Methods and systems for determining population size of nucleic acid fragments using digital assays
Through digital experiments and logistic regression models, the sensitivity and automation problems of nucleic acid fragment size determination in the prior art were solved, and the precise detection of DNA residues in biological products was achieved, which met the standards of FDA and WHO.
Patent Information
- Application Number
- CN202380042337.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2022-03-24
- Filing Date
- 2023-03-24
- Publication Date
- 2025-01-03
AI Technical Summary
The existing technology is used to determine the size of nucleic acid fragments in biological products, which have problems such as poor sensitivity and specificity, low throughput, high cost and difficulty in automation, making it difficult to effectively detect and meet the limitations of residual DNA by FDA and WHO.
Using digital experimental methods and systems, the target amplification reaction is carried out by forming partitions, and the target-defined subsets are enumerated using amplification data to predict the length characteristics of the nucleic acid fragment population. The logistic regression model is used to generate a size prediction algorithm to achieve the nucleic acid Accurate determination of fragment size.
It improves the sensitivity and accuracy of nucleic acid fragment size determination, reduces the sample usage, is suitable for automated processing, and generates a more accurate size prediction algorithm through machine learning, meeting the detection requirements for DNA residues in biological products.
Smart Images

Figure CN119256095A_ABST
Abstract
Description
Cross-reference to prior applications
[0001] This application is based upon and claims the benefit of 35 U.S.C. §119(e) U.S. Provisional Patent Application Serial No. 63 / 323,314 filed on March 24, 2022, which is incorporated herein by reference in its entirety for all purposes. introduction
[0002] Biological products (such as monoclonal antibodies, therapeutic proteins, hormones, etc.) can be biosynthesized on a commercial scale by cells cultured in bioreactors. After biosynthesis, the biological product is separated from the cells and then processed and purified to remove potentially harmful by-products. One of these by-products is fragmented DNA, which may be carcinogenic, immunogenic or infectious in some cases. In order to minimize the risk of application to humans, the U.S. Food and Drug Administration (FDA) and the World Health Organization (WHO) have restricted the content (e.g., less than 10 ng / dose) and size (e.g., median length less than 200 base pairs) of residual DNA present in biological products. The size restriction makes it unlikely that the residual DNA contains the entire gene. Each batch of biological products must be tested to ensure that the residual DNA does not exceed the size limit. However, there are various disadvantages in the prior art for measuring the size of nucleic acid fragments in samples, such as poor sensitivity and specificity, low throughput, high cost and / or automation difficulties. Summary of the invention
[0003] The present disclosure provides methods, systems and computer-readable media for determining the size of nucleic acid fragment populations. In an exemplary method for determining the size of nucleic acid fragment populations provided by a sample, partitions can be formed, each containing a portion of a sample. Each target in at least two targets can be present only in a fraction of nucleic acid fragments. Amplification reactions can be performed on each of at least two targets in a partition. Amplification data of at least two targets can be collected from a partition. Amplification data can be used to enumerate a subset of target limits of a partition. Enumeration results can be used to predict at least one length feature of a nucleic acid fragment population. BRIEF DESCRIPTION OF THE DRAWINGS
[0004] Figure 1 is a schematic flow chart presenting selected aspects of an illustrative method and system for determining the size of a population of nucleic acid fragments using a digital assay for first and second targets of varying lengths, the digital assay being performed with a fluid containing a sample wherein the nucleic acid fragments are substantially larger than each target.
[0005] Figure 2 It is presented Figure 1Another schematic flow diagram of selected aspects of methods and systems of the invention, which are performed using a different sample-containing fluid that is more fragmented so that the nucleic acid fragments are generally comparable in size to the first and second targets.
[0006] Figure 3 The degree of sample fragmentation is shown Figure 1 and Figure 2 A graph of an exemplary exponential relationship between target level ratios of a first and a second target.
[0007] Figure 4 yes Figure 1 Schematic diagram of an exemplary configuration of targets and reagents for methods and systems obtained in the presence of a nucleic acid fragment containing a first target and a second target linked to each other, and showing amplification of the two targets.
[0008] Figure 5 yes Figure 4 Another schematic diagram of a setup of , except that there is a nucleic acid fragment containing only the second target.
[0009] Figure 6 yes Figure 4 Yet another schematic diagram of a setup of , except that there is a nucleic acid fragment containing only the first target.
[0010] Figure 7 is a schematic flow chart showing a method for using Figure 1 Selected aspects of illustrative methods and systems for determining the size of a nucleic acid fragment population using a similar digital assay, except that Figure 4 The ligation targets are set and performed with a fluid containing a sample, wherein the nucleic acid fragments are much larger than each target.
[0011] Figure 8 is used Figure 4 and Figure 7 Another schematic flow chart of a digital assay and ligation target of FIG. 1 , except that it is performed using a fluid containing a sample that exhibits a greater amount of fragmentation.
[0012] Figures 9 to 16 is a scatter plot of amplification data collected from a digital amplification experiment performed in droplets (as partitions), using Figure 4-6 The targets and reagents are configured on a series of standards, each standard having a different median length nucleic acid fragment measured separately.
[0013] Fig.17 is drawn by Figure 9-16 The scatter plot presents the target concentration determined from the amplification data as a function of the median fragment length of each standard in a bar graph.
[0014] Fig.18 The following is a plot of size predictions for Benzonase fragmentation samples, where the size predictions were generated using a machine learning model based on logistic regression.
[0015] Fig.19 is a flow chart listing exemplary steps of an illustrative method for determining the size of a population of nucleic acid fragments using a digital assay.
[0016] Fig. 20 is a flow chart listing exemplary steps for an illustrative method of generating a size prediction algorithm for determining the size of a population of nucleic acid fragments using digital experiments. DETAILED DESCRIPTION
[0017] The following describes various aspects and examples of methods for measuring the size of nucleic acid fragment colonies and related systems and computer-readable media, and is illustrated in the related drawings. Unless otherwise stated, the method, system or computer-readable medium for measuring the size of nucleic acid fragments comprises at least one of the structures, components, functions and / or variations described, illustrated and / or incorporated herein. In addition, unless specifically excluded, the processing steps, structures, components, functions and / or variations described, illustrated and / or incorporated herein may be included in other similar methods, systems and / or computer-readable media, including interchangeable between disclosed examples. The following description of various examples is essentially merely illustrative and is by no means intended to limit examples, their applications or their uses. In addition, the advantages provided by the examples described below are essentially illustrative, and not all examples provide the same advantages or the same degree of advantages.
[0018] The present disclosure provides methods, systems and computer readable media for determining the size of nucleic acid fragment populations using digital experiments. Other aspects of the present disclosure are described in the following subsections: (I) Definitions, (II) Overview, (III) Examples, Components and Alternatives, (IV) Illustrative Combinations and Additional Examples, (V) Advantages and Benefits, and (VI) Conclusion.
[0019] The features, functions, and advantages can be achieved independently in various examples of the present disclosure or may be combined in yet other examples, further details of which can be seen with reference to the following description and drawings. I. Definitions
[0020] The technical terms used in the present disclosure have meanings recognized by those skilled in the art. However, the following terms may be further defined as follows.
[0021] An "amplicon" is the product of an amplification reaction. A copy of an amplicon can be generated by amplifying a target sequence so that the amplicon corresponds to the target sequence (i.e., matches and / or is complementary to the target sequence). However, the sequence of the amplicon, such as the sequence at the primer binding site, may not completely match and / or be not completely complementary to the target sequence.
[0022] "Amplification" is a process in which multiple copies are prepared by matching, complementing and / or otherwise corresponding to the amplicon with the target sequence. The process is interchangeably referred to as an amplification reaction. As the amplification proceeds, the amplification may produce an exponential increase in the number of copies. Typical amplification may increase the number of copies of the amplicon to greater than 100 times or 1,000 times. Exemplary amplification reactions of the methods disclosed herein may include polymerase chain reaction (PCR) or ligase chain reaction (LCR), each of which is driven by thermal cycling. The method may also use or alternatively use other amplification reactions, which may be carried out under isothermal conditions, such as branched probe DNA assays, cascade RCA, helicase-dependent amplification, loop-mediated isothermal amplification (LAMP), nucleic acid-based amplification (NASBA), nickase amplification reaction (NEAR), PAN-AC, Q-β replicase amplification, rolling circle replication (RCA), self-sustaining sequence replication, strand displacement amplification and / or similar means. Amplification may utilize linear or circular templates.
[0023] An "amplifier" is any reagent that promotes the generation of an amplicon by amplifying a target sequence. The reagent may include any combination of the following: at least one pair of primers for amplifying at least one target sequence, at least one marker for detecting the amplification of the at least one target (e.g., at least one probe including a marker and / or an intercalating dye as a marker), at least one polymerase and / or ligase (which may be thermostable), and nucleoside triphosphates (dNTPs and / or NTPs), etc.
[0024] A "carrier fluid" is a fluid that contacts the partitions, optionally surrounding the partitions. The fluid may be a liquid or a gas. The carrier fluid may be described as a continuous phase, and wherein the partitions are described as dispersed phases. The carrier fluid may be immiscible with the partitions and encapsulate the partitions. In some instances, the carrier fluid may be an oil, such as, for example, a fluorocarbon oil or a silicone oil.
[0025] "Complementary" means involving the rules of base pairing. If the first nucleic acid polymer or its region can hybridize with the second nucleic acid polymer in an antiparallel manner by forming a continuous (uninterrupted) or almost continuous base pair series (e.g., at least 5, 6, 7, 8, 9 or 10 continuous base pairs), the first nucleic acid polymer or its region is "complementary" to the second nucleic acid polymer. If the hybridization of the first nucleic acid polymer (or its region) with the second nucleic acid polymer forms a continuous base pair series with each nucleotide of the first nucleic acid polymer or its region, the first nucleic acid polymer (or its region) is referred to as "completely complementary" to the second nucleic acid polymer. The "complement" of the first nucleic acid polymer or its region is a second nucleic acid polymer or its region that is completely complementary to the first nucleic acid polymer or its region. The "complementarity" between the first nucleic acid polymer (or its region) and the second nucleic acid polymer (or its region) refers to the number or percentage of base pairs that can be formed when the first nucleic acid polymer (or its region) is optimally arranged to hybridize with the second nucleic acid polymer (or its region) in an antiparallel manner. The first nucleic acid polymer or its region that is complementary to the second nucleic acid polymer or its region generally has a complementarity of at least 80% or 90%.
[0026] “Comprising,” “including,” and “having” (and variations thereof) are used interchangeably and mean including but not necessarily limited to, and are open-ended terms that are not intended to exclude additional, unrecited elements or method steps.
[0027] "Computer-readable media" includes any removable / non-removable, volatile / non-volatile storage media that is readable by a machine. Exemplary types of computer-readable media include magnetic media, optical media, and electronic media, such as hard drives, optical disks, magnetic tapes, disks, solid-state random access memory (RAM), solid-state read-only memory (ROM), etc. Computer-readable media may be transient or non-transient.
[0028] "Coupled" refers to a relationship whereby the properties of one affect the properties of another, and may include permanent or releasable connection, whether direct or indirect through intermediate components, and is not necessarily limited to physical connections.
[0029] A "digital assay" is a research procedure that is capable of detecting a single copy of an analyte in a collection of partitions. A "digital amplification assay" is a digital assay that utilizes an amplification reaction to facilitate the detection of a single copy of a target. A digital assay can be performed with any suitable number of partitions that give a statistically significant result, such as at least twenty, one hundred, one thousand, or ten thousand, etc.
[0030] A "droplet" is a small volume of liquid surrounded by an immiscible liquid (e.g., encapsulated by an immiscible liquid that may form a continuous phase of an emulsion). The immiscible liquid may include oil and / or may consist primarily of oil. The droplets disclosed herein may, for example, have an average volume of less than about 500 nL, 100 nL, 10 nL, or 1 nL, etc.
[0031] An "enumerated value" is any value generated by enumerating a subset of a target definition of a partition set in a digital assay. For example, an enumerated value may represent the number of partitions in a partition set that are positive for the presence of a given target or two or more given targets, negative for the presence of a given target or two or more given targets, positive only for a specified subset of one or more targets in a target set, negative only for a specified subset of one or more targets in a target set, etc.
[0032] “Exemplary” means “illustrative” or “serving as an example.” Likewise, the term “example” (or “exemplary”) means “illustrative.” Neither word implies preference or superiority.
[0033] "First," "second," and similar terms are used to distinguish or identify individual members of a group, etc., in the order in which they are introduced in a particular context, and are not intended to indicate an order or numerical limitation.
[0034] "Fluorescence" is optical radiation emitted in response to the absorption of light. As used herein, fluorescence is intended to encompass any form of photoluminescence in which the absorption of one or more photons promotes an electron to an excited state and results in the subsequent emission of new photons, whether from a singlet, triplet, or other state. The excited state resulting from the absorption may have any suitable lifetime.
[0035] "Fluorophore" refers to any atom, functional group, molecule or substance that is capable of fluorescing.
[0036] "Fragmentation" or "fragmenting" of nucleic acid polymers (DNA and / or RNA) is the process of breaking nucleic acid polymers into smaller fragments (called "fragments"), which are also nucleic acid polymers (or monomers). The nucleic acid polymer state produced by this process is called "fragmentation" or "degradation". Fragmentation / fragmentation can be carried out by physical processes, chemical processes or combinations thereof. Physical fragmentation is usually carried out by applying a force sufficient to cause fracture to the nucleic acid polymer. These forces can be fluid dynamic shear forces, which can be generated by any suitable method, such as by ultrasonic treatment, acoustic energy pulses, centrifugation, passing through a needle, point slot shearing, etc. Chemical fragmentation is carried out by a chemical reaction using nucleic acid polymers as reactants. The chemical reaction can be enzyme-catalyzed, such as with an endonuclease. Exemplary chemical fragmentation is carried out by a hydrolysis reaction of the phosphodiester backbone of the cleavage nucleic acid polymer.
[0037] The "inverse" of a subset of a set of partitions is the remainder of the set. For example, if a first subset of a set of partitions is positive for a first target and negative for a second target, then the inverse of the first subset is composed of all other partitions of the set, i.e., partitions where the first target and the second target are negative, partitions where the first target and the second target are positive, and partitions where the first target is negative and the second target is positive. Similarly, if a second subset of a set of partitions is positive for the second target and negative for the first target, then the inverse of the second subset is composed of all other partitions of the set, i.e., partitions where the first target and the second target are negative, partitions where the first target and the second target are positive, and partitions where the first target is negative and the second target is positive. Furthermore, if each partition of the third subset of the set of partitions is positive for both the first target and the second target, then the reciprocal of the third subset includes all other partitions of the set, i.e., partitions where the first target and the second target are negative, partitions where the first target is positive and the second target is negative, and partitions where the first target is negative and the second target is positive. The partition count of the reciprocal of a given subset of the set of partitions is equal to the total partition count of the set of partitions minus the partition count of the given subset.
[0038] "Label" is a marker or identifier associated with a structure (such as a primer, probe, amplicon, partition, etc.) for identification and / or differentiation. The label can be covalently linked to the structure (such as a label coupled to an oligonucleotide) or non-covalently linked (e.g., by embedding, hydrogen bonding, electrostatic interaction, encapsulation, etc.). Exemplary labels include optical labels, radioactive labels, magnetic labels, electrical labels, epitopes, enzymes, antibodies, etc. Optical labels can be optically detected by their interaction with light. Possibly suitable exemplary optical labels include fluorophores and quenchers, etc.
[0039] "Length characteristic" is any measurement or description of the length of a nucleic acid fragment population. For example, a length characteristic can be expressed relative to base pairs / nucleotides, molecular weight, nanometers, etc., and can be the average length, median length, quartile length, or quintile length of nucleic acid fragments in a population. In some cases, a length characteristic can be a classification about one or more threshold sizes, such as a binary classification about a threshold size, such as 200 nucleotides (for example, less than (below), or greater than (above), or greater than or equal to a threshold size). In some cases, a length characteristic can be a fragment percentage (in terms of quantity or quality) above or below the threshold size. The term "size" can be used as a synonym for length in this article.
[0040] The method disclosed herein can predict any suitable combination of the length characteristics of nucleic acid fragment colonies. The combination of length characteristics can include median size estimation (e.g., in bp or nucleotides). The combination can also include an estimated confidence interval of the median (such as 90%, 95% or 99% confidence interval of the median), which can be calculated using Poisson statistics to provide the upper and lower limit input values of the prediction function also used to estimate the median size. The combination can also or alternatively include a percentage value of the nucleic acid fragment colony, which is less than or greater than (or equal to) a threshold size, such as 200bp. The combination can further include a confidence interval (such as 90%, 95% or 99% confidence interval) less than or greater than the threshold size.
[0041] A "prediction function" is any function that provides an output value of a dependent variable based on one or more input values of one or more independent variables, wherein the output value represents a prediction from the one or more input values. A prediction function may include an expression having any suitable number of variable terms, coefficients for each variable term, and a constant term. The expression may include 1, 2, 3, or more independent variables, an equal or greater number of variable terms including the independent variables, and a constant term. The values of the coefficients and constant term of the prediction function may be estimated by machine learning using a computer and a data set containing data points, and the prediction function is fitted by estimating the values of the coefficients and constant term of the prediction function.
[0042] The "Linear Prediction Function" has the following general form: y=β1x1+…+β p x p +b (1) The expression on the right side of the equal sign has p variables and one constant term, where x k (k=1,...,p) is the kth independent (input / explanatory) variable β1,....,β p is the coefficient corresponding to the independent variable (and variable term), b is the constant term (intercept), and y is the dependent variable (output). Given data point i, the independent variable value is xi1 ,...,x ip , the dependent variable value of data point i is yi. The linear prediction function of the present disclosure may have only one (p=1), only two (p=2), only three (p=3) or more (p>3) independent variables. A linear prediction function with two or more independent variables is a linear combination of the independent variables. The input values of each variable are multiplied by the corresponding coefficients, and then the products are added (and the intercept) to obtain the value of the dependent variable. The value of the corresponding coefficient of the independent variable can be estimated based on a data set including data points, where each data point has the value of the dependent variable and the associated value of the respective variable. The process of estimating the coefficients and / or intercepts can be performed by machine learning.
[0043] "Linear regression" is a linear method for modeling the relationship between a dependent variable (output variable (e.g., scalar response)) and one or more independent variables (input variables (also called explanatory variables)). Simple linear regression is performed using a linear prediction function with one independent variable, while multiple linear regression is performed using a linear prediction function with two or more independent variables. Linear regression creates a linear regression model by fitting a linear prediction function to a data set containing data points, where each data point has a value for the dependent variable and an associated value for the respective variable. The fitting process may involve estimating the values of the coefficients and the intercept value of the linear prediction function (such as by the method of least squares).
[0044] "Logistic regression" is a method of modeling the relationship between a dependent variable (output variable) and one or more independent variables (input variables) that have only a limited number of allowed values. Binary regression is a logistic regression in which the dependent variable is a binary variable that has only two allowed values (e.g., 0 or 1). Logistic regression can create a logistic regression model by fitting a linear prediction function of a general form to a data set containing data points, where each data point has a value for the dependent variable and an associated value for each variable. The linear prediction function can be a linear combination of weighted input variables. The fitting process can involve estimating the values of each coefficient and the intercept value of the linear prediction function (such as by maximum likelihood estimation (MLE)).
[0045] "Machine learning" is any computer-implemented program or set of programs that can generate a prediction algorithm based on collected data. Machine learning can involve estimating the values of the coefficients of a prediction function and the value of a constant (e.g., an intercept).
[0046] A "nucleic acid polymer" is a molecule or molecular duplex of any length comprising naturally occurring nucleotides (e.g., the polymer is DNA and / or RNA), or a synthetically produced compound that can hybridize with DNA or RNA in a sequence-specific manner similar to two naturally occurring nucleic acids, e.g., can participate in Watson-Crick base pairing interactions. Nucleic acid polymers can contain any suitable number (such as at least about 5, 10, 100, or 1000, etc.) of nucleotides.
[0047] Nucleic acid polymers can have natural or artificial structures, or a combination of the two. Nucleic acid polymers with natural structures, i.e., deoxyribonucleic acid (DNA) and ribonucleic acid (RNA), generally have a backbone of alternating pentose groups and phosphate groups. Each pentose group is connected to a core base (e.g., purine (e.g., adenine (A) or guanine (G)) or a pyrimidine (e.g., cytosine (C), thymine (T) or uracil (U))). Nucleic acid polymers with artificial structures are analogs of natural nucleic acids, and can, for example, be produced by changing the pentose and / or phosphate groups and / or one or more core bases of the natural backbone. Exemplary artificial nucleic acid polymers include glycol nucleic acids (GNA), peptide nucleic acids (PNA), locked nucleic acids (LNA), threose nucleic acids (TNA), heterologous nucleic acids (XNA), etc.
[0048] The sequence of a nucleic acid polymer is determined by the order in which the nucleobases are arranged along the backbone. This order generally determines the ability of a nucleic acid polymer to hybridize with another nucleic acid through hydrogen bonding. Specifically, adenine pairs with thymine (or uracil), and guanine pairs with cytosine.
[0049] "Oligonucleotide" is a relatively short and / or chemically synthesized nucleic acid polymer. For example, the length of the oligonucleotide can be 3 to 1000 nucleotides, etc. In some cases, the oligonucleotide can be labeled with at least one marker, which can be coupled to any suitable structure of the oligonucleotide. The at least one marker can include at least one fluorophore, and therefore can be a fluorescent marker. Each marker can be coupled to the oligonucleotide at any suitable position, including 5' end, 3' end or middle 5' and 3' end.
[0050] "Optical radiation" means electromagnetic radiation in the spectrum, namely ultraviolet light, visible light and / or infrared light. The term "light" without modification has the same meaning as optical radiation.
[0051] "Partially occupied" means present or contained in only a subset of a set of partitions. A set of partitions containing a partially occupied target means that at least one copy of the target is present in each partition of only a subset of the partitions. In other words, one or more partitions do not contain any copies of the target. Copies of the target may be randomly distributed between partitions, but the supply may be limited so that each partition is unable to receive a copy of the target.
[0052] "Partitions" are discrete portions of a fluid (i.e., fluid portions) that are separated from one another (also referred to as separated portions). Each partition in a set of partitions may contain portions of the same sample. Partitions may be separated from one another by fluids (e.g., oil or air), walls of a device, or a combination thereof, etc. Thus, partitions may be droplets of an emulsion, or portions held by a well, chamber (e.g., a nanochamber with a capacity of less than 1 μL), or tube (e.g., a microtube with a diameter of less than 1 mm), etc. Partitions may have the same diameter as one another and / or may contain the same amount of fluid as one another.
[0053] A "partition count" is a value that specifies the number of partitions in a group, such as a subset of partitions with specific target content. The partition count can be an enumerated value.
[0054] The term "positive" when used to refer to the target content of a fragment, partition, or subset of a partition set indicates that each partition of the fragment, partition, or subset contains (or at least appears and / or is considered to contain) at least one copy of a given target or each target of a given target set. The term "negative" when used to refer to the target content of a fragment, partition, or subset of a partition set indicates that each partition of the fragment, partition, or subset does not contain (or at least appears and / or is considered to contain) at least one copy of a given target or each target of a given target set.
[0055] "Prediction algorithm" is a set of instructions followed by a machine to predict features or results."Size prediction algorithm" is a prediction algorithm that predicts at least one length feature. The size prediction algorithm can be set to utilize a prediction function, such as a linear prediction function, to calculate the output value of the length feature representing a nucleic acid fragment colony. In some cases, the size prediction algorithm can be set to receive the input values of the respective variables of the prediction function (the values of its coefficients and intercept have been estimated), and calculate the output value of the prediction function dependent variable. In some cases, the size prediction algorithm can be set to utilize at least two prediction functions, using the same input value, to generate at least two corresponding output values for at least two independent variables representing different length features.
[0056] A "primer" is an oligonucleotide (DNA / RNA or its analogue) that can serve as a starting point for template-directed nucleic acid synthesis under appropriate reaction conditions (e.g., in the presence of a template for oligonucleotide annealing, a nucleoside triphosphate, and a polymerization catalyst (such as a DNA or RNA polymerase or reverse transcriptase), in an appropriate buffer and at an appropriate temperature). A primer can have any suitable length, such as 5 to 500 nucleotides, etc. The primer can be a member of a "primer pair", which includes a "forward primer" and a "reverse primer", which defines the end of the amplicon produced in the amplification reaction. (The adjectives "forward" and "reverse" are arbitrary terms relative to each other). The forward primer hybridizes with a complementary sequence to the 5' end region of the template sequence to be amplified, and the reverse primer hybridizes with the 3' end region of the target sequence. The term "primer binding site" refers to the portion of the template (or its complement) to which the primer anneals. The full-length sequence of the primer does not need to be completely complementary to the primer binding site, as long as there is sufficient complementarity under the reaction conditions to anneal. Thus, a primer may have a 3' terminal region that is complementary to the primer binding site and a 5' terminal region that is not complementary to the primer binding site (and forms a "5'-tail").
[0057] A "probe" is a labeled oligonucleotide that is configured to report the occurrence of an amplification reaction and / or the amplicon formed by the amplification reaction. The probe can be a fluorescent probe comprising an oligonucleotide labeled with a fluorophore. The probe can be configured to hybridize with at least a portion of the amplicon produced by the amplification. The probe (e.g., a hydrolysis probe) can be configured to hybridize with at least a portion of the amplicon during the annealing / extension phase of the amplification cycle, or the probe (e.g., a molecular beacon probe or a strand displacement probe) can be configured to hybridize with the amplicon after the amplification reaction is completed, etc.
[0058] A "reagent" is any substance or reagent used in a reaction for any purpose other than the analyte (eg, a population of nucleic acid fragments and / or a target) being analyzed.
[0059] A "standard" is a sample containing nucleic acid fragments that has been (or is to be) subjected to two different sizing techniques. One sizing technique involves performing a digital assay on at least a target pair provided by the nucleic acid fragments of the standard.
[0060] "Substantially" means conforming primarily to the particular size, range, shape, concept or other aspect modified by the term, such that the feature or component does not need to conform exactly as long as it is suitable for its intended purpose or function. For example, a "substantially cylindrical" object means that the object resembles a cylinder, but may have one or more deviations from a true cylinder.
[0061] A "target" (also referred to as a "target sequence") is a nucleic acid polymer sequence (DNA and / or RNA) of any suitable length to be amplified in an amplification reaction. Exemplary target sequences are about 20-1000 nucleotides, or about 30-500 nucleotides, etc.
[0062] A "target-defined subset" of a set of partitions contains partitions with predefined target content (product positive / negative) for one or more targets.
[0063] A "template" is a nucleic acid polymer that includes a target sequence and / or its complementary sequence. II. Overview
[0064] This section provides an overview of methods, systems, and computer-readable media for determining the size of a population of nucleic acid fragments using digital assays.
[0065] An exemplary method for sizing a population of nucleic acid fragments of a sample is provided. In the method, partitions can be formed, each partition comprising a portion of the sample. Each of at least two targets can be present only in a portion of the nucleic acid fragment. An amplification reaction can be performed on each of the at least two targets in the partition. Amplification data for at least two targets can be collected from the partition. Amplification data can be used to enumerate a subset of the target definitions of the partition. The enumeration results can be used to predict at least one length feature of the nucleic acid fragment population.
[0066] Provided is an exemplary method for obtaining a size prediction algorithm for determining the size of a nucleic acid fragment population in a sample. In the method, a series of standards can be provided. Each standard of the series can have varying degrees of nucleic acid fragmentation and can include at least two targets. A plurality of partition sets can be formed. Each partition set can contain a portion of one of the standards and can contain each of at least two targets that are partially occupied. An amplification reaction of each of the at least two targets can be performed in each partition set. Amplification data of at least two targets can be collected from each partition set. Amplification data can be used to enumerate a subset of the target limits of each partition set. For each standard, the length of the nucleic acid fragment present in the standard can be measured. The size prediction algorithm can be generated by machine learning based on the results of enumeration and measurement.
[0067] A computer readable medium is provided. The computer readable medium contains program instructions for measuring the size of nucleic acid fragments in a sample. The execution of the program instructions by one or more processors of a computer system causes the one or more processors to perform a method. In this method, at least one length feature of a nucleic acid fragment population can be predicted using a value obtained by counting a subset of a partition defined by a target of at least two targets in a digital amplification test or derived therefrom. The partition can contain each of at least two targets in a partially occupied state. Each of at least two targets can be provided by a nucleic acid fragment population.
[0068] A system for measuring the size of a nucleic acid fragment population is provided. The system may include a first primer pair for amplifying a first target and a second primer pair for amplifying a second target. The system may also include a computer-readable medium containing program instructions that can be executed by one or more processors of a computer system to predict at least one length characteristic of a nucleic acid fragment population using a value obtained or derived from a subset of partitions defined by counting targets at least in part in a digital amplification test. Each partition may contain a first primer pair and a second primer pair. The partition may contain each of the first target and the second target that are partially occupied. The first target and the second target may be provided by a nucleic acid fragment population. III. Examples, Components and Alternatives
[0069] Subsections AF below relate to methods, systems and computer-readable media for determining nucleic acid fragment population sizes using digital experiments. The examples of these subsections are intended to illustrate and should not be construed as limiting the full scope of the disclosure. Each subsection may include one or more different examples and / or contexts or related information, functions and / or structures. A. Digital assay method and system for determining nucleic acid population size Snippet
[0070] This section describes exemplary methods and systems for determining the size of a population of nucleic acid fragments using digital assays for targets of varying lengths; see Figure 1-3 .
[0071] Figure 1 A schematic flow chart for performing an experiment on a fluid 22 containing a sample to predict at least one length characteristic of a population 24 of nucleic acid fragments 26 therein is shown. Fig. 20 The nucleic acid fragments 26 provide copies of a first target 28 (interchangeably referred to as target B; black boxes) and copies of a second target 30 (interchangeably referred to as target W; white boxes). Only a portion of the nucleic acid fragments 26 contain the first target 28, and only a portion of the nucleic acid fragments contain the second target 30. In most cases, the majority of the population 24 contains negative target fragments 32, such that the portion of the population 24 containing one or both targets 28, 30 is only a small portion of the total population (e.g., less than 1%, 0.1%, or 0.01%, etc.). However, for simplicity, Figure 1 Only a small number of negative target fragments 32 are depicted.
[0072] The sample-containing fluid 22 can be prepared as a continuous phase contained by a container 34. The continuous phase can include a sample 36 that provides a population 24 and any suitable reagents for performing a digital experiment. These reagents can include amplification reagents 38, such as amplification primers for each target 28, 30, probes and / or labels capable of detecting amplification, nucleotide monomers (e.g., dNTPs) for primer extension, enzymes that catalyze amplification, etc.
[0073] The sample-containing fluid 22 is separated, as shown by separation arrows 40, to form partitions 42. Each partition 42 is a separate sample portion of the sample-containing fluid 22 and includes a portion of the sample 36 and each amplification reagent 38. Therefore, each partition 42 can contain multiple negative target fragments 32 from the sample-containing fluid 22, although not explicitly shown here. The partitions are generally dispersed phases generated at least in part from a continuous phase. In some instances, each partition can be formed at least in part by merging fluid volumes with each other. For simplicity, only a small number of partitions are shown here.
[0074] Only a subset of partitions 42 contain at least one copy of one or both of targets 28, 30. A first subset of partitions is a first target partition 44 containing at least one copy of the first target 28. A second subset of partitions is a second target partition 46 containing at least one copy of the second target 30. A third subset of partitions is a first negative target partition 48 containing no copies of the first target 28. A fourth subset of partitions is a second negative target partition 50 containing no copies of the second target 30. The first, second, third, and fourth subsets may overlap with each other. Figure 1 In the example shown, the first target and the second target are not connected to each other and are distributed independently of each other between the partitions. By this arrangement, the two targets are co-located to the same partition in a concentration-dependent manner by chance. For simplicity, the partitions comprising both targets 28,30 are not depicted, so the first subset and the second subset of the partition do not overlap in this example, but in other examples, these subsets of the first target partition 44 and the second target partition 46 do overlap each other. In some instances, the first target and the second target are closely connected to each other and are effectively co-located to the same partition, without a large amount of fragmentation (e.g., referring to subsection B). As another simplification, it is not shown that the partition comprises two or more copies of one of targets 28,30. In other examples, multiple copies of each target are present in one or more partitions.
[0075] Corresponding amplification reactions for each target 28, 30 are performed in partition 42, as indicated by amplification arrows 52. As indicated by post-amplification partition 54, the first amplification reaction produces a copy of a first amplicon 56 corresponding to the first target 28, and the second amplification reaction produces a copy of a second amplicon 58 corresponding to the second target 30. The first amplicon 56 is produced only in the first target partition 44, and the second amplicon 58 is produced only in the second target partition 46.
[0076] The occurrence of each amplification reaction is detected for the individual partitions 54, as shown by the detection arrows 60, to collect amplification data. In the depicted example, a first fluorescence 62 and a second fluorescence 64 are detected from the amplified partitions 54. The intensity of the first fluorescence 62 detected from each partition 54 indicates whether the first amplification reaction occurred in the partition, and the intensity of the second fluorescence 64 indicates whether the second amplification reaction occurred in the partition.
[0077] The partitions 54 are enumerated with respect to the presence / absence (positive / negative) of each target 28, 30, as indicated by the enumeration arrows 66. For example, the enumeration can include determining a partition count for partitions that are negative for a first target, a partition count for partitions that are positive for the first target, a partition count for partitions that are negative for a second target, a partition count for partitions that are positive for the second target, and / or a total number of partitions.
[0078] In some cases, the values obtained by enumeration can be used to derive other values. For example, the enumerated values can be used to calculate the concentration value of each target (e.g., as copies per unit volume), the score of the positive partition of each target (i.e., the score positive value of the target), the negative partition of each target (i.e., the negative score value of the target), the normalized partition count (e.g., according to the predefined total partition count conversion), the target ratio value (e.g., the ratio of the first target partition count to the second target partition count, or vice versa), and / or the like.
[0079] The values obtained by enumeration or derived at least in part from the enumeration can be used in various ways. In some cases, any one of these values can be the input of a size prediction algorithm, which uses these inputs to predict at least one length characteristic of a sample (e.g., see subsection D). In other cases, the sample can be a standard with a separately measured (and optionally, more directly measured) length characteristic, and any one of these values can be used, as well as corresponding values from other standards with a separately measured length characteristic, to generate such a size prediction algorithm (e.g., see subsection C).
[0080] As an illustrative example, the counting results may be used to determine a target ratio (eg, a concentration ratio, a partition count ratio, etc.) of a first target to a second target in the sample-containing fluid 22. Figure 1In the experiment, the nucleic acid polymer of the sample 36 is not over-fragmented, so that the median / average length of the nucleic acid fragment 26 is much greater than the length of the first and second targets 28, 30. In this specific example, the first target 28 and the second target 30 have the same copy number in the genome or transcriptome represented by the sample 36. For simplicity, and because the sample is not over-fragmented, only full-length, intact copies of the first target 28 and the second target 30 are shown to be present in the fluid 22 containing the sample and are assumed to be amplified and detected with the same efficiency. As a result, the copies of the first target (B) and the second target (W) are determined in the experiment to have a target ratio (R B:W ) is 1 (i.e. 1:1).
[0081] Figure 2 Shown with Figure 1 Another schematic flow chart of an assay very similar to the assay of 120. Figure 1 The same amplification reagent 138 as the amplification reagent 38 represents Figure 1 The same genome or transcriptome containing sample fluid 122 is performed Figure 2 However, the sample-containing fluid 122 has a population 124 of nucleic acid fragments 126 provided by the sample 136, which is much more fragmented than the nucleic acid fragments 26 (compare Figure 1 and Figure 2 In other words, the median / average length of the nucleic acid fragments 126 in sample 136 is much smaller than the length of the nucleic acid fragments 26 in sample 36, and is comparable to the lengths of the first and second targets 128, 130.
[0082] Nucleic acid fragments 126 provide unfragmented and fragmented forms of each target. An intact first target 128 (unfragmented target B; black frame) is identical to first target 28. A fragmented first target 129 (fragmented target B; disrupted black frame) is created by disrupting the intact first target 128. An intact second target 130 (unfragmented target W; white frame) is identical to second target 30. Finally, a fragmented second target 131 (fragmented target W; disrupted white frame) is created by disrupting the intact second target 130.
[0083] Execute the above for Figure 1 process Fig. 20The steps described. The fluid 122 containing the sample is separated, as shown by the separation arrow 140, to form a partition 142. The corresponding amplification reaction of each target is carried out in the partition 142, as shown by the amplification arrow 152. As shown in the amplified partition 154, the intact first and second targets 128, 130 can be amplified and detected using the amplification reagent 138. In contrast, the broken first and second targets 129, 131 are substantially non-amplifiable and / or non-detectable in the test. For example, using primer pairs for the target, exponential amplification of the broken first target or the broken second target will not occur in the partition 142, because the break separates the binding site of the primer pair from the individual nucleic acid fragments 126. In other words, the initial break separates the copies of the intact first target 128 or the intact second target 130 into the first target fragment 133a, 133b pair or the second target fragment 135a, 135b, and each primer has only the binding site of one primer in the corresponding primer pair. Partitions 142 that receive only target fragments 133a, 133b, 135a, or 135b, or only any combination of two or more of these target fragments, serve as negative target partitions 155 for amplification and detection.
[0084] The post-amplification partitions 154 are enumerated relative to the presence / absence (positive / negative) of each intact target 128, 130, as indicated by the enumeration arrows 166, as described above for Figure 1 The process Fig. 20 Due to the increased fragmentation of nucleic acid polymers in sample 136, the number / concentration of each unfragmented target is lower than Figure 1 . However, the second target 130 is significantly longer than the first target 128 (e.g., at least 25%, 50%, 75%, or 100% longer). Therefore, the frequency of fragmentation of the second target is greater than the frequency of fragmentation of the first target. In this simplified example, the amount of fragmentation of sample 136 has increased the measured target ratio (RB:W) of the first target (B) to the second target (W) to 2 (i.e., 2:1).
[0085] Figure 3 An example graph is shown that plots the generalized exponential relationship between the sample fragmentation extent and the ratio of the level of the intact target for targets of different lengths B (black) and W (white). The exponential relationship can be derived as shown below in Equation 2-22.
[0086] Equation 2-8 defines various variables and relationships: B≡Black Target (2) W≡White Target (3) L B ≡The length of B (4) L W ≡The length of W (5) L W >L B(6) λ t ≡Average number of breaks in the target (8) Assume that the average number of breaks in target W is proportional to the average number of breaks in target B, according to the ratio x of the target lengths, as shown in Equation 9: λ W = xλ B (9) Equation 10-18 defines other variables and relationships: N B0 ≡When λ B = The number of unbroken B at 0 (10) N W0 ≡When λ W = The number of unbroken W at 0 (11) When N≡λ B = the number of unbroken B at time i (12) N Wi ≡When λ B = the number of unbroken W at time i (13) K≡x-1 (14) f t ≡Fraction of targets that are broken (16) N Bi =N B0 (1-f B ) (17) N Wi =N W0 (1-f W ) (18) Substituting N in Equations 17 and 18 Bi and N Wi Substituting into Equation 14 yields Equation 19: The fragmentation of each target is expected to have a Poisson distribution as shown in Equation 20: Therefore, the proportion of each target that is fragmented (i.e., has at least one fragmentation) is given by Equation 21: Substituting Equation 21 for f in Equation 18 B and f W , and simplifying it, we get Equation 22:
[0087] Figure 3 The target ratio R plotted based on Equation 22 is shown B:W 168 as a function of the average number of fragments B in target B. In the resulting curve 168, as the fragmentation of nucleic acid polymers in the sample increases (and therefore the average / median fragment length decreases), at least within a certain range of fragmentation, R B:W Thus, the target levels of a pair (or more) of targets vary relative to each other as a function of the mean / median fragment length and provide information about the size of nucleic acid fragments in the sample, as further explained below. B. Illustrative Target / Reagent Setup
[0088] This section describes an exemplary setup of ligated targets and corresponding reagents for performing the digital assay of Section A, as well as an exemplary flow chart for performing a digital assay using a sample-containing fluid containing ligated targets and representing different degrees of nucleic acid fragmentation; see Figure 4-8 .
[0089] Figure 4-6 Schematically illustrates Figure 1 and Figure 2 The assay contains three partitions 254a-254c corresponding to nucleic acid segments 270a-270c. Each nucleic acid segment 270a-270c includes at least one unfragmented target of a pair of linked targets. The linked targets are a first target 228 (unfragmented target B) and a second target 230 (unfragmented target W) of different lengths. Nucleic acid segment 270a includes unfragmented targets B and W (see Figure 4 ). Nucleic acid fragment 270b includes target W, but only includes the broken portion 233b of target B (see Figure 5 ). Nucleic acid fragment 270c includes target B, but only includes the broken portion 235a of target W. Targets B and W can have any length described elsewhere herein (such as in subsection A).
[0090] Figure 4 Partition 254a contains nucleic acid fragment 270a, wherein unbroken target B and W are connected to each other by spacer 272 (interchangeably referred to as spacer sequence). Spacer 272 can have any suitable length, such as at least 1,2,5,10,20,50 or 100 nucleotides, and / or less than 5000,2000,1000,500,200,100,50,20 or 10 nucleotides etc.Alternatively or additionally, the length of the spacer can be greater than or less than the length of one or two targets.Longer spacers may be conducive to more accurately measuring the size of larger fragments, and shorter spacers may be conducive to more accurately measuring the size of smaller fragments.In some instances, target B and W can be directly attached to each other without any intermediate spacer.
[0091] Each partition 254a-254c contains a primer pair for amplifying each target (see Figure 4-6 ). Primers 274a, 274b are forward and reverse primers, respectively, to initiate amplification of target B to form a first amplicon 256. Primers 276a, 276b are forward and reverse primers, respectively, to initiate amplification of target W to form a second amplicon 258. The occurrence of each amplification reaction using a pair of primers is indicated by an amplification arrow 252.
[0092] Each partition 254a-254c also contains a pair of optical probes to be able to detect the occurrence of each amplification reaction. The first probe 278 is configured to hybridize with target B and / or the first amplicon 256, and the second probe 280 is configured to hybridize with target W and / or the second amplicon 258. Each probe includes a fluorophore 282a or 282b and a quencher 284a or 284b of the fluorophore. The generation of the first amplicon 256 separates the fluorophore 282a from the quencher 284a to increase the first fluorescence 262 of the fluorophore 282a. The generation of the second amplicon 258 separates the fluorophore 282b from the quencher 284b to increase the second fluorescence 264 of the fluorophore 282b. In the depicted example, the fluorophore and the quencher are separated from each other by the cracking of the first probe 278 or the second probe 280 to generate a degraded probe 278d or 280d. In other examples, when the probe hybridizes with the corresponding amplicon (e.g., if the probe is a molecular beacon probe or a strand displacement probe), separation can occur without degradation. In some examples, the fluorophore 282a and the fluorophore 282b can be identical to each other and can be distinguished by their fluorescence intensity. In other examples, the probe can be replaced by an intercalating dye.
[0093] Subareas 254a-254c have distinguishable fluorescence patterns. Subarea 254a contains targets B and W and therefore exhibits fluorescence from fluorophores 282a and 282b (see Figure 4 ). Partition 254b contains target W but only the broken portion of target B and therefore exhibits the fluorescence of fluorophore 282b but not the fluorescence of fluorophore 282a (see Figure 5 ). Partition 254c contains target B but only the fragmented portion of target W and therefore exhibits the fluorescence of fluorophore 282a but not the fluorescence of fluorophore 282b (see Figure 6 ).
[0094] Figure 7 A schematic flow chart 220 (with reference to FIG. 220 ) is shown for conducting an experiment on a fluid 222 containing a sample to predict at least one length characteristic of a population 224 of nucleic acid fragments 226 therein. Figure 1Nucleic acid fragments 226 include a population of target-containing fragments 226a that provide copies of a first target 228 (B) and a second target 230 (W) (see also Figure 4 ). Only a subset of nucleic acid fragments 226 contain targets B and W. In most cases, the majority of population 224 contains negative target fragments 232, but for simplicity, Figure 7 Only a small number of negative target fragments 232 are depicted in the figure. The average / median length of the nucleic acid fragments 226 is much greater than the length of the target region 286 of the fragments 226a containing the target, wherein the target region 286 is defined by the targets B and W and the spacer 272 (if present). Therefore, none of the copies of the target region 286 are broken. After partitioning 240, amplification 252 and detection 260, enumeration 266 obtains partition counts of 4, 0 and 0 for the B+W positive partition, the B positive partition only and the W positive partition only. The low percentages of the B positive partition only and the W positive partition only indicate that the median / average length of the nucleic acid fragments 226 in the population 224 is larger relative to the length of the target region 286.
[0095] Figure 8 A schematic flow chart 320 (with reference to FIG. 1 ) is shown for conducting an assay on a fluid 322 containing a sample to predict at least one length characteristic of a population 324 of nucleic acid fragments 326 therein. Figure 1 , 2 Compared with 7). The mean / median length ratio of population 324 Figure 7 As a result, the target region 386 (B+W) provided by the nucleic acid fragment 326 is often destroyed. (The target region 386 is identical in sequence to Figure 4 and Figure 7 326a). In this highly simplified example, the targets B and W of the target-only fragment 326a are not broken. The target-containing fragment 326b has the target W not broken, but only has the broken portion of the target B. The target-containing fragment 326c has the target B not broken, but only has the broken portion of the target W. The fragment 326d containing part of the target has only a portion of each target B and W. After partitioning 340, amplification 352 and detection 360, enumeration 366 obtains partition counts of 1, 2 and 1 for the B+W positive partition, the B-only positive partition and the W-only positive partition, respectively. The higher percentages of the B-only positive partition and the W-only positive partition are due to the fact that the degree of fragmentation of the population 324 is much higher than Figure 7 224 in the population, so the median / average length of the nucleic acid fragments 326 in the population 324 is comparable to or shorter than the length of the target region 386.
[0096] The partition counts of the three target-defined partition subsets (B+W, B only, and W only) can directly or indirectly provide inputs for the respective variables of the size prediction algorithm. The size prediction algorithm linearly combines these partition counts or values derived therefrom to produce an output representing at least one length characteristic of the population of nucleic acid fragments in the sample. Therefore, in some examples, exactly three values (three partition counts or three values derived therefrom, respectively) can be used as input, which is sufficient for the size prediction algorithm to produce an output. C. Exemplary amplification data used to generate the size prediction algorithm
[0097] This section describes exemplary amplification data collected from a series of standards in the digital experiments of Sections A and B using the target and reagent settings of Section B; see Figure 9-17 The amplification data demonstrates variation in the relative abundance of ligated (B+W) and unligated targets (B only or W only) as a function of median fragment length and can be used to generate size prediction algorithms, such as by machine learning.
[0098] Figure 9-16 is used Figure 4-6 Target and reagent settings for scatter plot of amplification data collected from digital experiments performed in droplets (as partitions). Target B is 79 bp, target W is 200 bp, and the spacer separating the B and W targets is 9 bp. The PCR products were purchased from MilliporeSigma, Burlington, MA. Each standard in a series of standards generated by endonuclease fragmentation of an aliquot of a genomic HEK cell DNA sample was subjected to a digital amplification assay. The median length of the nucleic acid fragments in each standard is indicated for each figure and was independently measured by electrophoresis using a bioanalyzer commercially available from Agilent, Santa Clara, CA.
[0099] The scatter plots represent each droplet as a point located according to the fluorescence amplitude (expressed in relative fluorescence units (RFU)) measured from the droplet at each of two different wavelengths ("Channel 1" and "Channel 2", respectively). Channel 1 fluorescence indicates whether the droplet contains at least one copy of target B, and channel 2 fluorescence indicates whether the droplet contains at least one copy of target W. In each plot, four clusters of points / droplets are circled and identified according to B / W content: - / - (B and W are negative), + / - (only B is positive), - / + (only W is positive), and + / + (both B and W are positive).
[0100] Fig.17 A bar graph is shown, which plots the median fragment lengths of each standard from Figure 9-16The estimated target concentrations (B fragment only, W fragment only, and B+W fragment) are shown in the scatter plot. With the increase of fragmentation, the concentration of B+W gradually decreases, the concentration of W only increases and then decreases, and the concentration of B only increases and then decreases more slowly than the concentration of W only.
[0101] Include Figure 9-17 The data set of data points produced by the assay and the individual measured length characteristics of each standard can be used to estimate the values of the coefficients and constant terms of the input expressions of the prediction functions used by the size prediction algorithm. Each data point represents a standard. The digital amplification assay and the individual size measurements performed on the standards provide the input and output values of the data points, respectively. Figure 9-16 Any suitable count value obtained in the amplification test or the value derived therefrom can be used as the input value of the data point of the data set. The coefficient of the prediction function and the value of the constant term can be estimated by linear regression, logistic regression or any other suitable statistical method. The median length (and / or other length characteristics) of the independent measurement of each standard is the output value of the prediction function of each group of input values from one of the standards. For example, the input value can be the original or normalized partition count of the partition subset defined by the target, the estimated concentration of each unconnected target and the target connected, the target ratio, the positive score (or negative score) of a given target (unconnected) or a given combination of two or more connected targets, etc. D. Evaluation of size prediction algorithms generated by logistic regression
[0102] This subsection describes the evaluation of a logistic regression model using a series of fragmentation standards, each with an independently measured median nucleic acid fragment length; see Fig.18 .
[0103] Sigma Millipore Corp., Burlington, Massachusetts Endonuclease treatment of genomic DNA, a series of standards are generated. The median length of the nucleic acid fragments present in each standard is measured by electrophoresis using a bioanalyzer commercially available from Agilent, Santa Clara, California, USA. Each standard has a different median fragment length. It is also tested in the digital amplification test using droplets (as partitions) whether each standard has unconnected target B, unconnected target W and connected target B+W, as described in subsections B and C above. The raw counts of the droplet subsets that are only positive for target B, only for target W and target B+W of each standard are used as the input of the size prediction algorithm, which is generated in a logistic regression model by machine learning, as generally described in subsection C. The size prediction algorithm provides output (i.e., model value) for each group of inputs (raw counts) of each standard. Each standard is tested repeatedly in the digital amplification test to obtain a pair of model (output) values of the standard.
[0104] Fig.18A pair of model (output) values for each standard is plotted against the median fragment length of the independently measured base pairs of the standard. The size prediction algorithm was generated using machine learning based on a threshold median length of 200bp. In other words, the coefficients and intercept of the size prediction algorithm were estimated so that output model values below zero predict a median fragment length greater than or equal to 200bp, and model values above zero predict a median fragment length less than 200bp. In general, the more negative the model value for a standard, the more likely it is that the median fragment length of the standard is greater than or equal to the threshold length of 200bp. Fig.18 In the graph of , circles representing model values above zero are unfilled, and circles representing model values below zero are filled. Therefore, the logistic regression model accurately predicts that the median fragment length of the two most fragmented standards (i.e., 93bp and 161bp median fragment lengths) is less than 200bp. The logistic regression model also accurately predicts that the median fragment length of the five least fragmented standards (i.e., median fragment lengths of 490bp, 930bp, 2260bp, 4094bp, and 4573bp, respectively) is greater than or equal to 200bp, with increasing certainty as the median fragment length increases. In some cases, extensively degraded samples may be more difficult to accurately size because degradation reduces the total number of detectable targets, thereby increasing statistical variation. In this case, shorter targets may be more advantageous. E. Methods for determining size
[0105] This section describes exemplary methods for sizing a population of nucleic acid fragments provided by a sample; see Fig.19 400. The methods may include any suitable combination of the steps listed in flowchart 400, performed in any suitable order using any suitable aspects and features of the present disclosure (eg, see subsections AD).
[0106] Partitions can be set in step 402. Each partition can contain a portion of a sample, which includes a nucleic acid fragment population. The nucleic acid fragment provides at least two targets of different lengths from each other, such as a first target and a second target. The nucleic acid fragment can be a nucleic acid polymer, such as DNA (e.g., genomic DNA), reverse transcribed RNA (i.e., DNA produced therefrom), RNA, etc. The partition set can be formed by dividing a mixture of a sample and a reagent that is sufficient to amplify each target in at least two targets. Alternatively, the sample can be separated from at least one amplification reagent and then combined with at least one amplification reagent. In some cases, the nucleic acid fragment can provide at least one other target other than the first target and the second target, such as a third target, a fourth target, etc. Therefore, the reagent used for amplification can also be sufficient to amplify at least one other target.
[0107] Each of the at least two targets may be present in the set of partitions in a partially occupied manner. In other words, only a subset of the partitions contains each target.
[0108] A pair (or more) of targets in a sample may be at least partially connected to each other. This association results in the target pair being co-localized to the same partition at a statistically significant frequency (i.e., co-localization greater than chance). A disruption event that interrupts a copy of the first target of the pair, a copy of the second target of the pair, or a copy of the spacer between the pair of targets reduces the ratio of the first and second targets co-localized to each other in the partition (see subsections B and C).
[0109] In step 404, an amplified reaction for at least two targets can be carried out in the group of partitions. The amplified reaction produces the corresponding amplicon corresponding to each target. The first amplicon can only be produced in the partition containing at least one copy of the first target, and the second amplicon can only be produced in the partition containing at least one copy of the second target, etc. The amplified reaction can be driven by heating and / or this partition set of thermal cycles. Thermal cycles can drive polymerase chain reaction or ligase chain reaction, etc.
[0110] In step 406, amplification data of at least two targets can be collected from the subregion. Amplification data is produced by the amplification reaction carried out in step 404, and can be detected continuously or in parallel from the subregion by any suitable technology. In some instances, amplification data can be optically detected by fluorescence. The generation of each amplicon can be optically distinguished by wavelength or intensity from the generation of other amplicon in each subregion. In some cases, the generation of the first amplicon can be detected by the first marker in the subregion set, and the generation of the second amplicon can be detected by the second marker. In some cases, the generation of each amplicon in the first and second amplicon can be detected by the same marker (e.g., intensity difference).
[0111] In step 408, the subsets of the target definitions of the partition set can be enumerated using the amplification data to obtain partition counts. The enumeration can include determining the first partition count of the first subset of partitions that are positive for the first target, the second partition count of the second subset of partitions that are positive for the second target, the third partition count of the third subset of partitions that are positive for both the first target and the second target, the fourth partition count of the fourth subset of partitions that are negative for both the first target and the second target, and / or the total target independent count of the entire partition set. The determination of the total number of partitions allows the calculation of the score of the positive or negative partitions of each target or target combination, and the concentration of the target or target combination is estimated from the score. The first subset of partitions can include all partitions that are positive for the first target or can be limited to partitions that are positive for the first target and negative for the second target. The second subset of partitions can include all partitions that are positive for the second target or can be limited to partitions that are positive for the second target and negative for the first target. Because positive counts and negative counts are opposite to one another, enumeration can include determining a partition count that is negative for the first target, a partition count that is negative for the second target, and in some cases, a partition count that is negative for either the first target or the second target.
[0112] At least one length feature of a nucleic acid fragment colony can be predicted based on the enumeration result (i.e., the result from step 408) in step 410. Each length feature can be directly predicted using the partition counts of each target and target combination, or can be predicted using the value derived from the partition count. Possibly suitable exemplary derived values include the positive score / negative score value of each target and / or target combination. Other exemplary derived values include normalized values, such as the partition count converted by a conversion factor. Possibly suitable other exemplary derived values include the concentration-related values (e.g., the average copy number or molar number of each partition or unit volume) of each target and / or target combination. Positive scores or negative score values of each target and / or target combination can be used to calculate concentration values. Possibly suitable another exemplary derived value includes one or more positive target ratios and / or negative target ratios of targets / target combinations.
[0113] Each length feature can be predicted using a size prediction algorithm. The size prediction algorithm can be generated by machine learning as described elsewhere herein. The input value of the size prediction algorithm can be, for example, any of the above-mentioned partition counts, normalized values, positive / negative score values, concentration values, or target ratio values. F. Methods for generating size prediction algorithms
[0114] This section describes an exemplary method for generating a size prediction algorithm; see Fig. 20Methods may include any suitable combination of the steps listed in flowchart 500, performed in any suitable order using any suitable aspects and features of the present disclosure.
[0115] For each standard in a series of standards, at least one length characteristic of the nucleic acid fragment population can be obtained in step 502. At least one length characteristic can be obtained by any suitable measurement technique, such as electrophoresis, mass spectrometry, flow cytometry, electron microscopy, etc. Each standard can represent different degrees of nucleic acid fragmentation. Each standard can be produced by physical or chemical fracture of nucleic acid polymers. In some instances, more than one length characteristic can be obtained for each standard. For example, the median length of each standard and the percentage of fragments above / below a threshold size can be obtained.
[0116] A digital test can also be used to test a series of standards of step 502. A corresponding set of partitions can be formed for each standard, each partition of the set including a portion of the standard. At least two targets (i.e., the same at least two targets) can be amplified in each set of partitions, and amplification data of the at least two targets can be collected from each set of partitions.
[0117] The partition subsets of the target definitions for each set of partitions may be enumerated at step 504. The enumeration may be performed as described above in subsection E of step 408, and any suitable partition count values and / or values derived therefrom may be obtained.
[0118] At least one data set may be used in step 506 to generate a size prediction algorithm, the at least one data set comprising data points generated by steps 502 and 504 and corresponding to a series of standards. Each data point is defined by an input value of one or more independent variables and an output value of a dependent variable. The input value is obtained or derived by enumeration for one of the standards (i.e., from step 504). The output value is obtained or derived by measuring at least one length characteristic of the same standard (i.e., from step 502). Generating a size prediction algorithm may be performed by linear regression or logistic regression (e.g., binary regression), etc. Generating a size prediction algorithm may include estimating the coefficient values of the variable terms and constant terms of the regression model by machine learning. The value of the coefficient may include each weighted value and / or the intercept value of the y intercept in one, two, three or more independent variables (model inputs). Any value or derived value described in conjunction with step 408 in subsection E above may be applicable to generating a size prediction algorithm.
[0119] In some cases, at least two data sets can be used to generate a size prediction algorithm, or at least two size prediction algorithms can be used to predict at least two different length characteristics of a sample. The at least two data sets can be derived from the same series of standards or different series of standards. IV. Illustrative combinations and additional examples
[0120] This section describes additional aspects and features of the methods, systems, and computer-readable media of the present disclosure, which are presented in the form of a series of paragraphs, but are not limited to this, some or all of which may be indexed alphanumericy to improve clarity and efficiency. Each of these paragraphs can be combined with one or more other paragraphs in any suitable manner, and / or with the disclosure elsewhere in this application. Some of the paragraphs below explicitly mention and further limit other paragraphs, and provide some examples of suitable combinations, but are not limited to this.
[0121] A1. A method for determining the size of a population of nucleic acid fragments provided by a sample, the method comprising: forming partitions each containing a portion of the sample, wherein each of at least two targets is present only in a portion of the nucleic acid fragment, and wherein, optionally, the at least two targets have different lengths relative to each other; performing an amplification reaction on each of the at least two targets in the partition; collecting amplification data for the at least two targets from the partition; using the amplification data to enumerate a target-defined subset of the partition; and using the enumeration results to predict at least one length feature of the population of nucleic acid fragments.
[0122] A2. The method according to paragraph A1, wherein the prediction includes predicting whether the nucleic acid fragment population meets one or more predefined size criteria.
[0123] A3. A method according to paragraph A2, wherein the prediction includes classifying the length of the population relative to a threshold size.
[0124] A4. A method according to paragraph A3, wherein the prediction includes predicting whether the median length or average length of the nucleic acid fragment population is above or below a threshold size.
[0125] A5. A method according to paragraph A4, wherein the prediction includes calculating the logarithmic probability that the median length or average length of the nucleic acid fragment population is above or below a threshold size.
[0126] A6. A method according to any of paragraphs A1 to A5, wherein the prediction includes predicting a percentage value of the nucleic acid fragment population that meets one or more predefined size criteria.
[0127] A6.1. A method according to paragraph A6, wherein the percentage value represents the percentage of the nucleic acid fragment population that has a predefined relationship with the threshold length.
[0128] A6.2. A method according to paragraph A6.1, wherein the threshold length is 200 base pairs or nucleotides.
[0129] A6.3. A method according to claim A6.1 or A6.2, wherein the at least two targets include a first target and a second target, and wherein the first target is longer than the second target and has a length within 10%, 5% or 2% of (or equal to) the threshold length.
[0130] A6.4. A method according to any of paragraphs A6 to A6.3, wherein the prediction includes predicting a confidence interval for the percentage value.
[0131] A7. A method according to any of paragraphs A1 to A6, wherein the prediction includes predicting the mean, median, quartile and / or quintile length of the nucleic acid fragment population.
[0132] A7.1. A method according to any of paragraphs A1 to A7, wherein the prediction includes predicting the median length of the nucleic acid fragments in the population.
[0133] A7.2. A method according to paragraph A7.1, wherein the prediction includes predicting a confidence interval for the median length.
[0134] A7.3. A method according to paragraph 7.2, wherein the confidence interval is a 95% confidence interval of the median length.
[0135] A8. A method according to any of paragraphs A1 to A7, wherein enumeration includes obtaining two or more enumeration values, and wherein prediction includes combining the two or more enumeration values or values derived therefrom as input values of independent variables of a predictor function to generate a length feature representing the at least one length feature.
[0136] A9. A method according to paragraph A8, wherein the combination includes linearly combining the enumerated values or each value derived therefrom with each other according to the prediction function, and wherein the linear combination includes multiplying each enumerated value or each value derived therefrom by a corresponding coefficient and summing the products.
[0137] A10. A method according to paragraph A9, wherein the corresponding coefficients are not equal to each other.
[0138] A11 A method according to paragraph A9 or A10, wherein none of the coefficients is equal to 1.
[0139] A12. A method according to any of paragraphs A9 to A11, wherein summing the products of the multiplications includes summing the products with a constant, and wherein the constant is the intercept of the prediction function.
[0140] A13. A method according to any of paragraphs A8 to A12, wherein the prediction comprises combining at least three enumerated values or values derived therefrom with one another according to the prediction function.
[0141] A14. A method according to paragraph A13, wherein the prediction comprises a linear combination of exactly three enumerated values according to a prediction function, or exactly three corresponding values derived therefrom.
[0142] A15. A method according to paragraph A14, wherein at least three enumeration values are obtained directly or indirectly from a subset of the partitioned target definition, and the subset of the partitioned target definition is defined only for the positive / negative of the first target and the second target among at least two targets.
[0143] A16. A method according to any one of paragraphs A1 to A15, wherein the prediction is performed by a size prediction algorithm utilizing a linear prediction function.
[0144] A17. A method according to any of paragraphs A1 to A16, wherein the enumeration includes obtaining a first partition count from a first subset of the partitions or from the inverse of the first subset of the partitions, and obtaining a second partition count from a second subset of the partitions or from the inverse of the second subset of the partitions, wherein the first subset of the partitions is positive for a first target among at least two targets, wherein the second subset of the partitions is positive for a second target among at least two targets, and wherein the prediction uses the first partition count and the second partition count or one or more values derived therefrom as input values of independent variables of a prediction function to produce an output value of a length feature representing at least one length feature.
[0145] A18. The method of paragraph A17, wherein a first subset of the partitions is negative for the second target, and wherein a second subset of the partitions is negative for the first target.
[0146] A19. A method according to paragraph A17 or A18, wherein the prediction uses the first partition count and the second partition count as direct inputs to the prediction function.
[0147] A20. A method according to any of paragraphs A17 to A19, wherein the prediction uses values derived from the first partition count and the second partition count, respectively, as inputs to the prediction function.
[0148] A21. A method according to paragraph A20, wherein the prediction uses one or more target concentration values, positive / negative score values, normalized values and / or target ratio values derived from the first partition counts and the second partition counts as inputs to the prediction function.
[0149] A22. The method of paragraph A21, wherein the prediction uses a positive score value or a negative score value, each of which is calculated using one of the first and second partition counts and the total non-target-dependent count of the partition.
[0150] A23. A method according to paragraph A21, wherein the prediction uses a target ratio value calculated using the first partition count and the second partition count.
[0151] A24. A method according to any one of paragraphs A17 to A23, wherein the first target and the second target have a certain degree of association with each other in the sample, wherein the first subset of the partitions is negative for the second target, wherein the second subset of the partitions is negative for the first target, wherein the enumeration includes obtaining a third partition count that is positive for the first target and the second target from a third subset of the partitions, or obtaining a third partition count from the inverse of the third subset of the partitions, and wherein the prediction uses the first partition count, the second partition count and the third partition count or values derived therefrom as inputs of the independent variables of the prediction function.
[0152] A25. A method according to paragraph A24, wherein the prediction uses three inputs to the prediction function.
[0153] A26. A method according to paragraph A24 or A25, wherein the prediction uses the first, second and third partition counts as direct inputs to the prediction function.
[0154] A27. A method according to paragraph A24 or A25, wherein the prediction uses one or more values derived from the first, second and third partition counts as inputs to the prediction function.
[0155] A28. A method according to any of paragraphs A24 to A27, wherein the sample represents a genome wherein the first target and the second target are separated from each other by less than 200 base pairs.
[0156] A29. A method according to any one of paragraphs A24 to A28, wherein the degree of association between the first target and the second target in the sample is related to the median length of the nucleic acid fragment population.
[0157] A30. A method according to any of paragraphs A1 to A29, wherein the prediction utilizes a prediction function having coefficient values estimated from a data set comprising a plurality of data points, wherein each data point represents one standard in a series of standards having different amounts of nucleic acid fragments, wherein the data points include input values of two or more independent variables of the prediction function and output values of a dependent variable of the prediction function, wherein the values of the two or more independent variables of the prediction function are derived from amplification data of at least two targets collected from a set of partitions in a digital experiment, each partition comprising a portion of a standard, and wherein the output value of the prediction function and the dependent variable of the prediction function are obtained by measuring the length of the standard nucleic acid fragment separately from the digital experiment.
[0158] A31. A method according to paragraph A30, wherein logistic regression is used to estimate the coefficient values.
[0159] A32. A method according to paragraph A30, wherein linear regression is used to estimate the coefficient values.
[0160] A33. A method according to any of paragraphs A1 to A32, wherein the at least two targets include a first target and a second target, and wherein the second target is at least 25%, 50%, 75% or 100% longer than the first target.
[0161] A34. A method according to any one of paragraphs A1 to A33, wherein the at least two targets include a first target and a second target, wherein the forming includes forming the partitions to each contain a first probe for detecting amplification of the first target and a second probe for detecting amplification of the second target, and wherein the first probe includes a first marker and the second probe includes a second marker.
[0162] A35. A method according to paragraph A34, wherein the first marker and the second marker comprise fluorophores that are different from each other.
[0163] A36. A method according to paragraph A35, wherein collecting amplification data includes detecting fluorescence from the first marker and the second marker, wherein enumeration of the first target is based on fluorescence from the first marker, and wherein enumeration of the second target is based on fluorescence from the second marker.
[0164] A37. A method according to any of paragraphs A1 to A36, wherein the at least two targets include a first target and a second target, and wherein the forming includes forming the partitions to each contain a first primer pair for amplifying the first target and a second primer pair for amplifying the second target.
[0165] A38. A method according to any of paragraphs A1 to A37, wherein the at least two targets include a first target and a second target, and wherein the second target is at least 25, 50, 75 or 100 nucleotides longer than the first target.
[0166] A39. A method according to any of paragraphs A1 to A38, wherein the at least two targets include a first target and a second target, wherein the sample represents a genome, and wherein the first target and the second target are separated from each other in the genome by less than 10, 5, 2, 1, 0.5, 0.2, 0.1, 0.05 or 0.02 kilobases.
[0167] A40. A method according to any of paragraphs A1 to A39, wherein the nucleic acid fragment comprises a fragment of genomic DNA, RNA or reverse transcribed RNA.
[0168] A41. A method according to any one of paragraphs A1 to A40, wherein the at least two targets include a first target and a second target, and wherein the first target is longer than the second target and has a length within 10%, 5% or 2% of 200 nucleotides.
[0169] B1. A method for obtaining a size prediction algorithm for determining the size of a population of nucleic acid fragments in a sample, the method comprising: providing a series of standards, each standard in the series having different degrees of nucleic acid fragmentation and including at least two targets; forming a partition set, each partition set containing a portion of one of the standards and containing each of at least two targets partially occupied; performing an amplification reaction on each of the at least two targets in each partition set; collecting amplification data of the at least two targets from each partition set; using the amplification data to enumerate a subset of target definitions for each partition set; for each standard, measuring the length of the nucleic acid fragments present in the standard; and generating a size prediction algorithm based on the results of the enumeration and measurement.
[0170] B2. A method according to paragraph B1, wherein generating includes estimating values of coefficients and an intercept of a prediction function used by the size prediction algorithm.
[0171] B3. A method according to paragraph B2, wherein generating uses values obtained or derived by enumeration as input values of independent variables of the prediction function, and wherein measuring uses values obtained or derived by measurement as output values of the independent variables of the prediction function.
[0172] B4. A method according to paragraph B2 or B3, wherein generating includes estimating values of coefficients and intercept by logistic regression.
[0173] B5. A method according to any of paragraphs B1 to B3, wherein generating includes estimating values of coefficients and intercept by linear regression.
[0174] B6. A method according to any of paragraphs B1 to B5, wherein measuring includes electrophoresis of the nucleic acid fragments of each standard.
[0175] B7. The method according to any of paragraphs B1 to B6, further comprising fragmenting the nucleic acid polymer to generate each standard.
[0176] B8. A method according to paragraph B7, wherein the fragmentation comprises chemically fragmenting the nucleic acid polymer.
[0177] B9. A method according to paragraph B8, wherein the chemical fragmentation comprises performing an enzyme-catalyzed chemical reaction to break the nucleic acid polymer.
[0178] B10. The method according to paragraph B7, wherein the fragmentation comprises physically breaking the nucleic acid polymer.
[0179] C1. A computer-readable medium containing program instructions for determining the size of nucleic acid fragments in a sample, wherein execution of the program instructions by one or more processors of a computer system causes the one or more processors to perform a method comprising: predicting at least one length characteristic of a population of nucleic acid fragments using values obtained or derived by enumerating a subset of target definitions of partitions of at least two targets in a digital amplification experiment, wherein the partitions contain partially occupied portions of each of the at least two targets, each of the at least two targets being provided by a population of nucleic acid fragments.
[0180] C2. A computer-readable medium according to paragraph C1, wherein the computer-readable medium is non-transitory.
[0181] C3. A computer-readable medium according to paragraph C1, wherein the computer-readable medium is transient.
[0182] C4. A computer-readable medium according to any of paragraphs C1 to C3, wherein the prediction uses a first partition count and a second partition count, or one or more values derived therefrom, as input to a size prediction algorithm, wherein the first partition count represents a first subset of the partitions, or represents the inverse of the first subset of the partitions, and the second partition count represents a second subset of the partitions, or represents the inverse of the second subset of the partitions, wherein the first subset of the partitions is positive for a first target of at least two targets, and wherein the second subset of the partitions is positive for a second target of at least two targets.
[0183] C5. The method of paragraph C4, wherein a first subset of the partitions is negative for the second target, and wherein a second subset of the partitions is negative for the first target.
[0184] C6. A method according to paragraph C4 or C5, wherein the prediction uses the first partition count and the second partition count as direct inputs to the size prediction algorithm.
[0185] C7. A method as described in any of paragraphs C4 or C5, wherein the prediction uses one or more values derived from the first partition count and the second partition count as input to the size prediction algorithm.
[0186] C8. A method according to paragraph C7, wherein the prediction uses one or more target concentration values, normalized values and / or positive / negative score values derived from the first and second partition counts as input to the size prediction algorithm.
[0187] C9. A method according to paragraph C8, wherein the prediction uses positive score values, each of which is calculated using one of the first and second partition counts and the total non-target-dependent counts of the partition as input to the size prediction algorithm.
[0188] C10. A method according to any one of paragraphs C4 to 9, wherein the first target and the second target have a certain degree of correlation with each other in the sample, wherein the first subset of the partitions is negative for the second target, wherein the second subset of the partitions is negative for the first target, wherein the third partition count represents a third subset of the partitions that are positive for the first target and the second target, or represents the inverse of the third subset, and wherein the prediction uses the first partition count, the second partition count and the third partition count or values derived therefrom as input to the size prediction algorithm.
[0189] C11. A method according to any of paragraphs C1 to C10, wherein the prediction uses three inputs, optionally exactly three inputs for the size prediction algorithm.
[0190] C12. A method according to paragraph C10 or C11, wherein the prediction uses the first, second and third partition counts as direct input to the size prediction algorithm.
[0191] C13. A computer-readable medium as described in any of paragraphs C10 or C11, wherein the prediction uses one or more values derived from the first partition count, the second partition count, and the third partition count as input to a size prediction algorithm.
[0192] C14. A method according to any of paragraphs C1 to C13, wherein the prediction includes predicting the mean, median, quartile and / or quintile length of the nucleic acid fragment population.
[0193] C15. A method according to any of paragraphs C1 to C14, wherein the prediction includes predicting whether the median length or average length of the nucleic acid fragment population is above or below a threshold size.
[0194] C16. A method according to paragraph C15, wherein the prediction includes calculating the logarithmic probability that the median length or average length of the nucleic acid fragment population is above or below a threshold size.
[0195] C17. A computer-readable medium according to any of paragraphs C1 or C16, wherein the prediction is made using a size prediction algorithm generated by logistic regression.
[0196] C18. A computer-readable medium according to any of paragraphs C1 to C16, wherein the prediction is made using a size prediction algorithm generated by linear regression.
[0197] C19. A computer-readable medium according to any of paragraphs C1 to C18, wherein the prediction is performed using a size prediction algorithm, which uses quantitative data obtained from each of the at least two targets obtained from each standard in a series of standards representing different amounts of nucleic acid fragments.
[0198] C20. A computer-readable medium according to paragraph C19, wherein the size prediction algorithm is generated using a data set comprising the quantitative data and a length characteristic of each standard in the series of standards.
[0199] D1. A system for determining the size of a population of nucleic acid fragments, the system comprising: a first primer pair for amplifying a first target; a second primer pair for amplifying a second target; and a computer-readable medium containing program instructions executable by one or more processors of a computer system, for using values obtained or derived from a subset of target definitions of enumerated partitions in a digital amplification experiment to predict at least one length characteristic of a population of nucleic acid fragments, each partition containing a first primer pair and a second primer pair, the partitions containing each of partially occupied first and second targets, the first and second targets being provided by a population of nucleic acid fragments.
[0200] D2. The system of paragraph D1, further comprising a first probe for detecting amplification of the first target using the first primer pair and a second probe for detecting amplification of the second target using the second primer pair. V. Advantages and Benefits
[0201] Different examples of size determination methods and systems and related computer readable media provide several advantages over known solutions for determining the size of nucleic acid fragment populations. For example, the exemplary embodiments described herein can predict the length characteristics of nucleic acid fragments in a sample with higher sensitivity, less sample usage, and / or better suited for automation.
[0202] Furthermore, among other benefits, the examples described herein utilize machine learning to generate size prediction algorithms that are more accurate than other prediction strategies.
[0203] Furthermore, among other benefits, the illustrative examples described herein may also predict a length characteristic of a sample based on one or more input values generated by digitally detecting the sample without the need to electrophoresing the sample.
[0204] Furthermore, among other benefits, the illustrative examples described herein generate or use size prediction algorithms that can predict length features from two or more inputs, and in some cases, more inputs than tags, to detect target amplification.
[0205] Furthermore, the illustrative examples described herein can provide high sensitivity by predicting the length characteristics of a sample based on digital amplification of two or more targets, among other benefits.
[0206] Furthermore, the illustrative examples described herein allow, among other benefits, prediction of length characteristics of a sample nucleic acid fragment population, as well as sample target quantification based on digital amplification of two or more targets.
[0207] Furthermore, among other benefits, the illustrative examples described herein enable detection of two linked targets of different lengths, and co-localization of linked targets in the same partition provides additional input to size prediction algorithms.
[0208] In addition, among other benefits, the illustrative examples described herein also provide a test for the degree of nucleic acid degradation (DNA and / or RNA) of a sample. Thus, the test can allow a user to identify overly degraded samples and exclude these samples from other tests.
[0209] There is no known system or method that can perform these functions.However, not all examples described herein provide the same advantages or the same degree of advantages. VI. in conclusion
[0210] The above disclosure may include multiple different examples with independent utility. Although various of these disclosures have been disclosed in their preferred forms, the specific examples of the invention disclosed and illustrated herein should not be considered restrictive, because multiple changes can be made. With respect to the section titles used in this disclosure, such titles are only used for organizational purposes. The subject matter of the present disclosure includes all novel and non-obvious combinations and sub-combinations of the various elements, features, functions and / or characteristics disclosed herein. The claims specifically point out certain combinations and sub-combinations that are novel and non-obvious. Other combinations and sub-combinations of features, functions, elements and / or characteristics may be claimed in applications claiming priority to this application or a related application. These claims, whether their scope is wider, narrower, the same or different than the scope of the original claims, are also deemed to be included in the subject matter of the present disclosure.
Claims
1. A method for determining the size of a nucleic acid fragment population provided by a sample, the method comprising: forming partitions each containing a portion of the sample, wherein each of the at least two targets is present only in a portion of the nucleic acid fragment; performing an amplification reaction on each of at least two targets in the partition; collecting amplification data for at least two targets from the partitions; enumerating a subset of the target definitions for the partitions using the amplification data; as well as The enumeration results are used to predict at least one length characteristic of the population of nucleic acid fragments. 2 . The method of claim 1 , wherein the predicting comprises predicting whether the population of nucleic acid fragments meets one or more predefined size criteria.
3. The method of claim 2, wherein the predicting comprises classifying the length of the population relative to a threshold size.
4. The method of claim 3, wherein the prediction comprises predicting whether the median length or the average length of the nucleic acid fragment population is above or below a threshold size.
5. The method of claim 4, wherein the prediction comprises calculating the log odds that the median length or average length of the nucleic acid fragment population is above or below a threshold size.
6. The method of claim 1, wherein the predicting comprises predicting a percentage value of a population of nucleic acid fragments that meet one or more predefined size criteria.
7. The method of claim 6, wherein the percentage value represents the percentage of a population of nucleic acid fragments having a predefined relationship to a threshold length.
8. The method of claim 7, wherein the threshold length is 200 base pairs or nucleotides.
9. The method of claim 7, wherein the at least two targets include a first target and a second target, and wherein the first target is longer than the second target and has a length within 10% of the threshold length.
10. The method of claim 7, wherein the prediction includes predicting a confidence interval for the percentage value.
11. The method of claim 1, wherein the prediction comprises predicting the mean, median, quartile and / or quintile length of the nucleic acid fragment population.
12. The method of claim 11, wherein the predicting comprises predicting the median length of nucleic acid fragments in the population.
13. The method of claim 12, wherein the predicting comprises predicting a confidence interval for the median length.
14. The method of claim 1, wherein the enumeration comprises obtaining two or more enumeration values, and wherein the prediction comprises combining the two or more enumeration values or values derived therefrom as input values of independent variables of a prediction function to generate a length feature representing the at least one length feature.
15. A method according to claim 14, wherein the combination comprises linearly combining the enumerated values or values derived therefrom with each other according to the prediction function, and wherein the linear combination comprises multiplying each enumerated value or each value derived therefrom by a corresponding coefficient and summing the products. The method of claim 15 , wherein the coefficients are unequal to each other. The method according to claim 15 , wherein the coefficients are not equal to 1.
18. The method of claim 15, wherein summing the products of the multiplications comprises summing the products and a constant with each other, and wherein the constant is an intercept of the prediction function.
19. The method of claim 14, wherein the predicting comprises combining at least three enumerated values or corresponding values derived therefrom with one another according to the prediction function.
20. The method of claim 19, wherein the prediction comprises linearly combining exactly three enumerated values according to a prediction function, or exactly three corresponding values derived therefrom.
21. The method of claim 19, wherein at least three enumeration values are obtained directly or indirectly from a subset of partitioned target definitions, wherein the subset of partitioned target definitions is defined only for positive / negative of a first target and a second target of at least two targets.
22. The method according to claim 1, wherein the prediction is performed by a size prediction algorithm using a linear prediction function.
23. A method according to claim 1, wherein the enumeration includes obtaining a first partition count from a first subset of the partitions or from the inverse of the first subset of the partitions, and obtaining a second partition count from a second subset of the partitions or from the inverse of the second subset of the partitions, wherein the first subset of the partitions is positive for a first target among at least two targets, wherein the second subset of the partitions is positive for a second target among at least two targets, and wherein the prediction uses the first partition count and the second partition count or one or more values derived therefrom as input values of independent variables of a prediction function to produce an output value of a length feature representing at least one length feature.
24. The method of claim 23, wherein a first subset of the partitions are negative for the second target, and wherein a second subset of the partitions are negative for the first target.
25. The method of claim 23, wherein the predicting uses the first partition count and the second partition count as direct inputs to the prediction function.
26. The method of claim 23, wherein the prediction uses values derived from the first partition count and the second partition count, respectively, as inputs to the prediction function.
27. The method of claim 26, wherein the prediction uses one or more target concentration values, positive / negative score values, normalized values, and / or target ratio values derived from the first partition counts and the second partition counts as inputs to the prediction function.
28. The method of claim 27, wherein the prediction uses a positive score value or a negative score value, each of which is calculated using one of the first and second partition counts and the total non-target-dependent count of the partition.
29. The method of claim 27, wherein the prediction uses a target ratio value calculated using the first partition count and the second partition count.
30. A method according to claim 23, wherein the first target and the second target have a certain degree of association with each other in the sample, wherein the first subset of the partitions is negative for the second target, wherein the second subset of the partitions is negative for the first target, wherein the enumeration includes obtaining a third partition count that is positive for the first target and the second target from the third subset of the partitions, or obtaining a third partition count from the inverse of the third subset of the partitions, and wherein the prediction uses the first partition count, the second partition count and the third partition count or values derived therefrom as inputs of the independent variables of the prediction function.
31. The method of claim 30, wherein the prediction uses three inputs to the prediction function.
32. The method of claim 30, wherein the predicting uses the first, second, and third partition counts as direct inputs to a prediction function.
33. The method of claim 30, wherein the predicting uses one or more values derived from the first, second, and third partition counts as inputs to a prediction function.
34. The method of claim 30, wherein the sample represents a genome wherein the first target and the second target are separated from each other by less than 200 base pairs.
35. The method of claim 30, wherein the degree of association of the first target and the second target with each other in the sample is correlated with the median length of the nucleic acid fragment population.
36. A method according to claim 1, wherein the prediction utilizes a prediction function having coefficient values estimated from a data set comprising a plurality of data points, wherein each data point represents one standard in a series of standards having different amounts of nucleic acid fragmentation, wherein the data points include input values of two or more independent variables of the prediction function and output values of a dependent variable of the prediction function, wherein the values of the two or more independent variables of the prediction function are derived from amplification data of at least two targets collected from a set of partitions in a digital experiment, each partition comprising a portion of a standard, and wherein the output value of the dependent variable of the prediction function is obtained by measuring the length of the nucleic acid fragment of the standard separately from the digital experiment.
37. The method of claim 36, wherein logistic regression is used to estimate the coefficient values.
38. The method of claim 36, wherein linear regression is used to estimate the coefficient values.
39. The method of claim 1, wherein the at least two targets include a first target and a second target, and wherein the second target is at least 50% longer than the first target.
40. The method of claim 1, wherein the at least two targets comprise a first target and a second target, wherein forming comprises forming the partitions to each contain a first probe for detecting amplification of the first target and a second probe for detecting amplification of the second target, and wherein the first probe comprises a first marker and the second probe comprises a second marker.
41. The method of claim 40, wherein the first marker and the second marker comprise fluorophores that are different from each other.
42. The method of claim 41, wherein collecting amplification data comprises detecting fluorescence from the first marker and the second marker, wherein enumeration of the first target is based on fluorescence from the first marker, and wherein enumeration of the second target is based on fluorescence from the second marker.
43. The method of claim 1, wherein the at least two targets include a first target and a second target, and wherein forming includes forming the partitions to each contain a first primer pair for amplifying the first target and a second primer pair for amplifying the second target.
44. The method of claim 1, wherein the at least two targets include a first target and a second target, and wherein the second target is at least 50 nucleotides longer than the first target.
45. The method of claim 1, wherein the at least two targets include a first target and a second target, wherein the sample represents a genome, and wherein the first target and the second target are separated from each other by less than 100 nucleotides in the genome.
46. The method of claim 1, wherein the nucleic acid fragments comprise fragments of genomic DNA, RNA, or reverse transcribed RNA.
47. A method for obtaining a size prediction algorithm for determining the size of a population of nucleic acid fragments in a sample, the method comprising: providing a series of standards, each standard in the series having a different degree of nucleic acid fragmentation and comprising at least two targets; forming sets of partitions, each set of partitions containing a portion of one of the standards and containing each of the at least two targets that are partially occupied; performing an amplification reaction on each of at least two targets in each set of partitions; collecting amplification data for at least two targets from each set of partitions; using the amplification data to enumerate subsets of target definitions for each partition set; For each standard, measuring the length of the nucleic acid fragments present in the standard; as well as A size prediction algorithm is generated based on the results of the enumeration and measurements.
48. The method of claim 47, wherein the generating comprises estimating values of coefficients and an intercept of a prediction function used by the size prediction algorithm.
49. The method of claim 48, wherein the generating uses values obtained or derived by enumeration as input values of independent variables of the prediction function, and wherein, Measurement uses values obtained or derived through measurement as output values of the independent variables of the prediction function.
50. The method of claim 48, wherein said generating comprises estimating values of coefficients and an intercept by logistic regression.
51. The method of claim 48, wherein said generating comprises estimating values of coefficients and an intercept by linear regression.
52. The method of claim 47, wherein the measuring comprises electrophoresis of the nucleic acid fragments of each standard.
53. The method of claim 47, further comprising fragmenting the nucleic acid polymers to generate each standard.
54. The method of claim 53, wherein the fragmenting comprises chemically fragmenting the nucleic acid polymers.
55. The method of claim 54, wherein the chemical fragmentation comprises performing an enzyme-catalyzed chemical reaction to break the nucleic acid polymer.
56. The method of claim 53, wherein the fragmenting comprises physically breaking the nucleic acid polymer.
57. A computer readable medium comprising program instructions for determining the size of nucleic acid fragments in a sample, wherein execution of the program instructions by one or more processors of a computer system causes the one or more processors to perform a method comprising: At least one length characteristic of a population of nucleic acid fragments is predicted using values obtained or derived by enumerating a target-defined subset of a partition of at least two targets in a digital amplification assay, wherein the partition contains partially occupied each of the at least two targets, each of the at least two targets provided by the population of nucleic acid fragments.
58. The computer readable medium of claim 57, wherein the computer readable medium is non-transitory.
59. The computer readable medium of claim 57, wherein the computer readable medium is transient.
60. A computer-readable medium according to claim 57, wherein the prediction uses a first partition count and a second partition count, or one or more values derived therefrom, as input to a size prediction algorithm, wherein the first partition count represents a first subset of the partitions, or represents the inverse of the first subset of the partitions, and the second partition count represents a second subset of the partitions, or represents the inverse of the second subset of the partitions, wherein the first subset of the partitions is positive for a first target of at least two targets, and wherein the second subset of the partitions is positive for a second target of at least two targets.
61. The computer readable medium of claim 60, wherein a first subset of the partitions are negative for the second target, and wherein a second subset of the partitions are negative for the first target.
62. The computer readable medium of claim 60, wherein the predicting uses the first partition count and the second partition count as direct input to a size prediction algorithm.
63. The computer readable medium of claim 60, wherein the prediction uses one or more values derived from the first partition count and the second partition count as input to a size prediction algorithm.
64. The computer-readable medium of claim 63, wherein the prediction uses one or more target concentration values, normalization values, and / or positive / negative score values derived from the first and second partition counts as input to the size prediction algorithm.
65. The computer readable medium of claim 64, wherein the prediction uses positive score values, each calculated using one of the first and second partition counts and the total non-target-dependent counts of the partition, as input to a size prediction algorithm.
66. A computer-readable medium according to claim 60, wherein the first target and the second target have a certain degree of correlation with each other in the sample, wherein the first subset of the partitions is negative for the second target, wherein the second subset of the partitions is negative for the first target, wherein the third partition count represents a third subset of the partitions that are positive for the first target and the second target, or represents the inverse of the third subset, and wherein the prediction uses the first partition count, the second partition count and the third partition count, or values derived therefrom, as input to the size prediction algorithm.
67. The computer readable medium of claim 66, wherein the prediction uses three inputs, optionally exactly three inputs for the size prediction algorithm.
68. The computer readable medium of claim 66, wherein the prediction uses the first, second, and third partition counts as direct input to a size prediction algorithm.
69. The computer-readable medium of claim 66, wherein the prediction uses one or more values derived from the first, second, and third partition counts as input to a size prediction algorithm.
70. The computer-readable medium of claim 57, wherein the predicting comprises predicting a mean, median, quartile, and / or quintile length of the population of nucleic acid fragments.
71. The computer-readable medium of claim 57, wherein the predicting comprises predicting whether a median length or an average length of the population of nucleic acid fragments is above or below a threshold size.
72. The computer-readable medium of claim 71, wherein the predicting comprises calculating the log odds that the median or average length of the population of nucleic acid fragments is above or below a threshold size.
73. The computer readable medium of claim 57, wherein the prediction is made using a size prediction algorithm generated by logistic regression.
74. The computer readable medium of claim 57, wherein the prediction is made using a size prediction algorithm generated by linear regression.
75. The computer-readable medium of claim 57, wherein the prediction is performed using a size prediction algorithm that uses quantified data obtained for each of the at least two targets obtained from each standard in a series of standards representing different amounts of nucleic acid fragments.
76. The computer readable medium of claim 75, wherein the size prediction algorithm is generated using a data set comprising the quantitative data and length characteristics of each standard in the series of standards.
77. A system for determining the size of a population of nucleic acid fragments, the system comprising: a first primer pair for amplifying a first target; a second primer pair for amplifying a second target; and A computer-readable medium containing program instructions executable by one or more processors of a computer system for using values obtained or derived from enumerating a subset of target definitions of partitions in a digital amplification experiment to predict at least one length characteristic of a population of nucleic acid fragments, each partition containing a first primer pair and a second primer pair, the partition containing partially occupied each of a first target and a second target, the first target and the second target being provided by the population of nucleic acid fragments.
78. The system of claim 77, further comprising a first probe for detecting amplification of the first target using the first primer pair and a second probe for detecting amplification of the second target using the second primer pair.
