality model, and the target parameter quality includes at least one of a target inserted base quality, a target mismatched base quality, and a target deleted base quality. For the same set of sequence feature data to be evaluated for each base variation, a dedicated sequence parameter quality model is used to perform quality assessment, thereby solving the problem of unclear quality factors of traditional base quality values ​​and improving the granularity and accuracy of quality assessment of nucleic acid sequences.

[0020] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present invention, nor is it intended to limit the scope of the present invention. Other features of the present invention will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0022] Figure 1 A flowchart of a sequencing quality assessment method provided by one embodiment of the present invention;

[0023] Figure 2 A flowchart of another sequencing quality assessment method provided by one embodiment of the present invention;

[0024] Figure 3 A scatter plot showing the distribution of a test insertion base error rate and a standard insertion base error rate of an insertion base quality model provided by one embodiment of the present invention;

[0025] Figure 4 A scatter plot showing the distribution of the test mismatch base error rate and the standard mismatch base error rate of a mismatch base quality model provided by one embodiment of the present invention;

[0026] Figure 5 A scatter plot showing the distribution of the test missing base error rate and the standard missing base error rate of a missing base quality model provided by one embodiment of the present invention;

[0027] Figure 6A schematic diagram of determining a target quality level based on a probability density curve provided by one embodiment of the present invention;

[0028] Figure 7 A schematic diagram of determining a target parameter quality score based on a probability density curve provided by one embodiment of the present invention;

[0029] Figure 8 A schematic structural diagram of a sequencing quality assessment device provided by one embodiment of the present invention;

[0030] Figure 9 The present invention provides a schematic structural diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0032] In the embodiments of this application, the term "sequencing" may also be referred to as "nucleic acid sequencing" or "gene sequencing." The three terms are interchangeable and all refer to the determination of the types and order of bases or nucleotides (including nucleotide analogs) in nucleic acid molecules. Sequencing as such includes the process of binding nucleotides to a template and collecting the corresponding signals emitted by the nucleotides (including analogs). Sequencing as such includes sequencing by synthesis (sequencing by synthesis, SBS) and / or sequencing by ligation (sequencing by ligation, SBL), including DNA sequencing and / or RNA sequencing, and including long-fragment sequencing and / or short-fragment sequencing. The terms "long" and "short" are relative. For example, nucleic acid molecules longer than 1 Kb, 2 Kb, 5 Kb, or 10 Kb are referred to as long fragments, while those shorter than 1 Kb or 800 bp are referred to as short fragments.

[0033] Sequencing generally involves multiple rounds of determining the types and order of multiple bases or nucleotides on a nucleic acid template. In the present embodiments, each round of determining the types and order of multiple bases or nucleotides on a nucleic acid template is referred to as a "sequencing round." A "sequencing round" (also known as a "sequencing round") can be defined as a single base extension of four nucleotides / bases. In other words, a "sequencing round" can be defined as the complete determination of the type of base or nucleotide at any given position on the template. For sequencing platforms that use polymerization or ligation reactions to achieve sequencing, a sequencing round involves the process of binding four nucleotides (including nucleotide analogs) to the nucleic acid template via base complementarity and collecting the corresponding signals emitted. For platforms that use polymerization reactions to achieve sequencing, the reaction system includes reaction substrate nucleotides, a polymerase, and a nucleic acid template. The nucleic acid template is bound to a sequence (sequencing primer). Based on the principles of base pairing and polymerization, the added reaction substrate nucleotides are catalyzed by the polymerase and ligated to the sequencing primer, enabling the nucleotides to bind to a specific position on the nucleic acid template.

[0034] Generally, one round of sequencing may include one or more base extensions (repeat). For example, four nucleotides are added to the reaction system in sequence, and base extension and corresponding reaction signal collection are performed respectively. One round of sequencing includes four base extensions. For another example, any combination of the four nucleotides is added to the reaction system, such as two-by-two combinations or one-by-three combinations, and base extension and corresponding reaction signal collection are performed for two combinations respectively. One round of sequencing includes two base extensions. For another example, four nucleotides are added to the reaction system at the same time for base extension and reaction signal collection. One round of sequencing includes one base extension.

[0035] In the examples of this application, the term "template" refers to a "nucleic acid template," which refers to a master nucleic acid molecule or a master nucleic acid molecule fragment to which nucleotides or nucleotide analogs are attached through multiple rounds of base extension. The so-called nucleic acid template can be the entire sequence of a nucleic acid molecule or a partial fragment of a nucleic acid molecule. In the examples of this application, the number of templates refers to the number of nucleic acid templates. It should be noted that in nucleic acid sequencing performed after amplification, the nucleic acid molecule cluster formed after amplification serves as a nucleic acid template.

[0036] In the embodiments of this application, the term "channel" refers to pathways formed in different ways during the sequencing process that are capable of screening and distinguishing the four bases derived from adenine A, cytosine C, guanine G, and thymine T (or uracil U). For example, a channel may refer to an optical channel of fluorescent signals that is formed during the sequencing image capture process by using different excitation lights, different fluorescent filters, and the like, capable of screening and distinguishing the four fluorescent bases derived from adenine A, cytosine C, guanine G, and thymine T (or uracil U).

[0037] In the embodiments of this application, "spots" (also known as "peaks") refer to locations on an optical image where the optical signal is relatively strong. For example, the signal at this location is stronger than the surrounding area, and appears as a relatively bright spot or point on the image. A spot or location occupies one or more pixels. The signal at the spot / location may originate from the target molecule or non-target substances. Detection of "spots" includes detection of the optical signal of the target molecule, such as an extended base or base cluster.

[0038] It should be noted that the terms "first", "second", "initial", "target", "reference", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0039] Figure 1 This is a flowchart of a sequencing quality assessment method provided by one embodiment of the present invention. This embodiment is applicable to the case of quality assessment of sequencing data of nucleic acid samples, and is particularly applicable to the case of sequence quality assessment of sequencing data. The method can be performed by a sequencing quality assessment device, which can be implemented in the form of hardware and / or software, and can be configured in a terminal device. Figure 1 As shown, the method includes:

[0040] S110 , determining sequence feature data to be evaluated based on the sequencing data to be evaluated.

[0041] Specifically, the sequencing data to be evaluated is used to represent sequencing data obtained by sequencing the nucleic acid sample to be evaluated, and the sequencing data to be evaluated includes at least one sequencing sequence and / or images respectively acquired based on at least one excitation light channel.

[0042] Exemplary, the sequencing technology used in the sequencing process includes but is not limited to Sanger sequencing technology, high-throughput sequencing technology and single-molecule sequencing technology, etc. The sequencing technology used in the sequencing process is not limited here, and can be customized according to actual needs. The sequencing method can be selected as single-end sequencing or double-end sequencing. Sequencing can be performed by a sequencing platform, and the sequencing platform can be selected but not limited to Illumina's Hiseq / Miseq / Nextseq / Novaseq sequencing platform, Thermo Fisher / Life Technologies' Ion Torrent platform, BGISEQ and MGISEQ platforms and single-molecule sequencing platforms of BGI, and Geno Lab / Geno Care / FASTA / SURF Sec sequencing platforms of Zhenmai Biotechnology, but is not limited thereto.

[0043] In this embodiment, the sequence feature data to be evaluated includes at least one sequence feature to be evaluated. Specifically, the sequence feature to be evaluated is used to represent data features obtained after performing sequence-dimensional data analysis on the sequencing data to be evaluated.

[0044] In an optional embodiment, the sequence feature data to be evaluated includes at least one of the ratio of the number of sequences of the first sequence to the total number of sequences (Output Ratio), the average base quality value of the first sequence (Clean Avg Read Score), the number of sequences of the second sequence (Good Read Num), the ratio of the number of sequences of the second sequence to the total number of sequences (Good Read Ratio), the number of base pairs of the second sequence (Good Read_Base), the average base quality value of the sequencing data to be evaluated (Raw Avg ReadScore), the single-channel base recognition ratio (Total Match Ratio), the multi-channel base recognition ratio (Total DoubleCalled Ratio), the channel cycle recognition base ratio (Base Repeat Ratio), the channel image quality (I Score) and the template density (Density).

[0045] In this embodiment, the first sequence is a sequence whose sequence read length is greater than or equal to the read length threshold. Specifically, the ratio of the number of sequences of the first sequence to the total number of sequences can be called the sequence throughput (Output Ratio), the sequence read length is used to characterize the length of the nucleic acid sequence fragment measured by sequencing the nucleic acid sample to be evaluated, and the total number of sequences is used to characterize the total number of sequencing sequences contained in the sequencing data to be evaluated. Exemplarily, the read length threshold can be 25 base pairs (base-pair, bp) or 30 base pairs. The specific parameter value of the read length threshold can be customized according to actual needs, such as the species origin of the nucleic acid sample to be tested. In some embodiments, the inventors found that when the read length threshold is set to 20-30bp, it can better reflect the sequencing quality and is associated with the sequencing error rate.

[0046] In this embodiment, the average base quality value of the first sequence is used to characterize the average value of the base quality values ​​corresponding to the first sequence in the sequencing data to be evaluated. Specifically, the base quality value (Quality Score, Q) characterizes the integer mapping of the probability of base recognition error, which can be used to quantitatively evaluate the credibility of the recognized base. In one embodiment, when the base is identified by an optical signal, it can be confirmed based on the optical signal intensity presented by the four bases in the same light spot. The optical signal intensity can be the absolute value of the optical intensity measured under certain conditions, or it can be a relative optical intensity. Exemplarily, the base quality value can be determined by one of the following methods:

[0047] 1) Among the four optical signal intensities corresponding to the four bases, take the ratio of the maximum optical signal intensity to the second maximum optical signal intensity;

[0048] 2) Among the four optical signal intensities corresponding to the four bases, determine the sum of the maximum optical signal intensity and the second largest optical signal intensity, and calculate the ratio of the maximum optical signal intensity to the sum;

[0049] 3) Among the four optical signal intensities corresponding to the four bases, take the ratio of the maximum optical signal intensity to the sum of the four optical signal intensities;

[0050] 4) Among the four optical signal intensities corresponding to the four bases, determine the difference between the maximum optical signal intensity and the second maximum optical signal intensity, as well as the sum of the four optical signal intensities, and calculate the ratio of the difference to the sum.

[0051] 5) Among the standard deviations of the four optical signal intensities corresponding to the four bases, take the ratio of the standard deviation corresponding to the maximum optical signal intensity to the standard deviation corresponding to the second largest optical signal intensity;

[0052] 6) Among the standard deviations of the four optical signal intensities corresponding to the four bases, determine the summed standard deviation between the standard deviation corresponding to the maximum optical signal intensity and the standard deviation corresponding to the second largest optical signal intensity, and calculate the ratio of the standard deviation corresponding to the maximum optical signal intensity to the summed standard deviation;

[0053] 7) Among the standard deviations of the four optical signal intensities corresponding to the four bases, take the ratio of the standard deviation corresponding to the maximum optical signal intensity to the sum of the standard deviations corresponding to the four optical signal intensities;

[0054] 8) Among the standard deviations of the four optical signal intensities corresponding to the four bases, determine the standard value of the difference between the standard deviation corresponding to the maximum optical signal intensity and the standard deviation corresponding to the second largest optical signal intensity, and the standard deviation of the sum of the standard deviations corresponding to the four optical signal intensities, and calculate the ratio of the standard deviation of the difference to the standard deviation of the sum.

[0055] The larger the above ratio is, the greater the base quality value is.

[0056] The average base quality value of the sequencing data to be evaluated in this embodiment is used to represent the arithmetic mean of the base quality values ​​corresponding to all sequencing sequences in the sequencing data to be evaluated.

[0057] In this embodiment, the second sequence is the first sequence whose base quality value is greater than or equal to the base quality threshold. For example, the base quality threshold can be 65 or 70. The specific parameter value of the base quality threshold can be customized based on actual needs, such as the species origin of the nucleic acid sample to be tested. In some embodiments, the inventors have found that setting the base quality threshold between 60 and 70 can better reflect sequencing quality and correlate with the sequencing error rate.

[0058] In this embodiment, the single-channel base recognition ratio is the ratio of the number of bright spots recognized as bases in a single excitation light channel to the total number of bright spots detected by the excitation light channel.

[0059] In some embodiments, the sequence characteristic data to be evaluated includes at least one single-channel base recognition ratio. The number of single-channel base recognition ratios is determined by the number of excitation light wavelengths used when sequencing the nucleic acid sample to be evaluated.

[0060] In one embodiment, sequencing uses dual excitation lights with different excitation wavelengths to determine base types. In this case, the sequence characteristic data to be evaluated includes the first single-channel base recognition ratio of the first excitation light channel corresponding to the first excitation wavelength, and the second single-channel base recognition ratio of the second excitation light channel corresponding to the second excitation wavelength. In another embodiment, sequencing uses three excitation lights with different excitation wavelengths to determine base types. In this case, the sequence characteristic data to be evaluated includes the first single-channel base recognition ratio of the first excitation light channel corresponding to the first excitation wavelength, the second single-channel base recognition ratio of the second excitation light channel corresponding to the second excitation wavelength, and the third single-channel base recognition ratio of the third excitation light channel corresponding to the third excitation wavelength. In another embodiment, sequencing uses four excitation lights with different excitation wavelengths to determine base types. In this case, the sequence characteristic data to be evaluated includes the first single-channel base recognition ratio of the first excitation light channel corresponding to the first excitation wavelength, the second single-channel base recognition ratio of the second excitation light channel corresponding to the second excitation wavelength, the third single-channel base recognition ratio of the third excitation light channel corresponding to the third excitation wavelength, and the fourth single-channel base recognition ratio of the fourth excitation light channel corresponding to the fourth excitation wavelength.

[0061] In this embodiment, the multi-channel base recognition ratio is the ratio of the number of bright spots recognized as bases in at least two excitation light channels to the number of templates. The number of excitation light channels in the multi-channel base recognition ratio is determined by the number of excitation light wavelengths used when sequencing the nucleic acid sample to be evaluated.

[0062] In one embodiment, sequencing uses dual excitation lights with different excitation lights to determine the base type. In this case, the multi-channel base recognition ratio is the ratio of the number of bright spots that are recognized as bases in both the first excitation light channel corresponding to the first excitation wavelength and the second excitation light channel corresponding to the second excitation wavelength to the number of templates. In another embodiment, sequencing uses three excitation lights with different excitation lights to determine the base type. In this case, the multi-channel base recognition ratio is the ratio of the number of bright spots that are recognized as bases in both the first excitation light channel corresponding to the first excitation wavelength, the second excitation light channel corresponding to the second excitation wavelength, and the third excitation light channel corresponding to the third excitation wavelength to the number of templates. In another embodiment, sequencing uses four excitation lights with different excitation lights to determine the base type. In this case, the multi-channel base recognition ratio is the ratio of the number of bright spots that are recognized as bases in both the first excitation light channel corresponding to the first excitation wavelength, the second excitation light channel corresponding to the second excitation wavelength, the third excitation light channel corresponding to the third excitation wavelength, and the fourth excitation light channel corresponding to the fourth excitation wavelength to the number of templates.

[0063] In this embodiment, the channel cycle recognition base ratio is the ratio of the number of base pairs that are continuously recognized in a preset number of rounds in a single excitation light channel to the number of templates.

[0064] In some embodiments, the sequence feature data to be evaluated includes at least one channel cycle recognition base ratio, and the specific number of channel cycle recognition base ratios is determined by the number of excitation light wavelengths used when evaluating the nucleic acid sample for sequencing.

[0065] In one embodiment, sequencing uses dual excitation lights with different excitation lights to achieve base type determination. In this case, the sequence characteristic data to be evaluated includes the first cycle identification base ratio of the first excitation light channel corresponding to the first excitation wavelength, and the second cycle identification base ratio of the second excitation light channel corresponding to the second excitation wavelength. In another embodiment, sequencing uses three excitation lights with different excitation lights to achieve base type determination. In this case, the sequence characteristic data to be evaluated includes the first cycle identification base ratio of the first excitation light channel corresponding to the first excitation wavelength, the second cycle identification base ratio of the second excitation light channel corresponding to the second excitation wavelength, and the third cycle identification base ratio of the third excitation light channel corresponding to the third excitation wavelength. In another embodiment, sequencing uses four excitation lights with different excitation lights to achieve base type determination. In this case, the sequence characteristic data to be evaluated includes the first cycle identification base ratio of the first excitation light channel corresponding to the first excitation wavelength, the second cycle identification base ratio of the second excitation light channel corresponding to the second excitation wavelength, the third cycle identification base ratio of the third excitation light channel corresponding to the third excitation wavelength, and the fourth cycle identification base ratio of the fourth excitation light channel corresponding to the fourth excitation wavelength. For example, the preset number of rounds can be 8 or 9 rounds. The specific parameter value of the preset number of rounds can be customized according to actual needs, such as the species origin of the nucleic acid sample to be tested. In some embodiments, the inventors have found that when the preset number of rounds is set to be greater than or equal to 8, it can better reflect the sequencing quality and be correlated with the sequencing error rate.

[0066] In the embodiment of the present application, the template density is the average number of templates per square micron.

[0067] In an alternative embodiment, the first excitation wavelength has a wavelength of 500 nm to 580 nm. In an alternative embodiment, the second excitation wavelength has a wavelength of 600 nm to 680 nm. In an alternative embodiment, the first excitation wavelength has a wavelength of 500 nm to 580 nm, and the second excitation wavelength has a wavelength of 600 nm to 680 nm.

[0068] In an optional embodiment, the sequencing data to be evaluated includes an image acquired based on an excitation light channel, and the sequence features to be evaluated include the channel image quality corresponding to the excitation light channel.

[0069] In an optional embodiment, the sequence feature data to be evaluated is determined based on the sequencing data to be evaluated, including: obtaining first pixel values ​​corresponding to at least two pixels of interest in the image; for each pixel of interest, obtaining second pixel values ​​of pixels adjacent to the pixel of interest, and determining a pixel score of the pixel of interest based on the first pixel value and the second pixel value; and determining the channel image quality of the excitation light channel based on the pixel scores corresponding to each pixel of interest.

[0070] Specifically, the pixel of interest can be used to represent the pixel corresponding to the bright spot identified as the base in the image. For example, the adjacent pixel can be a pixel defined based on a 4-neighborhood or a pixel defined based on an 8-neighborhood.

[0071] In an optional embodiment, taking the adjacent pixels as pixels defined based on 8 neighborhoods as an example, the pixel scores are Satisfies the following formula:

[0072]

[0073] in, represents the first pixel value, Indicates the sum of the 8 second pixel values.

[0074] In one optional embodiment, determining the channel image quality of the excitation light channel based on the pixel scores corresponding to the pixels of interest includes sorting the pixel scores corresponding to the pixels of interest and using the pixel scores corresponding to a preset value ratio in the sorted results as the channel image quality of the excitation light channel. For example, when the sorting is in ascending order, the preset value ratio may be 90%; when the sorting is in descending order, the preset value ratio may be 10%. The preset value ratio is not limited here and can be customized according to actual needs.

[0075] In another optional embodiment, the channel image quality of the excitation light channel is determined based on the pixel scores corresponding to each pixel of interest, including: filtering the pixel scores corresponding to each pixel of interest according to a preset score threshold, and using the statistical value corresponding to at least one pixel score obtained by filtering as the channel image quality of the excitation light channel.

[0076] For example, the preset score threshold may be 80%, and the statistical value may be the arithmetic mean or median. There is no limitation on the preset score threshold and the statistical value, and they can be customized according to actual needs.

[0077] Based on the above embodiment, optionally, the model architecture of the sequence parameter quality model is a linear regression model. After determining the sequence feature data to be evaluated based on the sequencing data to be evaluated, the method further includes: obtaining feature weights corresponding to at least two sequence features to be evaluated in the sequence parameter quality model; and screening at least two sequence features to be evaluated in the sequence feature data to be evaluated based on the at least two feature weights.

[0078] Exemplarily, the model principle of the sequence parameter quality model can be expressed as:

[0079]

[0080] in, represents the target parameter quality, 、 … express The sequence features to be evaluated, represents the intercept term, Indicates the feature weight corresponding to each sequence feature to be evaluated.

[0081] The advantage of this setting is that it reduces the amount of input data for the sequence parameter quality model, thereby reducing the amount of calculation for the sequence parameter quality model and improving the efficiency of sequencing quality assessment.

[0082] S120 , determining the target parameter quality of the sequencing data to be evaluated based on the characteristics of each sequence to be evaluated and a pre-trained sequence parameter quality model.

[0083] In this embodiment, the sequence parameter quality model is at least selected from at least one of an inserted base quality model, a mismatched base quality model, and a deleted base quality model, and when the sequence parameter quality model is an inserted base quality model, the target parameter quality is the target inserted base quality, when the sequence parameter quality model is a mismatched base quality model, the target parameter quality is the target mismatched base quality, and when the sequence parameter quality model is a deleted base quality model, the target parameter quality is the target deleted base quality.

[0084] Among them, the inserted base quality is used to characterize the situation where the sequencing sequence in the sequencing data to be evaluated has one or more additional bases relative to the reference sequence; the mismatched base quality is used to characterize the situation where one or more bases in the sequencing sequence in the sequencing data to be evaluated are mismatched relative to the reference sequence; the missing base quality is used to characterize the situation where the sequencing sequence in the sequencing data to be evaluated has one or more bases missing relative to the reference sequence.

[0085] The reference sequence can be a human reference genome sequence or a reference genome sequence of another species. There is no limitation on the source of the reference sequence and it can be customized according to actual needs.

[0086] Specifically, the at least two target parameter quality expression forms can be the same or different. Exemplarily, the target parameter quality expression form can be the presence of base variation, quality probability, quality grade, or quality score. The quality probability can be an accuracy rate or an error rate, and a higher quality grade or quality score indicates better sequencing quality of the sequencing data to be evaluated. Conversely, a lower quality grade or quality score indicates worse sequencing quality of the sequencing data to be evaluated. The expression form of the target parameter quality is not limited here and can be customized according to actual needs.

[0087] Specifically, the model architectures of the at least two sequence parameter quality models may be the same or different. For example, the model architecture of the sequence parameter quality model may be a logistic regression model, a linear regression model, a decision tree, a support vector machine, a random forest, or a naive Bayes classifier. The model architecture of the sequence parameter quality model is not limited here and can be customized according to actual needs.

[0088] In an optional embodiment, the target parameter quality of the sequencing data to be evaluated is determined based on each sequence feature to be evaluated and a pre-trained sequence parameter quality model, including: inputting each sequence feature to be evaluated into the pre-trained sequence parameter quality model to obtain the output target parameter quality of the sequencing data to be evaluated.

[0089] The technical solution of this embodiment determines sequence feature data to be evaluated based on the sequencing data to be evaluated, and determines the target parameter quality of the sequencing data to be evaluated based on at least one sequence feature to be evaluated in the sequence feature data to be evaluated and a pre-trained sequence parameter quality model, wherein the sequence parameter quality model is at least selected from at least one of an inserted base quality model, a mismatched base quality model, and a deleted base quality model, and the target parameter quality includes at least one of a target inserted base quality, a target mismatched base quality, and a target deleted base quality. For the same set of sequence feature data to be evaluated for each base variation, a dedicated sequence parameter quality model is used to perform quality assessment, thereby solving the problem of unclear quality factors in traditional base quality values ​​and improving the granularity and accuracy of nucleic acid sequence quality assessment.

[0090] Figure 2 This is a flow chart of another sequencing quality assessment method provided by one embodiment of the present invention. This embodiment further refines the above embodiment of "determining the target parameter quality of the sequencing data to be assessed based on the characteristics of each sequence to be assessed and the pre-trained sequence parameter quality model." Figure 2 As shown, the method includes:

[0091] S210: Determine sequence feature data to be evaluated based on the sequencing data to be evaluated.

[0092] S210 in this embodiment is the same as that in the above embodiment. Figure 1 The steps in S110 shown are the same or similar and will not be described in detail in this embodiment.

[0093] S220 , inputting each sequence feature to be evaluated into a pre-trained sequence parameter quality model to obtain a reference parameter quality of the sequencing data to be evaluated.

[0094] The method for determining the “reference parameter quality” in this embodiment is the same as or similar to the method for determining the “target reference quality” in the above embodiment, and will not be described in detail in this embodiment.

[0095] Based on the above embodiment, the method may optionally further include: determining training sequence feature data based on the training sequencing data; wherein the training sequence feature data includes at least one training sequence feature; inputting each training sequence feature into an untrained initial parameter quality model to obtain the output predicted parameter quality of the training sequencing data; determining a loss function based on the predicted parameter quality and the training parameter quality of the training sequencing data, and adjusting the model parameters of the initial parameter quality model based on the loss function; until the training end condition is met, determining a trained sequence parameter quality model based on the initial parameter quality model in the current iteration process.

[0096] Specifically, the function types of the loss functions corresponding to the at least two sequence parameter quality models can be the same or different. Exemplary loss functions include squared loss function, logarithmic loss function, exponential loss function, mean squared error loss function, logistic regression loss function, Huber loss function, cross entropy loss function, and Kullback-Leibler divergence loss function. The function types of loss functions are not limited here and can be customized according to actual needs.

[0097] Specifically, the training end conditions include the number of iterations reaching an iteration threshold and / or the loss function convergence. Exemplarily, the iteration threshold can be 2000 times. The iteration threshold is not limited here and can be customized according to actual needs.

[0098] In an optional embodiment, determining the trained sequence parameter quality model according to the initial parameter quality model in the current iteration process includes: using the initial parameter quality model in the current iteration process as the trained sequence parameter quality model.

[0099] In another optional embodiment, a trained sequence parameter quality model is determined based on the initial parameter quality model in the current iteration process, including: determining training sequence feature data based on test sequencing data; wherein the training sequence features include at least one test sequence feature; inputting each test sequence feature into the initial parameter quality model in the current iteration process to obtain the test parameter quality of the output test sequencing data; determining the model test performance based on the test parameter quality and the standard parameter quality of the test sequencing data; and when the model test performance meets the preset performance conditions, using the initial parameter quality model in the current iteration process as the trained sequence parameter quality model.

[0100] In an optional embodiment, the model test performance includes absolute mean error and / or goodness of fit. The absolute mean error is used to characterize the degree of difference between the test parameter quality and the standard parameter quality, and the goodness of fit is used to characterize whether the target parameter quality model fits sufficient feature information.

[0101] For example, the mean absolute error It can be expressed by the following formula:

[0102]

[0103] For example, goodness of fit It can be expressed by the following formula:

[0104]

[0105] in, Indicates the amount of test sequencing data, Indicates the The standard parameter quality corresponding to the test sequencing data, Indicates the The quality of the test parameters corresponding to the test sequencing data, Indicates the average parameter quality corresponding to all standard parameter qualities.

[0106] Among them, the difference between the test parameter quality and the standard parameter quality can be used to characterize the characteristic information that the target parameter quality model does not fit, and the difference between the standard parameter quality and the average parameter quality can be used to characterize the characteristic information carried by the standard parameter quality. The closer it is to 1, the better the model test performance of the target reference quality model is, and the goodness of fit is The closer it is to 0, the worse the model testing performance of the target reference quality model.

[0107] Based on the above embodiment, optionally, the method further includes: when the model test performance does not meet the preset performance conditions, adjusting the hyperparameters of the initial parameter quality model and repeating the training steps of the sequence parameter quality model.

[0108] Figure 3 A scatter plot of the test insertion base error rate and the standard insertion base error rate of an insertion base quality model provided by one embodiment of the present invention, specifically, Figure 3 The horizontal axis of the distribution scatter plot represents the test sequencing data, the vertical axis represents the insertion base error rate, the blue scatter points represent the standard insertion base error rate corresponding to the test sequencing data, and the orange scatter points represent the test insertion base error rate corresponding to the test sequencing data. Figure 3 The distribution scatter plot shows the corresponding absolute mean error The goodness of fit is 0.000349. The value is 0.96, which indicates that the model test performance of the inserted base quality model in the current iteration is very good.

[0109] Figure 4 A scatter plot of the test mismatch base error rate and the standard mismatch base error rate of a mismatch base quality model provided by one embodiment of the present invention, specifically, Figure 4 The horizontal axis of the distribution scatter plot in represents the test sequencing data, the vertical axis represents the mismatch base error rate, the blue scatter points represent the standard mismatch base error rate corresponding to the test sequencing data, and the orange scatter points represent the test mismatch base error rate corresponding to the test sequencing data. Figure 4 The distribution scatter plot shows the corresponding absolute mean error is 0.00023, goodness of fit It is 0.92, which shows that the model testing performance of the mismatch base quality model in the current iteration process is very good.

[0110] Figure 5 A scatter plot of the test missing base error rate and the standard missing base error rate of a missing base quality model provided by one embodiment of the present invention, specifically, Figure 5 The horizontal axis of the distribution scatter plot in represents the test sequencing data, the vertical axis represents the missing base error rate, the blue scatter points represent the standard missing base error rate corresponding to the test sequencing data, and the orange scatter points represent the test missing base error rate corresponding to the test sequencing data. Figure 5 The distribution scatter plot shows the corresponding absolute mean error The goodness of fit is 0.00098. It is 0.69, which indicates that the model testing performance of the missing base quality model in the current iteration process still needs to be further improved.

[0111] S230. Determine the target parameter quality of the sequencing data to be evaluated based on the reference parameter quality.

[0112] In an optional embodiment, the target parameter quality of the sequencing data to be evaluated is determined based on the reference parameter quality, including: if the reference parameter quality is the reference base parameter error rate and the target parameter quality is the target parameter quality level, then the target parameter quality level of the sequencing data to be evaluated is determined based on the reference base parameter error rate and the grade division range corresponding to the probability density curve; wherein the probability density curve is obtained by performing probability statistics on the historical base parameter error rates of the historical sequencing data output by the sequence parameter quality model, and the grade division range is obtained according to the confidence level range.

[0113] The probability density curve is used to represent the frequency distribution of historical base parameter error rates. Specifically, the horizontal axis of the probability density curve represents the historical base parameter error rate, and the vertical axis represents the amount of historical sequencing data.

[0114] Figure 6 A schematic diagram of determining a target quality level based on a probability density curve is provided in one embodiment of the present invention. Specifically, Figure 6 The probability density curve in is based on the statistics of historical mismatch base error rates. The historical mismatch base error rates are sorted in ascending order and divided into levels according to the proportion, such as Figure 6 In the example shown, the grading ranges include 0-E90, E90-E99, and E99 and above. E90 and E99 represent confidence levels of 90% and 99%, respectively, corresponding to error rates ranked at 90% and 99%. The grading range 0-E90 corresponds to high quality, the grading range E90-E99 corresponds to medium quality, and the grading range E99 and above corresponds to low quality.

[0115] In this example, if the reference mismatch base error rate is within the historical mismatch base error rate range corresponding to the 0-E90 grade division range, the target mismatch base quality grade is a high quality grade; if the reference mismatch base error rate is within the historical mismatch base error rate range corresponding to the E90-E99 grade division range, the target mismatch base quality grade is a medium quality grade; if the reference mismatch base error rate is within the historical mismatch base error rate range corresponding to the E99 and above grade division range, the target mismatch base quality grade is a low quality grade.

[0116] There is no limit on the number of ranges for grade division and the confidence level range, and you can customize the settings according to actual needs.

[0117] In another optional embodiment, the target parameter quality of the sequencing data to be evaluated is determined based on the reference parameter quality, including: if the reference parameter quality is the reference base parameter error rate and the target parameter quality is the target parameter quality score, then the target parameter quality score of the sequencing data to be evaluated is determined based on the reference base parameter error rate and the probability density curve; wherein the probability density curve is obtained by performing probability statistics on the historical base parameter error rate of the historical sequencing data output by the sequence parameter quality model, and the target parameter quality score represents the probability that the historical base parameter error rate in the probability density curve is greater than the reference base parameter error rate.

[0118] Figure 7 A schematic diagram of determining the quality score of a target parameter based on a probability density curve is provided in one embodiment of the present invention. Specifically, Figure 7 The probability density curve in is based on the historical mismatch error rate statistics. Figure 7 The red line in the figure represents the reference mismatch error rate, and the filled area to the right of the red line represents the target mismatch quality score, which also represents the probability that the historical mismatch error rate in the probability density curve is greater than the reference mismatch error rate.

[0119] In another optional embodiment, the target parameter quality of the sequencing data to be evaluated is determined based on the reference parameter quality, including: if the reference parameter quality is the reference base parameter error rate and the target parameter quality is the target parameter quality score, then the target parameter quality score of the sequencing data to be evaluated is determined based on the maximum historical base parameter error rate, the minimum historical base parameter error rate and the reference base parameter error rate; wherein the maximum historical base parameter error rate and the minimum historical base parameter error rate are the maximum historical base parameter error rate and the minimum historical base parameter error rate respectively obtained by comparing the historical base parameter error rates of the historical sequencing data obtained according to the sequence parameter quality model.

[0120] In an optional embodiment, a target parameter quality score of the sequencing data to be evaluated is determined based on the maximum historical base parameter error rate, the minimum historical base parameter error rate, and the reference base parameter error rate, including: if the reference base parameter error rate is greater than or equal to the minimum historical base parameter error rate and less than or equal to the maximum historical base parameter error rate, then the reference base parameter error rate is normalized according to the maximum historical base parameter error rate and the minimum historical base parameter error rate to obtain the target parameter quality score of the sequencing data to be evaluated; if the target parameter quality probability is greater than the maximum historical base parameter error rate, or the target parameter quality probability is less than the minimum historical base parameter error rate, then a preset value is used as the target parameter quality score of the sequencing data to be evaluated.

[0121] Taking the reference base parameter error rate as the reference mismatch base error rate as an example, the target quality score It is expressed by the following formula:

[0122]

[0123] in, represents the maximum historical mismatch base error rate, represents the minimum historical mismatch base error rate, represents the reference mismatch base error rate.

[0124] If the target parameter quality probability is greater than the maximum historical base parameter error rate, or if the target parameter quality probability is less than the minimum historical base parameter error rate, the sequencing data being evaluated may be abnormal. For example, the preset value can be 0 or 0.1. This preset value is not limited here and can be customized according to actual needs.

[0125] Based on the above embodiment, optionally, the method further includes: obtaining at least two target parameter quality levels corresponding to the sequencing data to be evaluated, and using the minimum target parameter quality level as the sequencing quality level of the sequencing data to be evaluated; or, obtaining at least two target parameter quality scores corresponding to the sequencing data to be evaluated, and using the score statistics corresponding to the at least two target parameter quality scores as the sequencing quality scores of the sequencing data to be evaluated.

[0126] For example, assuming that the target inserted base quality level, the target mismatched base quality level, and the target deleted base quality level are high quality level, low quality level, and medium quality level, respectively, the sequencing quality level of the sequencing data to be evaluated is set to low quality level; assuming that the target inserted base quality level, the target mismatched base quality level, and the target deleted base quality level are high quality level, high quality level, and medium quality level, respectively, the sequencing quality level of the sequencing data to be evaluated is set to medium quality level.

[0127] Exemplarily, the score statistic value may be a minimum value or an average value. There is no limitation on the score statistic value here and it can be customized according to actual needs.

[0128] The advantage of this setting is that the quality of the sequencing data to be evaluated can be evaluated by integrating the variation qualities corresponding to at least two base variations, thereby further improving the accuracy of the sequencing quality results.

[0129] The technical solution of this embodiment, by inputting each sequence feature to be evaluated into a pre-trained sequence parameter quality model, obtains the reference parameter quality of the sequencing data to be evaluated. Based on the reference parameter quality, the target parameter quality of the sequencing data to be evaluated is determined. This solves the problem of the unintuitive representation of the reference parameter quality output by a single sequence parameter quality model. The use of a probability density curve can further reflect the probability of the parameter quality occurring in the sequencing data to be evaluated. The use of normalization processing can ensure the consistency of the quality scores of multiple target parameters, further improving the accuracy of subsequent sequencing quality scores.

[0130] The following is an embodiment of a sequencing quality assessment device provided by an embodiment of the present invention. This device and the sequencing quality assessment method of the above embodiment are based on the same inventive concept. For details not fully described in the embodiment of the sequencing quality assessment device, please refer to the details of the sequencing quality assessment method in the above embodiment.

[0131] Figure 8 This is a schematic diagram of the structure of a sequencing quality assessment device provided by one embodiment of the present invention. Figure 8 As shown, the apparatus includes: a module 310 for determining sequence characteristic data to be evaluated and a module 320 for determining target parameter quality.

[0132] The module 310 for determining the sequence feature data to be evaluated is used to determine the sequence feature data to be evaluated based on the sequencing data to be evaluated; wherein the sequence feature data to be evaluated includes at least one sequence feature to be evaluated;

[0133] The target parameter quality determination module 320 is used to determine the target parameter quality of the sequencing data to be evaluated based on the characteristics of each sequence to be evaluated and a pre-trained sequence parameter quality model.

[0134] Specifically, the sequencing data to be evaluated is used to represent sequencing data obtained by sequencing the nucleic acid sample to be evaluated, and the sequencing data to be evaluated includes at least one sequencing sequence and / or images respectively acquired based on at least one excitation light channel.

[0135] Exemplary, the sequencing technology used in the sequencing process includes but is not limited to Sanger sequencing technology, high-throughput sequencing technology and single-molecule sequencing technology, etc. The sequencing technology used in the sequencing process is not limited here, and can be customized according to actual needs. The sequencing method can be selected as single-end sequencing or double-end sequencing. Sequencing can be performed by a sequencing platform, and the sequencing platform can be selected but not limited to Illumina's Hiseq / Miseq / Nextseq / Novaseq sequencing platform, Thermo Fisher / Life Technologies' Ion Torrent platform, BGISEQ and MGISEQ platforms and single-molecule sequencing platforms of BGI, and Geno Lab / Geno Care / FASTA / SURF Sec sequencing platforms of Zhenmai Biotechnology, but is not limited thereto.

[0136] In this embodiment, the sequence feature data to be evaluated includes at least one sequence feature to be evaluated. Specifically, the sequence feature to be evaluated is used to represent data features obtained after performing sequence-dimensional data analysis on the sequencing data to be evaluated.

[0137] In an optional embodiment, the sequence feature data to be evaluated includes at least one of the ratio of the number of sequences of the first sequence to the total number of sequences (Output Ratio), the average base quality value of the first sequence (Clean Avg Read Score), the number of sequences of the second sequence (Good Read Num), the ratio of the number of sequences of the second sequence to the total number of sequences (Good Read Ratio), the number of base pairs of the second sequence (Good Read_Base), the average base quality value of the sequencing data to be evaluated (Raw Avg ReadScore), the single-channel base recognition ratio (Total Match Ratio), the multi-channel base recognition ratio (Total DoubleCalled Ratio), the channel cycle recognition base ratio (Base Repeat Ratio), the channel image quality (I Score) and the template density (Density).

[0138] In this embodiment, the first sequence is a sequence whose sequence read length is greater than or equal to the read length threshold. Specifically, the ratio of the number of sequences of the first sequence to the total number of sequences can be called the sequence throughput (Output Ratio), the sequence read length is used to characterize the length of the nucleic acid sequence fragment measured by sequencing the nucleic acid sample to be evaluated, and the total number of sequences is used to characterize the total number of sequencing sequences contained in the sequencing data to be evaluated. Exemplarily, the read length threshold can be 25 base pairs (base-pair, bp) or 30 base pairs. The specific parameter value of the read length threshold can be customized according to actual needs, such as the species origin of the nucleic acid sample to be tested. In some embodiments, the inventors found that when the read length threshold is set to 20-30bp, it can better reflect the sequencing quality and is associated with the sequencing error rate.

[0139] In this embodiment, the average base quality value of the first sequence is used to characterize the average value of the base quality values ​​corresponding to the first sequence in the sequencing data to be evaluated. Specifically, the base quality value (Quality Score, Q) characterizes the integer mapping of the probability of base recognition error, which can be used to quantitatively evaluate the credibility of the recognized base. In one embodiment, when the base is identified by an optical signal, it can be confirmed based on the optical signal intensity presented by the four bases in the same light spot. The optical signal intensity can be the absolute value of the optical intensity measured under certain conditions, or it can be a relative optical intensity. Exemplarily, the base quality value can be determined by one of the following methods:

[0140] 1) Among the four optical signal intensities corresponding to the four bases, take the ratio of the maximum optical signal intensity to the second maximum optical signal intensity;

[0141] 2) Among the four optical signal intensities corresponding to the four bases, determine the sum of the maximum optical signal intensity and the second largest optical signal intensity, and calculate the ratio of the maximum optical signal intensity to the sum;

[0142] 3) Among the four optical signal intensities corresponding to the four bases, take the ratio of the maximum optical signal intensity to the sum of the four optical signal intensities;

[0143] 4) Among the four optical signal intensities corresponding to the four bases, determine the difference between the maximum optical signal intensity and the second maximum optical signal intensity, as well as the sum of the four optical signal intensities, and calculate the ratio of the difference to the sum.

[0144] 5) Among the standard deviations of the four optical signal intensities corresponding to the four bases, take the ratio of the standard deviation corresponding to the maximum optical signal intensity to the standard deviation corresponding to the second largest optical signal intensity;

[0145] 6) Among the standard deviations of the four optical signal intensities corresponding to the four bases, determine the summed standard deviation between the standard deviation corresponding to the maximum optical signal intensity and the standard deviation corresponding to the second largest optical signal intensity, and calculate the ratio of the standard deviation corresponding to the maximum optical signal intensity to the summed standard deviation;

[0146] 7) Among the standard deviations of the four optical signal intensities corresponding to the four bases, take the ratio of the standard deviation corresponding to the maximum optical signal intensity to the sum of the standard deviations corresponding to the four optical signal intensities;

[0147] 8) Among the standard deviations of the four optical signal intensities corresponding to the four bases, determine the standard value of the difference between the standard deviation corresponding to the maximum optical signal intensity and the standard deviation corresponding to the second largest optical signal intensity, and the standard deviation of the sum of the standard deviations corresponding to the four optical signal intensities, and calculate the ratio of the standard deviation of the difference to the standard deviation of the sum.

[0148] The larger the above ratio is, the greater the base quality value is.

[0149] The average base quality value of the sequencing data to be evaluated in this embodiment is used to represent the arithmetic mean of the base quality values ​​corresponding to all sequencing sequences in the sequencing data to be evaluated.

[0150] In this embodiment, the second sequence is the first sequence whose base quality value is greater than or equal to the base quality threshold. For example, the base quality threshold can be 65 or 70. The specific parameter value of the base quality threshold can be customized based on actual needs, such as the species origin of the nucleic acid sample to be tested. In some embodiments, the inventors have found that setting the base quality threshold between 60 and 70 can better reflect sequencing quality and correlate with the sequencing error rate.

[0151] In this embodiment, the single-channel base recognition ratio is the ratio of the number of bright spots recognized as bases in a single excitation light channel to the total number of bright spots detected by the excitation light channel.

[0152] In some embodiments, the sequence characteristic data to be evaluated includes at least one single-channel base recognition ratio. The number of single-channel base recognition ratios is determined by the number of excitation light wavelengths used when sequencing the nucleic acid sample to be evaluated.

[0153] In one embodiment, sequencing uses dual excitation lights with different excitation wavelengths to determine base types. In this case, the sequence characteristic data to be evaluated includes the first single-channel base recognition ratio of the first excitation light channel corresponding to the first excitation wavelength, and the second single-channel base recognition ratio of the second excitation light channel corresponding to the second excitation wavelength. In another embodiment, sequencing uses three excitation lights with different excitation wavelengths to determine base types. In this case, the sequence characteristic data to be evaluated includes the first single-channel base recognition ratio of the first excitation light channel corresponding to the first excitation wavelength, the second single-channel base recognition ratio of the second excitation light channel corresponding to the second excitation wavelength, and the third single-channel base recognition ratio of the third excitation light channel corresponding to the third excitation wavelength. In another embodiment, sequencing uses four excitation lights with different excitation wavelengths to determine base types. In this case, the sequence characteristic data to be evaluated includes the first single-channel base recognition ratio of the first excitation light channel corresponding to the first excitation wavelength, the second single-channel base recognition ratio of the second excitation light channel corresponding to the second excitation wavelength, the third single-channel base recognition ratio of the third excitation light channel corresponding to the third excitation wavelength, and the fourth single-channel base recognition ratio of the fourth excitation light channel corresponding to the fourth excitation wavelength.

[0154] In this embodiment, the multi-channel base recognition ratio is the ratio of the number of bright spots recognized as bases in at least two excitation light channels to the number of templates. The number of excitation light channels in the multi-channel base recognition ratio is determined by the number of excitation light wavelengths used when sequencing the nucleic acid sample to be evaluated.

[0155] In one embodiment, sequencing uses dual excitation lights with different excitation lights to determine the base type. In this case, the multi-channel base recognition ratio is the ratio of the number of bright spots that are recognized as bases in both the first excitation light channel corresponding to the first excitation wavelength and the second excitation light channel corresponding to the second excitation wavelength to the number of templates. In another embodiment, sequencing uses three excitation lights with different excitation lights to determine the base type. In this case, the multi-channel base recognition ratio is the ratio of the number of bright spots that are recognized as bases in both the first excitation light channel corresponding to the first excitation wavelength, the second excitation light channel corresponding to the second excitation wavelength, and the third excitation light channel corresponding to the third excitation wavelength to the number of templates. In another embodiment, sequencing uses four excitation lights with different excitation lights to determine the base type. In this case, the multi-channel base recognition ratio is the ratio of the number of bright spots that are recognized as bases in both the first excitation light channel corresponding to the first excitation wavelength, the second excitation light channel corresponding to the second excitation wavelength, the third excitation light channel corresponding to the third excitation wavelength, and the fourth excitation light channel corresponding to the fourth excitation wavelength to the number of templates.

[0156] In this embodiment, the channel cycle recognition base ratio is the ratio of the number of base pairs that are continuously recognized in a preset number of rounds in a single excitation light channel to the number of templates.

[0157] In some embodiments, the sequence feature data to be evaluated includes at least one channel cycle recognition base ratio, and the specific number of channel cycle recognition base ratios is determined by the number of excitation light wavelengths used when evaluating the nucleic acid sample for sequencing.

[0158] In one embodiment, sequencing uses dual excitation lights with different excitation lights to achieve base type determination. In this case, the sequence characteristic data to be evaluated includes the first cycle identification base ratio of the first excitation light channel corresponding to the first excitation wavelength, and the second cycle identification base ratio of the second excitation light channel corresponding to the second excitation wavelength. In another embodiment, sequencing uses three excitation lights with different excitation lights to achieve base type determination. In this case, the sequence characteristic data to be evaluated includes the first cycle identification base ratio of the first excitation light channel corresponding to the first excitation wavelength, the second cycle identification base ratio of the second excitation light channel corresponding to the second excitation wavelength, and the third cycle identification base ratio of the third excitation light channel corresponding to the third excitation wavelength. In another embodiment, sequencing uses four excitation lights with different excitation lights to achieve base type determination. In this case, the sequence characteristic data to be evaluated includes the first cycle identification base ratio of the first excitation light channel corresponding to the first excitation wavelength, the second cycle identification base ratio of the second excitation light channel corresponding to the second excitation wavelength, the third cycle identification base ratio of the third excitation light channel corresponding to the third excitation wavelength, and the fourth cycle identification base ratio of the fourth excitation light channel corresponding to the fourth excitation wavelength. For example, the preset number of rounds can be 8 or 9 rounds. The specific parameter value of the preset number of rounds can be customized according to actual needs, such as the species origin of the nucleic acid sample to be tested. In some embodiments, the inventors have found that when the preset number of rounds is set to be greater than or equal to 8, it can better reflect the sequencing quality and be correlated with the sequencing error rate.

[0159] In the embodiment of the present application, the template density is the average number of templates per square micron.

[0160] In an alternative embodiment, the first excitation wavelength has a wavelength of 500 nm to 580 nm. In an alternative embodiment, the second excitation wavelength has a wavelength of 600 nm to 680 nm. In an alternative embodiment, the first excitation wavelength has a wavelength of 500 nm to 580 nm, and the second excitation wavelength has a wavelength of 600 nm to 680 nm.

[0161] In an optional embodiment, the sequencing data to be evaluated includes an image acquired based on an excitation light channel, and the sequence features to be evaluated include the channel image quality corresponding to the excitation light channel.

[0162] In an optional embodiment, the module 310 for determining the characteristic data of the sequence to be evaluated includes: a first pixel value acquisition unit, used to obtain first pixel values ​​corresponding to at least two pixels of interest in the image; a pixel score determination unit, used to obtain, for each pixel of interest, the second pixel value of the pixel adjacent to the pixel of interest, and determine the pixel score of the pixel of interest based on the first pixel value and the second pixel value; and a channel image quality determination unit, used to determine the channel image quality of the excitation light channel based on the pixel scores corresponding to each pixel of interest.

[0163] Specifically, the pixel of interest can be used to represent the pixel corresponding to the bright spot identified as the base in the image. For example, the adjacent pixel can be a pixel defined based on a 4-neighborhood or a pixel defined based on an 8-neighborhood.

[0164] In an optional embodiment, taking the adjacent pixels as pixels defined based on 8 neighborhoods as an example, the pixel scores are Satisfies the following formula:

[0165]

[0166] in, represents the first pixel value, Indicates the sum of the 8 second pixel values.

[0167] In an optional embodiment, the channel image quality determination unit is specifically configured to sort the pixel scores corresponding to each pixel of interest, and use the pixel scores corresponding to a preset value ratio in the sorted results as the channel image quality of the excitation light channel. For example, when the sorting is in ascending order, the preset value ratio may be 90%; when the sorting is in descending order, the preset value ratio may be 10%. The preset value ratio is not limited here and can be customized according to actual needs.

[0168] In another optional embodiment, the channel image quality determination unit is specifically used to: filter the pixel scores corresponding to each pixel of interest according to a preset score threshold, and use the statistical value corresponding to at least one pixel score obtained by filtering as the channel image quality of the excitation light channel.

[0169] For example, the preset score threshold may be 80%, and the statistical value may be the arithmetic mean or median. There is no limitation on the preset score threshold and the statistical value, and they can be customized according to actual needs.

[0170] Based on the above embodiment, optionally, the model architecture of the sequence parameter quality model is a linear regression model, and the device further includes: a sequence feature screening module to be evaluated, which is used to obtain feature weights corresponding to at least two sequence features to be evaluated in the sequence parameter quality model after determining the sequence feature data to be evaluated based on the sequencing data to be evaluated; and screen at least two sequence features to be evaluated in the sequence feature data to be evaluated based on the at least two feature weights.

[0171] Exemplarily, the model principle of the sequence parameter quality model can be expressed as:

[0172]

[0173] in, represents the target parameter quality, 、 … express The sequence features to be evaluated, represents the intercept term, Indicates the feature weight corresponding to each sequence feature to be evaluated.

[0174] The advantage of this setting is that it reduces the amount of input data for the sequence parameter quality model, thereby reducing the amount of calculation for the sequence parameter quality model and improving the efficiency of sequencing quality assessment.

[0175] In this embodiment, the sequence parameter quality model is at least selected from at least one of an inserted base quality model, a mismatched base quality model, and a deleted base quality model, and when the sequence parameter quality model is an inserted base quality model, the target parameter quality is the target inserted base quality, when the sequence parameter quality model is a mismatched base quality model, the target parameter quality is the target mismatched base quality, and when the sequence parameter quality model is a deleted base quality model, the target parameter quality is the target deleted base quality.

[0176] Among them, the inserted base quality is used to characterize the situation where the sequencing sequence in the sequencing data to be evaluated has one or more additional bases relative to the reference sequence; the mismatched base quality is used to characterize the situation where one or more bases in the sequencing sequence in the sequencing data to be evaluated are mismatched relative to the reference sequence; the missing base quality is used to characterize the situation where the sequencing sequence in the sequencing data to be evaluated has one or more bases missing relative to the reference sequence.

[0177] The reference sequence can be a human reference genome sequence or a reference genome sequence of another species. There is no limitation on the source of the reference sequence and it can be customized according to actual needs.

[0178] Specifically, the at least two target parameter quality expression forms can be the same or different. Exemplarily, the target parameter quality expression form can be the presence of base variation, quality probability, quality grade, or quality score. The quality probability can be an accuracy rate or an error rate, and a higher quality grade or quality score indicates better sequencing quality of the sequencing data to be evaluated. Conversely, a lower quality grade or quality score indicates worse sequencing quality of the sequencing data to be evaluated. The expression form of the target parameter quality is not limited here and can be customized according to actual needs.

[0179] Specifically, the model architectures of the at least two sequence parameter quality models may be the same or different. For example, the model architecture of the sequence parameter quality model may be a logistic regression model, a linear regression model, a decision tree, a support vector machine, a random forest, or a naive Bayes classifier. The model architecture of the sequence parameter quality model is not limited here and can be customized according to actual needs.

[0180] In an optional embodiment, the target parameter quality of the sequencing data to be evaluated is determined based on each sequence feature to be evaluated and a pre-trained sequence parameter quality model, including: inputting each sequence feature to be evaluated into the pre-trained sequence parameter quality model to obtain the output target parameter quality of the sequencing data to be evaluated.

[0181] The technical solution of this embodiment determines sequence feature data to be evaluated based on the sequencing data to be evaluated, and determines the target parameter quality of the sequencing data to be evaluated based on at least one sequence feature to be evaluated in the sequence feature data to be evaluated and a pre-trained sequence parameter quality model, wherein the sequence parameter quality model is at least selected from at least one of an inserted base quality model, a mismatched base quality model, and a deleted base quality model, and the target parameter quality includes at least one of a target inserted base quality, a target mismatched base quality, and a target deleted base quality. For the same set of sequence feature data to be evaluated for each base variation, a dedicated sequence parameter quality model is used to perform quality assessment, thereby solving the problem of unclear quality factors in traditional base quality values ​​and improving the granularity and accuracy of nucleic acid sequence quality assessment.

[0182] In an optional embodiment, the target parameter quality determination module 320 includes: a reference parameter quality output unit, which is used to input each sequence feature to be evaluated into a pre-trained sequence parameter quality model to obtain the reference parameter quality of the sequencing data to be evaluated; and a target parameter quality determination unit, which is used to determine the target parameter quality of the sequencing data to be evaluated based on the reference parameter quality.

[0183] The method for determining the “reference parameter quality” in this embodiment is the same as or similar to the method for determining the “target reference quality” in the above embodiment, and will not be described in detail in this embodiment.

[0184] Based on the above embodiment, optionally, the apparatus further includes a sequence parameter quality model training module, and the sequence parameter quality model training module includes:

[0185] A training sequence feature data determination unit is used to determine training sequence feature data based on the training sequencing data; wherein the training sequence feature data includes at least one training sequence feature;

[0186] The predicted parameter quality output unit is used to input each training sequence feature into the untrained initial parameter quality model to obtain the output predicted parameter quality of the training sequencing data;

[0187] a loss function determination unit, configured to determine a loss function based on the quality of the predicted parameters and the quality of the training parameters of the training sequencing data, and to adjust the model parameters of the initial parameter quality model based on the loss function;

[0188] The sequence parameter quality model determination unit is used to determine the trained sequence parameter quality model according to the initial parameter quality model in the current iteration process until the training end condition is met.

[0189] Specifically, the function types of the loss functions corresponding to the at least two sequence parameter quality models can be the same or different. Exemplary loss functions include squared loss function, logarithmic loss function, exponential loss function, mean squared error loss function, logistic regression loss function, Huber loss function, cross entropy loss function, and Kullback-Leibler divergence loss function. The function types of loss functions are not limited here and can be customized according to actual needs.

[0190] Specifically, the training end conditions include the number of iterations reaching an iteration threshold and / or the loss function convergence. Exemplarily, the iteration threshold can be 2000 times. The iteration threshold is not limited here and can be customized according to actual needs.

[0191] In an optional embodiment, the sequence parameter quality model determining unit is specifically configured to use the initial parameter quality model in the current iteration process as the trained sequence parameter quality model.

[0192] In another optional embodiment, the sequence parameter quality model determination unit is specifically used to: determine training sequence feature data based on test sequencing data; wherein the training sequence features include at least one test sequence feature; input each test sequence feature into the initial parameter quality model in the current iteration process to obtain the test parameter quality of the output test sequencing data; determine the model test performance based on the test parameter quality and the standard parameter quality of the test sequencing data; when the model test performance meets the preset performance conditions, use the initial parameter quality model in the current iteration process as the trained sequence parameter quality model.

[0193] In an optional embodiment, the model test performance includes absolute mean error and / or goodness of fit. The absolute mean error is used to characterize the degree of difference between the test parameter quality and the standard parameter quality, and the goodness of fit is used to characterize whether the target parameter quality model fits sufficient feature information.

[0194] For example, the mean absolute error It can be expressed by the following formula:

[0195]

[0196] For example, goodness of fit It can be expressed by the following formula:

[0197]

[0198] in, Indicates the amount of test sequencing data, Indicates the The standard parameter quality corresponding to the test sequencing data, Indicates the The quality of the test parameters corresponding to the test sequencing data, Indicates the average parameter quality corresponding to all standard parameter qualities.

[0199] Among them, the difference between the test parameter quality and the standard parameter quality can be used to characterize the characteristic information that the target parameter quality model does not fit, and the difference between the standard parameter quality and the average parameter quality can be used to characterize the characteristic information carried by the standard parameter quality. The closer it is to 1, the better the model test performance of the target reference quality model is, and the goodness of fit is The closer it is to 0, the worse the model testing performance of the target reference quality model.

[0200] Based on the above embodiment, the device optionally further includes a hyperparameter adjustment module, which is used to: when the model test performance does not meet the preset performance conditions, adjust the hyperparameters of the initial parameter quality model and repeat the training steps of the sequence parameter quality model.

[0201] In an optional embodiment, the target parameter quality determination unit includes: a target parameter quality grade determination subunit, which is used to determine the target parameter quality grade of the sequencing data to be evaluated based on the reference base parameter error rate and the grade division range corresponding to the probability density curve if the reference parameter quality is the reference base parameter error rate and the target parameter quality is the target parameter quality grade; wherein the probability density curve is obtained by performing probability statistics on the historical base parameter error rate of the historical sequencing data output by the sequence parameter quality model, and the grade division range is obtained according to the confidence level range.

[0202] The probability density curve is used to represent the frequency distribution of historical base parameter error rates. Specifically, the horizontal axis of the probability density curve represents the historical base parameter error rate, and the vertical axis represents the amount of historical sequencing data.

[0203] There is no limit on the number of ranges for grade division and the confidence level range, and you can customize the settings according to actual needs.

[0204] In another optional embodiment, the target parameter quality determination unit includes: a first target parameter quality score determination subunit, which is used to determine the target parameter quality score of the sequencing data to be evaluated based on the reference base parameter error rate and the probability density curve if the reference parameter quality is the reference base parameter error rate and the target parameter quality is the target parameter quality score; wherein the probability density curve is obtained by performing probability statistics on the historical base parameter error rate of the historical sequencing data output by the sequence parameter quality model, and the target parameter quality score represents the probability that the historical base parameter error rate in the probability density curve is greater than the reference base parameter error rate.

[0205] In another optional embodiment, the target parameter quality determination unit includes: a second target parameter quality score determination subunit, used to: if the reference parameter quality is the reference base parameter error rate and the target parameter quality is the target parameter quality score, then determine the target parameter quality score of the sequencing data to be evaluated based on the maximum historical base parameter error rate, the minimum historical base parameter error rate and the reference base parameter error rate; wherein the maximum historical base parameter error rate and the minimum historical base parameter error rate are the maximum historical base parameter error rate and the minimum historical base parameter error rate respectively obtained by comparing the historical base parameter error rates of the historical sequencing data obtained according to the sequence parameter quality model.

[0206] In an optional embodiment, the second target parameter quality score determination subunit is specifically used to: if the reference base parameter error rate is greater than or equal to the minimum historical base parameter error rate and less than or equal to the maximum historical base parameter error rate, then the reference base parameter error rate is normalized according to the maximum historical base parameter error rate and the minimum historical base parameter error rate to obtain the target parameter quality score of the sequencing data to be evaluated; if the target parameter quality probability is greater than the maximum historical base parameter error rate, or the target parameter quality probability is less than the minimum historical base parameter error rate, then the preset value is used as the target parameter quality score of the sequencing data to be evaluated.

[0207] Taking the reference base parameter error rate as the reference mismatch base error rate as an example, the target quality score It is expressed by the following formula:

[0208]

[0209] in, represents the maximum historical mismatch base error rate, represents the minimum historical mismatch base error rate, represents the reference mismatch base error rate.

[0210] If the target parameter quality probability is greater than the maximum historical base parameter error rate, or if the target parameter quality probability is less than the minimum historical base parameter error rate, the sequencing data being evaluated may be abnormal. For example, the preset value can be 0 or 0.1. This preset value is not limited here and can be customized according to actual needs.

[0211] Based on the above embodiment, the apparatus may optionally further include: a sequencing quality level determination module, configured to obtain at least two target parameter quality levels corresponding to the sequencing data to be evaluated, and use the minimum target parameter quality level as the sequencing quality level of the sequencing data to be evaluated; or a sequencing quality score determination module, configured to obtain at least two target parameter quality scores corresponding to the sequencing data to be evaluated, and use the score statistics corresponding to the at least two target parameter quality scores as the sequencing quality score of the sequencing data to be evaluated.

[0212] For example, assuming that the target inserted base quality level, the target mismatched base quality level, and the target deleted base quality level are high quality level, low quality level, and medium quality level, respectively, the sequencing quality level of the sequencing data to be evaluated is set to low quality level; assuming that the target inserted base quality level, the target mismatched base quality level, and the target deleted base quality level are high quality level, high quality level, and medium quality level, respectively, the sequencing quality level of the sequencing data to be evaluated is set to medium quality level.

[0213] Exemplarily, the score statistic value may be a minimum value or an average value. There is no limitation on the score statistic value here and it can be customized according to actual needs.

[0214] The advantage of this setting is that the quality of the sequencing data to be evaluated can be evaluated by integrating the variation qualities corresponding to at least two base variations, thereby further improving the accuracy of the sequencing quality results.

[0215] The technical solution of this embodiment, by inputting each sequence feature to be evaluated into a pre-trained sequence parameter quality model, obtains the reference parameter quality of the sequencing data to be evaluated. Based on the reference parameter quality, the target parameter quality of the sequencing data to be evaluated is determined. This solves the problem of the unintuitive representation of the reference parameter quality output by a single sequence parameter quality model. The use of a probability density curve can further reflect the probability of the parameter quality occurring in the sequencing data to be evaluated. The use of normalization processing can ensure the consistency of the quality scores of multiple target parameters, further improving the accuracy of subsequent sequencing quality scores.

[0216] The sequencing quality assessment device provided in the embodiment of the present invention can execute the sequencing quality assessment method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0217] Figure 9 This is a schematic diagram of the structure of an electronic device provided in accordance with one embodiment of the present invention. The electronic device 10 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are provided for illustrative purposes only and are not intended to limit the implementation of the present inventions described and / or claimed herein.

[0218] like Figure 9As shown, electronic device 10 includes at least one processor 11 and memory, such as read-only memory (ROM) 12 and random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor 11. Processor 11 can perform various appropriate actions and processes based on the computer programs stored in ROM 12 or loaded from storage unit 18 into RAM 13. RAM 13 can also store various programs and data required for the operation of electronic device 10. Processor 11, ROM 12, and RAM 13 are interconnected via bus 14. An input / output (I / O) interface 15 is also connected to bus 14.

[0219] Multiple components in the electronic device 10 are connected to the I / O interface 15, including an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a magnetic disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information or data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0220] The processor 11 can be any general-purpose and / or specialized processing component with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors that run machine learning model algorithms, a digital signal processor (DSP), and any other suitable processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the sequencing quality assessment method provided in the above embodiments.

[0221] In some embodiments, the sequencing quality assessment methods provided in the above embodiments can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps in the sequencing quality assessment method described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to execute the sequencing quality assessment method in any other appropriate manner (e.g., via firmware).

[0222] Various embodiments of the systems and techniques described herein can be implemented in the following systems or combinations thereof: digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard parts (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0223] Computer programs for implementing the sequencing quality assessment methods of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer program implements the functions / operations specified in the flowcharts and / or block diagrams. The computer program can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0224] In the context of this application, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, device, or apparatus. A computer-readable storage medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or apparatus, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable storage medium. Examples of machine-readable storage media may include an electrical connection based on at least one wire, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0225] To provide interaction with a user, the systems and techniques described herein can be implemented on a terminal device having: a display device (e.g., a cathode ray tube (CRT) or a liquid crystal display (LCD) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the terminal device. Other types of devices can also provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0226] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: Local Area Networks (LANs), Wide Area Networks (WANs), blockchain networks, and the Internet.

[0227] A computing system may include clients and servers. The clients and servers are typically remote from each other and typically interact via a communications network. This client-server relationship arises through computer programs running on the respective computers, creating a client-server relationship. The server may be a cloud server, also known as a cloud computing server or cloud host. This is a host product within the cloud computing service ecosystem that addresses the management difficulties and limited scalability of traditional physical hosts and virtual private server (VPS) services.

[0228] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in the present invention can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved. This is not limited herein.

[0229] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A sequencing quality assessment method, characterized in that: include: Determining sequence feature data to be evaluated according to the sequencing data to be evaluated; wherein the sequence feature data to be evaluated includes at least one sequence feature to be evaluated, and the sequence feature to be evaluated represents a data feature obtained after performing data analysis processing in the sequence dimension on the sequencing data to be evaluated; Determining the target parameter quality of the sequencing data to be evaluated according to the characteristics of each sequence to be evaluated and a pre-trained sequence parameter quality model; Wherein, the sequence parameter quality model is at least selected from at least two of an insertion base quality model, a mismatch base quality model and a deletion base quality model, and when the sequence parameter quality model is an insertion base quality model, the target parameter quality is a target insertion base quality, when the sequence parameter quality model is a mismatch base quality model, the target parameter quality is a target mismatch base quality, and when the sequence parameter quality model is a deletion base quality model, the target parameter quality is a target deletion base quality; The method further comprises: The statistical values ​​of at least two target parameter qualities corresponding to the sequencing data to be evaluated are used as the sequencing quality of the sequencing data to be evaluated.

2. The method according to claim 1, characterized in that The sequence feature data to be evaluated includes at least one of the ratio of the number of sequences of the first sequence to the total number of sequences, the average base quality value of the first sequence, the number of sequences of the second sequence, the ratio of the number of sequences of the second sequence to the total number of sequences, the number of base pairs of the second sequence, the average base quality value of the sequencing data to be evaluated, the ratio of single-channel base recognition, the ratio of multi-channel base recognition, the ratio of channel cycle recognition bases, channel image quality and template density; Among them, the first sequence is a sequence whose sequence read length is greater than or equal to the read length threshold, the second sequence is a first sequence whose base quality value is greater than or equal to the base quality threshold, the single-channel base recognition ratio is the ratio of the number of bright spots recognized as bases in a single excitation light channel to the total number of bright spots detected by the excitation light channel, the multi-channel base recognition ratio is the ratio of the number of bright spots recognized as bases in at least two excitation light channels to the number of templates, and the channel cycle recognition base ratio is the ratio of the number of base pairs continuously recognized for a preset number of rounds in a single excitation light channel to the number of templates.

3. The method according to claim 2, characterized in that The single-channel base recognition ratio is a first single-channel base recognition ratio of a first excitation light channel corresponding to a first excitation wavelength or a second single-channel base recognition ratio of a second excitation light channel corresponding to a second excitation wavelength; The multi-channel base recognition ratio is the ratio of the number of bright spots recognized as bases in both the first excitation light channel and the second excitation light channel to the number of templates.

4. The method according to claim 3, characterized in that The first excitation wavelength has a wavelength of 500 nm to 580 nm, and / or the second excitation wavelength has a wavelength of 600 nm to 680 nm.

5. The method according to claim 1, characterized in that The sequencing data to be evaluated includes an image acquired based on an excitation light channel, and the sequence features to be evaluated include a channel image quality corresponding to the excitation light channel.

6. The method according to claim 5, characterized in that Determining the sequence feature data to be evaluated based on the sequencing data to be evaluated includes: Obtaining first pixel values ​​respectively corresponding to at least two pixels of interest in the image; For each pixel of interest, obtain a second pixel value of a pixel adjacent to the pixel of interest, and determine a pixel score of the pixel of interest according to the first pixel value and the second pixel value; The channel image quality of the excitation light channel is determined according to the pixel points scores respectively corresponding to the pixels of interest.

7. The method according to claim 6, characterized in that The pixel score Satisfies the following formula: in, represents the first pixel value, Represents the sum of the 8 second pixel values.

8. The method according to any one of claims 1 to 7, characterized in that: The model architecture of the sequence parameter quality model is a linear regression model. After determining the sequence feature data to be evaluated based on the sequencing data to be evaluated, the method further includes: Obtaining feature weights corresponding to at least two sequence features to be evaluated in the sequence parameter quality model; At least two sequence features to be evaluated in the sequence feature data to be evaluated are screened according to at least two feature weights.

9. The method according to claim 1, characterized in that: Determining the target parameter quality of the sequencing data to be evaluated according to the characteristics of each sequence to be evaluated and a pre-trained sequence parameter quality model includes: Inputting each of the sequence features to be evaluated into a pre-trained sequence parameter quality model to obtain the reference parameter quality of the sequencing data to be evaluated as an output; According to the reference parameter quality, the target parameter quality of the sequencing data to be evaluated is determined.

10. The method according to claim 9, characterized in that Determining the target parameter quality of the sequencing data to be evaluated according to the reference parameter quality includes: If the reference parameter quality is a reference base parameter error rate and the target parameter quality is a target parameter quality grade, determining the target parameter quality grade of the sequencing data to be evaluated according to the reference base parameter error rate and the grade classification range corresponding to the probability density curve; The probability density curve is obtained by performing probability statistics on the historical base parameter error rate of the historical sequencing data output by the sequence parameter quality model, and the grading range is obtained according to the confidence level range.

11. The method according to claim 9, characterized in that Determining the target parameter quality of the sequencing data to be evaluated according to the reference parameter quality includes: If the reference parameter quality is a reference base parameter error rate and the target parameter quality is a target parameter quality score, determining the target parameter quality score of the sequencing data to be evaluated according to the reference base parameter error rate and a probability density curve; Among them, the probability density curve is obtained by performing probability statistics on the historical base parameter error rate of the historical sequencing data output by the sequence parameter quality model, and the target parameter quality score represents the probability that the historical base parameter error rate in the probability density curve is greater than the reference base parameter error rate.

12. The method according to claim 9, characterized in that Determining the target parameter quality of the sequencing data to be evaluated according to the reference parameter quality includes: If the reference parameter quality is a reference base parameter error rate and the target parameter quality is a target parameter quality score, determining the target parameter quality score of the sequencing data to be evaluated according to the maximum historical base parameter error rate, the minimum historical base parameter error rate and the reference base parameter error rate; Among them, the maximum historical base parameter error rate and the minimum historical base parameter error rate are the maximum historical base parameter error rate and the minimum historical base parameter error rate respectively obtained by comparing and statistically analyzing the historical base parameter error rates of historical sequencing data obtained according to the sequence parameter quality model.

13. The method according to claim 12, characterized in that Determining the target parameter quality score of the sequencing data to be evaluated according to the maximum historical base parameter error rate, the minimum historical base parameter error rate, and the reference base parameter error rate includes: If the reference base parameter error rate is greater than or equal to the minimum historical base parameter error rate and less than or equal to the maximum historical base parameter error rate, then the reference base parameter error rate is normalized according to the maximum historical base parameter error rate and the minimum historical base parameter error rate to obtain a target parameter quality score for the sequencing data to be evaluated; If the target parameter quality probability is greater than the maximum historical base parameter error rate, or the target parameter quality probability is less than the minimum historical base parameter error rate, the preset value is used as the target parameter quality score of the sequencing data to be evaluated.

14. The method according to any one of claims 10 to 12, characterized in that: The step of taking the statistical values ​​of at least two target parameter qualities corresponding to the sequencing data to be evaluated as the sequencing quality of the sequencing data to be evaluated comprises: Obtaining at least two target parameter quality levels corresponding to the sequencing data to be evaluated, and taking the minimum target parameter quality level as the sequencing quality level of the sequencing data to be evaluated; or, At least two target parameter quality scores corresponding to the sequencing data to be evaluated are obtained, and score statistics corresponding to the at least two target parameter quality scores are used as sequencing quality scores of the sequencing data to be evaluated.

15. The method according to claim 1, characterized in that The method further comprises: Determine training sequence feature data according to the training sequencing data; wherein the training sequence feature data includes at least one training sequence feature; Inputting each of the training sequence features into an untrained initial parameter quality model to obtain the predicted parameter quality of the output training sequencing data; Determining a loss function according to the prediction parameter quality and the training parameter quality of the training sequencing data, and adjusting the model parameters of the initial parameter quality model according to the loss function; When the training end condition is met, the training completed sequence parameter quality model is determined according to the initial parameter quality model in the current iteration process.

16. The method according to claim 15, characterized in that The step of determining the trained sequence parameter quality model according to the initial parameter quality model in the current iteration process includes: Determine test sequence feature data according to the test sequencing data; wherein the test sequence feature includes at least one test sequence feature; Inputting each of the test sequence features into the initial parameter quality model in the current iteration process to obtain the test parameter quality of the output test sequencing data; Determining model test performance according to the test parameter quality and the standard parameter quality of the test sequencing data; When the model test performance meets the preset performance conditions, the initial parameter quality model in the current iteration process is used as the trained sequence parameter quality model.

17. The method according to claim 16, characterized in that The model test performance includes absolute mean error and / or goodness of fit.

18. A sequencing quality assessment device, characterized in that: include: A module for determining feature data of a sequence to be evaluated, used to determine feature data of a sequence to be evaluated according to sequencing data to be evaluated; wherein the feature data of the sequence to be evaluated includes at least one feature of the sequence to be evaluated, and the feature of the sequence to be evaluated represents a data feature obtained after data analysis processing of the sequencing data to be evaluated in the sequence dimension; A target parameter quality determination module, used to determine the target parameter quality of the sequencing data to be evaluated according to the characteristics of each sequence to be evaluated and a pre-trained sequence parameter quality model; Wherein, the sequence parameter quality model is at least selected from at least two of an insertion base quality model, a mismatch base quality model and a deletion base quality model, and when the sequence parameter quality model is an insertion base quality model, the target parameter quality is a target insertion base quality, when the sequence parameter quality model is a mismatch base quality model, the target parameter quality is a target mismatch base quality, and when the sequence parameter quality model is a deletion base quality model, the target parameter quality is a target deletion base quality; The device also includes: The sequencing quality determination module is used to use the statistical values ​​of at least two target parameter qualities corresponding to the sequencing data to be evaluated as the sequencing quality of the sequencing data to be evaluated.

19. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can perform the sequencing quality assessment method according to any one of claims 1 to 17.

20. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the sequencing quality assessment method according to any one of claims 1 to 17 when executed.

21. A computer program product, comprising a computer program, which, when executed by a processor, implements the sequencing quality assessment method according to any one of claims 1 to 17.