Determine linear and circular forms of circulating nucleic acids

Abstract

A technique for analyzing circular DNA in a biological sample is provided, which includes cutting with a restriction enzyme or a transposase to linearize the circular DNA so that it can be sequenced; identifying the linearized DNA molecules using one or more standards; and analyzing the identified molecules to measure a property of the biological sample.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application is a non-provisional application and claims the benefit of U.S. Provisional Patent Application No. 62 / 823,567, filed on March 25, 2019, entitled “Determining Linear And Circular Forms of Circulating Nucleic Acids,” which is incorporated herein by reference in its entirety for all purposes.

[0003] background

[0004] Extrachromosomal circular DNA (eccDNA) is a circular form of DNA that exists independently of chromosomal DNA (Zhu et al., Sci Rep. 2017; 7 (1): 10968). According to electron microscopy, they were first found in wheat and wild boar DNA (Hotta et al., Proc Natl Acad Sci US A. 1965; 53: 356-62). Later researchers found that these forms of DNA are widely present in the tissues of all organisms (Gaubatz. Mutat Res. 1990; 237 (5-6): 271-292). In addition, it has been revealed that eccDNA can be detected in mice (Kumar et al., Mol Cancer Res. 2017; 15: 1197-1205) and human plasma (Zhu et al., Sci Rep. 2017; 7: 10968).

[0005] Mitochondrial DNA has been detected in plasma (Chiu et al., Clin Chem 2003; 49: 719-726 and Lo et al., Sci Transl Med 2010; 2: 61-ra91). Mitochondrial DNA has been measured in the plasma of cancer patients, but these measurements are inconsistent (Yu M et al., Mitochondrial DNA 2012; 23: 329-32; Zachariah RR et al., Obstet Gynecol 2008; 112: 843-50; Mehra N et al., Clin Cancer Res 2007; 23: 421-6; Kohler et al., Mol Cancer 2009; 8: 105; and Choudhuri et al., Mol Cell Biochem 2014; 386: 259-269).

[0006] Overview

[0007] Various embodiments of the present disclosure can provide techniques for analyzing circular DNA in biological samples, which may include cells and / or cell-free DNA, such as plasma. For example, in order to measure circular DNA, cutting can be performed to linearize the circular DNA so that they can be sequenced. Example cutting techniques include restriction enzymes and transposases. Then, one or more criteria can be used to identify linearized DNA molecules, for example, to distinguish them from linear DNA molecules. An example criterion is mapping a pair of reverse end sequences to a reference genome. Another example criterion is the identification of a cutting tag associated with a restriction enzyme or an adapter sequence added by a transposase.

[0008] Once circular DNA molecules (e.g., eccDNA and circular mitochondrial DNA) are identified, they can be analyzed (e.g., to determine counts, size profiles, and / or methylation) to measure properties of the biological sample. Example properties include detecting copy number aberrations in chromosomal regions, which in turn can be used to detect levels of disease (e.g., cancer). Disease levels can be detected directly, using aberrations, such as using methylation. Another example includes identifying tissue types or diseases in cell tissues based on the amount of eccDNA.

[0009] In addition, some embodiments can provide methods for simultaneously analyzing short linear and circular mtDNA molecules. For example, the present disclosure allows (1) quantification of the relative amounts between linear and circular forms of cell-free mtDNA molecules in a plasma DNA pool, for example, to determine disease levels; and (2) inference of the tissue of origin of linear and circular mtDNA molecules in a plasma DNA pool, for example, as part of determining whether a non-hematopoietic or hematopoietic tissue has a sequence variant. The identification of sequence variants can be further used to identify a disease (e.g., cancer) and original information about the disease.

[0010] These and other embodiments of the present disclosure are described in detail below.For example, other embodiments relate to systems, devices, and computer-readable media related to the methods described herein.

[0011] A better understanding of the nature and advantages of embodiments of the present disclosure may be obtained by reference to the following detailed description and accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 An exemplary technique for eccDNA identification according to an embodiment of the present disclosure is shown.

[0014] Figure 2A and 2B An illustrative method for a junction search method according to an embodiment of the present disclosure is shown.

[0015] Figures 3A-5B Size analysis of one pregnancy case (MspI treated) according to an embodiment of the present disclosure is shown.

[0016] Figures 6A-7B Size analysis of maternal-specific eccDNA and fetal-specific eccDNA (5 cases combined) is shown.

[0017] Figure 8 Annotation of the genomic location of eccDNA is shown.

[0018] Fig. 9 Shown is the genomic location of eccDNA from one pregnancy plasma sample treated with MspI and HpaII.

[0019] Fig. 10A and 10B Shown is an example of the principle of eccDNA identification using transposase-based tagmentation according to an embodiment of the present disclosure.

[0020] Fig.11 Shown is size analysis of eccDNA using a tagmentation scheme according to an embodiment of the present disclosure.

[0021] Fig.12 An example workflow for identification and methylation analysis of eccDNA by enzymatic conversion is shown, according to an embodiment of the present disclosure.

[0022] Fig.13A and 13B Shown are size analysis and cumulative frequency of eccDNA in human plasma according to embodiments of the present disclosure.

[0023] Fig.14 The methylation levels of linear and eccDNA molecules in different chromosomes were compared.

[0024] Fig.15 An example workflow for identification and methylation analysis of eccDNA by bisulfite conversion is shown, according to an embodiment of the present disclosure.

[0025] Fig.16 is a flow chart illustrating a technique for analyzing circular nuclear DNA according to an embodiment of the present disclosure.

[0026] Fig.17 An exemplary technique for distinguishing between cell-free circular and linear-derived mtDNA molecules in plasma is shown, in accordance with embodiments of the present disclosure.

[0027] Fig.18The use of the BfaI restriction enzyme to demonstrate various principles is demonstrated.

[0028] Fig.19A and 19B Shown is a comparison of plasma mtDNA metrics of plasma DNA with and without restriction enzyme (BfaI) digestion, according to embodiments of the present disclosure.

[0029] Fig. 20 Shown are the size profiles of mtDNA fragments sequenced under different treatments.

[0030] Fig.21 A and 21B show the proportion of mtDNA >200 bp in all autosomes and all mtDNA.

[0031] Fig. 22 Shows the Fig. 20 Enzymatic cleavage in 2020 and Fig.18 Size profile of mtDNA fragments with two cleaved ends in the simulation results.

[0032] Fig.23 Shown are MtDNA fragments in plasma DNA that was not treated with restriction enzymes sequenced by the Pacific Biosciences Single Molecule Real-Time (SMRT) sequencing platform.

[0033] Fig.24A The correlation between the proportion of mtDNA in plasma with and without enzymatic cleavage treatment according to embodiments of the present disclosure is shown. Fig. 24B Shown is the correlation of the proportion of linear-derived mtDNA in all DNA (nuclear and mitochondrial) of plasma DNA treated with enzymatic treatment to the proportion of all mtDNA in all plasma DNA (nuclear and mitochondrial) without treatment, according to embodiments of the present disclosure.

[0034] Fig.25A and 25B Shown is a comparison of plasma mtDNA metrics for plasma DNA of HBV and HCC samples with and without restriction enzyme digestion.

[0035] Fig. 27 Shown is a targeted capture based TopM analysis according to an embodiment of the present disclosure.

[0036] Fig.28 It is shown that enzymatic cleavage and target enrichment improve the detection of mtDNA fragments according to embodiments of the present disclosure.

[0037] Fig.29Shown are example techniques illustrated using a liver transplant model for analysis of linear and circular mtDNA, according to embodiments of the present disclosure.

[0038] Fig.30 A table 3000 illustrating statistics of liver transplant cases according to an embodiment of the present disclosure is shown.

[0039] Fig.31 Shown is the analysis of linear and circular mtDNA molecules in plasma of liver transplant cases according to embodiments of the present disclosure.

[0040] Fig.32 Shown is the correlation of mtDNA fraction (determined by sequencing without enzyme treatment) with liver DNA fraction (determined by methylation analysis using ddPCR) according to embodiments of the present disclosure.

[0041] Fig.33 The difference in the amount of linear and circular mtDNA between healthy controls and liver transplant patients according to embodiments of the present disclosure is shown.

[0042] Fig.34 The proportion of linear and circular mtDNA molecules among all mtDNA molecules according to embodiments of the present disclosure is shown.

[0043] Figures 35A-35C The differences between HCC and non-HCC subjects are shown for the percentage of mtDNA in all molecules (35A), the percentage of linear mtDNA in all mtDNA molecules (35B), and the percentage of circular mtDNA in all mtDNA molecules (35C).

[0044] Fig.36 An exemplary technique for analyzing linear and circular mtDNA using a pregnancy model according to an embodiment of the present disclosure is shown.

[0045] Fig.37 Shown is an example of mutation load calculation for a linear mtDNA molecule, according to an embodiment of the present disclosure.

[0046] Fig.38 is a flow chart illustrating a method 2700 for analyzing circular mitochondrial DNA (mtDNA) and linear mtDNA according to an embodiment of the present disclosure.

[0047] Fig.39 is a flow chart illustrating a method 2800 for analyzing circular mitochondrial DNA according to an embodiment of the present disclosure.

[0048] Fig.40A system according to an embodiment of the present disclosure is described.

[0049] Fig.41 A block diagram of an example computer system that may be used with systems and methods according to embodiments of the present disclosure is shown.

[0050] the term

[0051] A "tissue" corresponds to a group of cells grouped together as a functional unit. More than one type of cell may be found in a single tissue. Different types of tissues may consist of different types of cells (e.g., liver cells, alveolar cells, or blood cells), but may also correspond to tissues from different organisms (mother vs. fetus) or to healthy cells vs. tumor cells. A "reference tissue" may correspond to a tissue used to determine tissue-specific methylation levels. Multiple samples of the same tissue type from different individuals may be used to determine tissue-specific methylation levels for that tissue type.

[0052] "Biological sample" refers to any sample collected from a subject (e.g., a human (or other animal), such as a pregnant woman, a person with cancer or a person suspected of having cancer, an organ transplant recipient, or a subject suspected of having a disease process involving an organ (e.g., the heart of a myocardial infarction, the brain of a stroke, or the hematopoietic system of anemia)) and containing one or more target nucleic acid molecules. The biological sample can be a body fluid, such as blood, plasma, serum, urine, vaginal fluid, hydrocele (e.g., testicular) fluid, vaginal washing fluid, pleural fluid, ascites fluid, peritoneal fluid, cerebrospinal fluid, saliva, sweat, tears, sputum, bronchoalveolar lavage fluid, cervical lavage fluid, nipple discharge, aspirate from different parts of the body (e.g., thyroid, breast), intraocular fluid (e.g., aqueous humor), etc. Stool samples can also be used. In various embodiments, a majority of the DNA in a biological sample that has been enriched for cell-free DNA (e.g., a plasma sample obtained via a centrifugation protocol) can be cell-free (e.g., more than 50%, 60%, 70%, 80%, 90%, 95%, or 99% of the DNA can be cell-free). The centrifugation protocol can include, for example, 3,000 g x 10 minutes, obtaining a fluid portion, and centrifuging again at, for example, 30,000 g for 10 minutes to remove residual cells. As part of the analysis of the biological sample, at least 1,000 cell-free DNA molecules can be analyzed. As other examples, at least 10,000 or 50,000 or 100,000 or 500,000 or 1,000,000 or 5,000,000 cell-free DNA molecules or more can be analyzed.

[0053] As used herein, the term "fragment" (e.g., DNA fragment) may refer to a portion of a polynucleotide or polypeptide sequence that contains at least 3 consecutive nucleotides. A nucleic acid fragment may retain the biological activity and / or some characteristics of a parent polypeptide. A nucleic acid fragment may be double-stranded or single-stranded, methylated or unmethylated, intact or nicked, complexed or uncomplexed with other macromolecules (e.g., lipid particles, proteins). A nucleic acid fragment may be a linear fragment or a circular fragment. Tumor-derived nucleic acid may refer to any nucleic acid released from a tumor cell, including pathogenic nucleic acids from pathogens in a tumor cell.

[0054] The term "assay" generally refers to a technique for determining the properties of a nucleic acid or nucleic acid sample (e.g., a statistically significant number of nucleic acids) and the properties of an object (from which a sample is obtained). An assay (e.g., a first assay or a second assay) generally refers to a technique for determining the number of nucleic acids in a sample, the genomic identity of nucleic acids in a sample, the copy number variation of nucleic acids in a sample, the methylation state of nucleic acids in a sample, the fragment size distribution of nucleic acids in a sample, the mutation state of nucleic acids in a sample, or the fragmentation pattern of nucleic acids in a sample. Any assay known to those skilled in the art can be used to detect any of the properties of nucleic acids mentioned herein. The properties of nucleic acids include sequence, number, genomic identity, copy number, methylation state at one or more nucleotide positions, nucleic acid size, mutations in nucleic acids at one or more nucleotide positions, and fragmentation patterns of nucleic acids (e.g., nucleotide positions at which nucleic acid fragments are located). The term "assay" can be used interchangeably with the term "method". An assay or method can have a specific sensitivity and / or specificity (e.g., based on the selection of one or more cutoff values), and its relative effectiveness as a diagnostic tool can be measured using the area under the receiver operating characteristic (ROC) curve (AUC) statistic.

[0055] "Sequence read" refers to a nucleotide string sequenced from any part or all of a nucleic acid molecule. For example, a sequence read can be a whole nucleic acid fragment present in a biological sample. For example, a sequence read can be a short nucleotide string (e.g., 20-150 bases) sequenced from a nucleic acid fragment, a short nucleotide string at one or both ends of a nucleic acid fragment, or a sequencing of the whole nucleic acid fragment present in a biological sample. Sequence reads can be obtained in a variety of ways, such as using sequencing techniques or using probes (e.g., by hybridization arrays or capture probes) or amplification techniques (such as polymerase chain reaction (PCR) or linear amplification or isothermal amplification using a single primer), or based on biophysical measurements (e.g., mass spectrometry). Sequence reads can be obtained from single molecule sequencing. As part of a biological sample analysis, at least 1,000 sequence reads can be analyzed. As other examples, at least 10,000 or 50,000 or 100,000 or 500,000 or 1,000,000 or 5,000,000 sequence reads or more can be analyzed. The sequence reads can be aligned (mapped) to a reference genome to determine their position in the reference genome. Various software packages can be used to perform this alignment.

[0056] "Single molecule sequencing" refers to sequencing a single template DNA molecule to obtain a sequence read without interpreting the base sequence information from a cloned copy of the template DNA molecule. Single molecule sequencing can sequence the entire molecule or only a portion of the DNA molecule. Most of the DNA molecule can be sequenced, for example, greater than 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99%. The sequence read (or the read from the two ends) can be compared with the reference genome. When the two ends are compared (for example, as part of the read of the entire fragment or for paired ends), higher accuracy can be achieved in the comparison, and the length of the fragment can be obtained.

[0057] A sequence read may include a "termination sequence" (or "end sequence") associated with the end of a fragment. The termination sequence may correspond to the outermost N bases of the fragment, such as 2-30 bases at the end of the fragment. If the sequence read corresponds to the entire fragment, the sequence read may include two termination sequences. When paired-end sequencing provides two sequence reads corresponding to the ends of the fragment, each read sequence may include one termination sequence.

[0058] A "sequence motif" (or just "motif") can refer to a short, repeating pattern of bases in a DNA fragment (e.g., a cell-free DNA fragment). A sequence motif can occur at the end of a fragment, and thus be part of or include a terminator sequence. A "terminal motif" can refer to a sequence motif that preferentially occurs at a terminator sequence at the end of a DNA fragment (perhaps for a particular type of tissue). A terminal motif can also occur just before or after the end of a fragment, and thus still correspond to a terminator sequence.

[0059] A "cleavage tag" may refer to a short sequence at the end of a DNA fragment, indicating that the DNA has been cleaved at that location. For example, a cleavage tag may be generated by enzymatic digestion (restriction enzyme or transposase) and may be a different tag or a repeated sequence. When a restriction enzyme is used, a "cleavage tag" may be a "terminal motif".

[0060] As used herein, the term "locus" or its plural form "loci" is a position or address of nucleotides (or base pairs) of any length. A locus can have variation on a genome.

[0061] "Termination position" or "terminal position" (or just "terminal") may refer to the genomic coordinates or genomic identity or nucleotide identity of the outermost base (i.e., at the end) of a cell-free DNA molecule, such as a plasma DNA molecule. The terminal position may correspond to either end of a DNA molecule. In this manner, if one end refers to the beginning and end of a DNA molecule, both may correspond to termination positions. In practice, one terminal position is the genomic coordinates or nucleotide identity of the outermost base on one end of a cell-free DNA molecule detected or determined by an analytical method, such as (but not limited to) massively parallel sequencing or next generation sequencing, single molecule sequencing, double-stranded or single-stranded DNA sequencing library preparation protocols, polymerase chain reaction (PCR), or microarray. Such in vitro techniques may alter the true in vivo physical ends of cell-free DNA molecules. Thus, each detectable end may represent a biologically true end or the end may be one or more inwardly directed nucleotides or one or more nucleotides extending from the original end of the molecule, such as the overhang of a non-blunt-ended double-stranded DNA molecule by 5' blunting and 3' filling of a Klenow fragment. The genomic identity or genomic coordinates of the end position can be derived from the alignment of the sequence read with a human reference genome such as hg19. It can be derived from the catalog number of the index or code representing the initial coordinates of the human genome. It can refer to the position or nucleotide identity on the cell-free DNA molecule read by (but not limited to) target-specific probes, microsequencing, DNA amplification. The term "genomic position" can refer to the nucleotide position in a polynucleotide (e.g., a gene, a plasmid, a nucleic acid fragment, a viral DNA fragment, or a genome). The term "genomic position" is not limited to the nucleotide position within the genome (e.g., the haploid set of chromosomes in a gamete or microorganism or in each cell of a multicellular organism).

[0062] The term "allele" refers to an optional DNA sequence at the same physical genome locus, which may or may not result in different phenotypic traits. In any specific diploid organism, each chromosome has two copies (except the sex chromosomes in male human subjects), and the genotype of each gene includes a pair of alleles present on the locus, which are identical in homozygotes and different in heterozygotes. The colony or species of an organism usually comprises multiple alleles on each locus of each individual. The genomic locus of more than one allele is found in the colony and is called a polymorphic site. The allele variation at the locus can be measured as the number of alleles present (i.e., the degree of polymorphism) or the ratio of heterozygotes in the colony (i.e., heterozygosity). As used herein, the term "polymorphism" refers to any inter-individual variation in the human genome, regardless of its frequency. Examples of such variations include, but are not limited to, single nucleotide polymorphisms, simple tandem repeat polymorphisms, insertion-deletion polymorphisms, mutations (which may be disease-causing) and copy number variations. The term "haplotype" as used herein refers to a combination of alleles at multiple loci that are transmitted together on the same chromosome or chromosome region. A haplotype can refer to as few as a pair of loci or chromosome regions or an entire chromosome or chromosome arm.

[0063] The term "fetal DNA fraction concentration" is used interchangeably with the terms "fetal DNA ratio" and "fetal DNA fraction" and refers to the proportion of fetal DNA molecules present in a biological sample derived from the fetus (e.g., a maternal plasma or serum sample) (Lo et al., Am J Hum Genet. 1998; 62: 768-775; Lun et al., Clin Chem. 2008; 54: 1664-1672). Similarly, the tumor fraction or tumor DNA fraction can refer to the fractional concentration of tumor DNA in a biological sample.

[0064] The terms "size profile" and "size distribution" generally relate to the sizes of DNA fragments in a biological sample. A size profile can be a histogram that provides a distribution of the amounts of DNA fragments of various sizes. Various statistical parameters (also referred to as size parameters or simply parameters) can distinguish one size profile from another. One parameter is the percentage of DNA fragments of a particular size or size range relative to all DNA fragments or relative to DNA fragments of another size or range.

[0065] The term "parameter" as used herein means a numerical value characterizing a quantitative data set and / or a numerical relationship between a quantitative data set. For example, the ratio (or a function of a ratio) between the first amount of a first nucleic acid sequence and the second amount of a second nucleic acid sequence is a parameter. A separation value is an example of a parameter." Separation value" (or relative abundance) corresponds to two values, such as two amounts of a DNA molecule, two fractional contributions or two methylation levels, such as a difference or a ratio between a sample (mixture) methylation level and a reference methylation level. A separation value can be a simple difference or a ratio. As an example, the direct ratio of x / y and x / (x+y) are separation values. A separation value can include other factors, such as a multiplication factor. As other examples, a difference or a ratio of a function of a value can be used, such as a difference or a ratio of the natural logarithm (ln) of two values. A separation value can include a difference and / or a ratio. "Methylation level" is an example of a relative abundance between, for example, a methylated DNA molecule (for example, at a specific site) and other DNA molecules (for example, all other DNA molecules at a specific site or only unmethylated DNA molecules). The amount of other DNA molecules can serve as a normalization factor. As another example, the intensity (e.g., fluorescence or electrical intensity) of the methylated DNA molecules relative to the intensity of all or unmethylated DNA molecules can be determined. Relative abundance can also include the intensity per volume.

[0066] "Classification" can refer to any number or other character associated with a particular property of a sample. For example, a "+" symbol (or the term "positive") can indicate that a sample is classified as having a deletion or amplification. In another example, the term "classification" can refer to the amount of tumor tissue in an object and / or a sample, the size of a tumor in an object and / or a sample, the stage of a tumor in an object, the tumor load in an object and / or a sample, and the presence of tumor metastasis in an object. Classification can be binary (e.g., positive or negative) or have more classification levels (e.g., a scale of 1 to 10 or 0 to 1).

[0067] The terms "cutoff value" and "threshold value" may refer to a predetermined number used in an operation. A threshold value or cutoff value may be a value above or below which a specific classification is applicable, such as a classification of a condition, such as whether an object suffers from a condition or the severity of a condition. A cutoff value or threshold value may be a "reference value", or may be derived from a reference value representing a specific classification or distinguishing between two or more classifications. A cutoff value may be predetermined with reference to or without reference to the characteristics of a sample or object. For example, a cutoff value may be selected based on the age or sex of a test object. A cutoff value may be selected after the test data is output and based on the output of the test data. For example, when the sequencing of a sample reaches a certain depth, certain cutoff values ​​may be used. As another example, a reference object with a known classification of one or more conditions and a measured feature value (e.g., methylation level, statistical size value or count) may be used to determine a reference level to distinguish between different conditions and / or condition classifications (e.g., whether an object suffers from the condition). A reference value may be selected to represent a classification (e.g., an average value) or a value between two clusters of a metric (e.g., selected to obtain a desired sensitivity and specificity). As another example, a reference value can be determined based on statistical simulations of samples.Any of these terms may be used in any of these contexts.

[0068] "Non-hematopoietic tissue source" refers to any organ other than the blood system. Examples include liver, lung, heart, brain, non-hematopoietic cancers, placenta, etc.

[0069] The term "nuclear DNA" refers to DNA derived from the nucleus of a cell. "Nuclear genome" corresponds to nuclear DNA derived from the nucleus of a cell. "Mitochondrial genome" corresponds to DNA derived from the mitochondria in a cell.

[0070] The term "cancer level" (or more generally, "disease level" or "condition level" or "disorder level") can refer to whether cancer exists (i.e., exists or does not exist), the stage of cancer, tumor size, whether metastasis exists, the overall tumor load of the body, the response of cancer to treatment, and / or other measures of cancer severity (e.g., cancer recurrence). Cancer levels can be numbers (e.g., probabilities) or other markers, such as symbols, letters, and colors. The level can be zero. Cancer levels can also include pre-malignant or precancerous conditions (states). Cancer levels can be used in a variety of ways. For example, screening can check whether someone who was previously not aware of having cancer has cancer. Assessment can investigate someone who has been diagnosed with cancer to monitor the progression of cancer over time, study the effectiveness of treatment, or determine prognosis. In one embodiment, prognosis can be expressed as the chance of a patient dying of cancer, or the chance of cancer progression after a specific duration or time, or the chance of cancer metastasis. Detection can mean "screening", or can mean checking whether someone with cancer-suggestive features (e.g., symptoms or other positive tests) has cancer.

[0071] The term "sequence imbalance" or "aberration" as used herein means any significant deviation from a reference amount defined by at least one cutoff value in the amount of a clinically relevant chromosome region. Sequence imbalance can include chromosome dosage imbalance, allele imbalance, mutation dosage imbalance, copy number imbalance, haplotype dosage imbalance, and other similar imbalances. As an example, when one allele of a gene of a tumor is deleted or one allele of a gene is amplified or two alleles in its genome are differentially amplified, allele imbalance can occur, thereby generating an imbalance at a specific locus of the sample. As another example, a patient can have a genetic mutation in a tumor suppressor gene. Then, the patient can go on to develop a tumor in which the unmutated allele of the tumor suppressor gene is deleted. Therefore, within the tumor, there is a mutation dosage imbalance. When the tumor releases its DNA into the patient's plasma, the tumor DNA will mix with the constituent DNA (from normal cells) in the patient's plasma. Aberrations can include deletions or amplifications of chromosome regions.

[0072] "DNA methylation" in mammalian genomes generally refers to the addition of a methyl group to the 5' carbon of cytosine residues in CpG dinucleotides (i.e., 5-methylcytosine). In other cases, DNA methylation can occur in cytosines, such as CHG and CHH, where H is adenine, cytosine, or thymine. Cytosine methylation can also be in the form of 5-hydroxymethylcytosine. Non-cytosine methylations, such as N6-methyladenine, have also been reported.

[0073] A "site" (also referred to as a "genomic site") corresponds to a single site, which can be a single base position or a group of related base positions, e.g., a CpG site or a larger group of related base positions. A "locus" can correspond to a region that includes multiple sites. A locus can include only one site, which would make the locus equivalent to a site in this context.

[0074] The "methylation index" of each genomic site (e.g., CpG site) can refer to the proportion of DNA fragments (e.g., determined by sequence reading) showing methylation at the site to the total number of reads covering the site. "Reading" can include information obtained from DNA fragments (e.g., the methylation state at the site). Reading can be obtained using reagents (e.g., primers or probes) that preferentially hybridize to DNA fragments of a specific methylation state. Typically, such reagents are applied after treatment with a method (e.g., bisulfite conversion, or methylation-sensitive restriction enzymes, or methylation-binding proteins, or anti-methylcytosine antibodies) or a single-molecule sequencing technique that recognizes methylcytosine and hydroxymethylcytosine depending on its methylation state.

[0075] " methylation density " of region can refer to the number of readings showing methylation at the site in the region divided by the total number of readings covering the site in the region. These sites can have specific features, such as CpG sites. Therefore, " CpG methylation density " of region can refer to the number of readings showing CpG methylation divided by the total number of readings covering the regional CpG sites (such as, specific CpG sites, CpG sites in CpG islands or larger regions). For example, the methylation density of every 100-kb stack in the human genome can be determined from the total number of unconverted cytosines (corresponding to methylated cytosine) at CpG sites after bisulfite treatment, and the CpG sites are read as a part of all CpG sites covered by the sequence mapped to the 100-kb region. This analysis can also be performed for other stack sizes, such as 500bp, 5kb, 10kb, 50kb or 1-Mb, etc. A region can be a whole genome or a chromosome or a part of a chromosome (such as, a chromosome arm). When a region contains only CpG sites, the methylation index of the CpG sites is the same as the methylation density of the region. The "proportion of methylated cytosines" can refer to the number of cytosine sites "C" that are shown to be methylated (e.g., not converted after bisulfite conversion) compared to the total number of cytosine residues analyzed in the region (i.e., including cytosines outside the CpG range). The methylation index, methylation density, and the proportion of methylated cytosines are examples of "methylation levels." In addition to bisulfite conversion, other methods known to those skilled in the art can be used to interrogate the methylation state of a DNA molecule, including but not limited to enzymes sensitive to methylation state (e.g., methylation-sensitive restriction enzymes), methylation binding proteins, single molecule sequencing using a platform sensitive to methylation state (e.g., nanopore sequencing (Schreiber et al., Proc Natl Acad Sci 2013; 110: 18910-18915)) and by Pacific Biosciences single molecule real-time analysis (Flusberg et al., Nat Methods 2010; 7: 461-465)). The methylation metric of a DNA molecule can correspond to the percentage of methylated sites (e.g., CpG sites). The methylation metric can be specified as an absolute number or a percentage, which can be referred to as the methylation density of the molecule.

[0076] "Methylation-aware sequencing" refers to any sequencing method that allows the methylation state of a DNA molecule to be determined during the sequencing process, including but not limited to bisulfite sequencing or sequencing followed by methylation-sensitive restriction enzyme digestion, immunoprecipitation using anti-methylcytosine antibodies or methylation binding proteins, or single-molecule sequencing that allows the elucidation of methylation state. "Methylation-aware assays" or "methylation-sensitive assays" can include sequencing-based and non-sequencing methods, such as MSP, probe-based interrogation, hybridization, restriction enzyme digestion followed by densitometry, anti-methylcytosine immunoassays, mass spectrometry interrogation of the proportion of methylated cytosine or hydroxymethyl cytosine, immunoprecipitation without sequencing, etc.

[0077] The terms "control," "control sample," "reference," "reference sample," "normal," and "normal sample" can be used interchangeably to generally describe a sample that does not have a particular condition, or is otherwise healthy. In one example, the methods disclosed herein can be performed on a subject with a tumor, wherein the reference sample is a sample taken from healthy tissue of the subject. In another example, the reference sample is a sample taken from a subject with a disease (e.g., cancer or a particular stage of cancer). The reference sample can be obtained from a subject or a database. A reference generally refers to a reference genome, which is used to map sequence reads obtained by sequencing a sample from a subject. A reference genome generally refers to a haploid or diploid genome, with which sequence reads from biological samples and constituent samples can be compared and compared. For a haploid genome, each locus has only one nucleotide. For a diploid genome, a heterozygous locus can be identified, such a locus having two alleles, either of which can allow alignment and matching with the locus. A reference genome can correspond to a virus, for example, by including one or more viral genomes.

[0078] The phrase "healthy" as used herein generally refers to a subject having good health. Such a subject exhibits the absence of any malignant or non-malignant disease. A "healthy individual" may suffer from other diseases or conditions unrelated to the condition being measured (which may not generally be considered "healthy").

[0079] The terms "cancer" or "tumor" are used interchangeably and generally refer to an abnormal mass of tissue in which the growth of the mass exceeds and is inconsistent with the growth of normal tissue. Cancer or tumors can be defined as "benign" or "malignant" based on the following characteristics: degree of cell differentiation, including morphology and function, growth ratio, local invasion and metastasis. "Benign" tumors are usually well differentiated, typically grow slower than malignant tumors, and remain confined to the site of origin. In addition, benign tumors do not have the ability to infiltrate, invade or metastasize to distant sites. "Malignant" tumors are usually poorly differentiated (dysplastic), typically grow rapidly, accompanied by progressive infiltration, invasion and destruction of surrounding tissues. In addition, malignant tumors have the ability to metastasize to distant sites. "Stage" can be used to describe the degree of progression of a malignant tumor. Early stage cancer or malignancy is associated with less tumor burden in the body, usually with fewer symptoms, better prognosis and better treatment response than late stage malignancies. Late or advanced stage cancer or malignancy is often associated with distant metastasis and / or lymphatic spread.

[0080] The term "about" or "approximately" may mean within an acceptable range of deviation of a specific value as determined by a person skilled in the art, which will depend in part on how the value is measured or determined, i.e., the limitations of the measurement system. For example, according to practice in the art, "about" may mean within 1 or more than 1 standard deviation. Alternatively, "about" may mean a range of up to 20%, up to 10%, up to 5%, or up to 1% of a given value. Alternatively, particularly with respect to a biological system or method, the term "about" or "approximately" may mean within the order of magnitude of a value, within 5 times and more preferably within 2 times. In the case of describing a specific value in the present application and claims, unless otherwise stated, it should be assumed that the term "about" means within an acceptable error range for the specific value. The term "about" may have a meaning as commonly understood by those skilled in the art. The term "about" may refer to ±10%. The term "about" may refer to ±5%.

[0081] The terms used herein are for the purpose of describing a particular situation only and are not intended to be limiting. As used herein, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. The term "based on" is intended to mean "based at least in part on". In addition, to the extent that the terms "including", "includes", "having", "has", "with", or variations thereof are used in the detailed description and / or claims, these terms are intended to be included in a manner similar to the term "comprising".

[0082] Details

[0083] Since the discovery of extrachromosomal circular DNA (eccDNA), researchers have tried various methods to gain knowledge about this special form of DNA. Early reports used electron microscopy and Southern blotting in the visualization and quantification of eccDNA (Gaubatz. Mutat Res. 1990; 237: 271-292). This method can only obtain limited information about the number and part of the sequence of eccDNA (as known by probe hybridization). More detailed information, such as the accurate complete sequence and genomic location of eccDNA is not easy to obtain. Moreover, the use of eccDNA to measure the characteristics of biological samples (e.g., for diagnostic purposes) is limited. Through some embodiments of the present disclosure, we have proposed a method for eccDNA analysis.

[0084] Compared with the plasma DNA fragments derived from nuclear DNA molecules, the cell-free mitochondrial DNA (mtDNA) fragments derived from the mitochondrial genome show completely different size characteristics (Lo et al., Sci Transl Med. 2010; 2: 61ra91; Jiang et al., Proc Natl Acad Sci USA. 2015; 112: E1317-E1325). When compared with the situation of nuclear DNA in plasma, for circulating mitochondrial DNA, there is a much higher ratio of shorter DNA molecules. In addition, the modal peak of 166-bp and the periodic small peak of 10-bp are no longer shown in the size profile of mitochondrial DNA in plasma. This disappearance of the nucleosome characteristics of plasma mitochondrial DNA molecules may be due to the lack of histone packaging in the mitochondrial genome (Lo et al., Sci Transl Med. 2010; 2: 61ra91; Jiang et al., Proc Natl Acad Sci USA. 2015; 112: E1317-E1325).

[0085] Using massively parallel sequencing, Jiang et al. reported that an increase in cell-free mtDNA molecules was associated with patients with hepatocellular carcinoma (HCC) compared to healthy controls, HBV carriers, and subjects with cirrhosis (Jiang et al., Proc Natl Acad Sci USA. 2015; 112: E1317-E1325). However, this massively parallel sequencing can only effectively sequence linear fragmented DNA, such as naturally fragmented cell-free nuclear DNA and mitochondrial DNA, and artificially sheared or cut genomic DNA. It is known that the mitochondrial genome is a naturally occurring circular double-stranded DNA form with a size of approximately 16.5 kb. Here, we assume that the cell-free mtDNA analyzed by Jiang et al. (Jiang et al., Proc Natl Acad Sci USA. 2015; 112: E1317-E1325) may be degraded linear mtDNA, rather than circular mtDNA.

[0086] Through some embodiments of the present disclosure, we have proposed the analysis of circular mtDNA. For example, the present disclosure can provide a technique for analyzing circular DNA in a biological sample (e.g., including cell-free DNA, such as plasma). For example, in order to measure circular DNA, cutting can be performed to linearize the circular DNA so that it can be sequenced. Example cutting techniques include restriction enzymes and transposases. Then, one or more standards can be used to identify the linearized DNA molecules, for example, to distinguish them from linear DNA molecules. An example standard is to map a pair of reverse end sequences to a reference genome. Another example standard is the identification of a cutting tag associated with a restriction enzyme or an adapter sequence added by a transposase. Once circular DNA molecules (e.g., eccDNA and circular mitochondrial DNA) are identified, they can be analyzed (e.g., to determine counts, size profiles and / or methylation) to measure the characteristics of the biological sample, including genetic characteristics and disease levels.

[0087] I. Circular nuclear DNA

[0088] According to some embodiments, the workflow first reduces (e.g., to substantially eliminate) linear DNA in a plasma DNA sample by exonuclease digestion (e.g., using exonuclease V). Other techniques (e.g., cesium chloride-ethidium bromide (CsCl-EB) density gradient centrifugation) can also be used to reduce linear DNA.

[0089] Then, we follow up with a method for opening a ring (e.g., a ring of eccDNA or mitochondrial DNA) to form a linearized DNA molecule. The linearization of eccDNA can be performed in various ways. In one example, we use restriction enzyme digestion to open the ring at a specific cleavage site with a cleavage sequence motif (which is a type of cleavage tag). In another example, we use a transposase (e.g., via tagging) to open the ring, for example, inserting an identifiable cleavage tag, such as a cleavage sequence motif for restriction enzyme digestion. Library preparation and next generation sequencing of the resulting linearized DNA can then be performed.

[0090] In various examples using enzyme digestion, one embodiment can use the restriction enzyme Mspl (cuts CCGG sequence; methylation insensitive). In another embodiment, we use the restriction enzyme Hpall (cuts CCGG sequence; methylation sensitive). In another embodiment, we combined the data generated by using Mspl and Hpall to obtain new insights into eccDNA.

[0091] Restriction enzymes other than MspI and HpaII can be used. As an illustration, DpnI and DpnII can also be used, both of which recognize GATC sequences. DpnI only cuts when the recognition site (A base) is methylated. On the other hand, DpnII is insensitive to the methylation state. The number of bases recognized and cut can vary. For example, MspI and HpaII are both 4-base cutters. Restriction enzymes other than 4-base cutters, such as 6-base cutters, can be used.

[0092] When compared with rolling circle amplification (Shibata et al., Science.2012; 336: 82-86) and shearing (e.g., by a nebulizer) of eccDNA to form linearized DNA, a method using a cutting tag (e.g., a restriction enzyme or transposase method) can provide a more stringent standard in the definition (identification) of eccDNA reading. For example, two other anchors containing known sequences (cutting tags) can be used to accurately identify eccDNA molecules, where cutting (e.g., CCGG fragment ends) has been performed, and there is no gap between the two end sequences of the sequence reading. This feature anchor (signature anchor) can be used to accurately identify eccDNA reads and to determine their positions in the reference genome. The absence of a gap can be determined using a reference genome via an alignment program, as described in more detail below.

[0093] This information from the cleavage tag (e.g., CCGG read end) not only facilitates more accurate identification of eccDNA, but also the complementary information provided by the number of eccDNA detected from methylation-insensitive and methylation-sensitive restriction enzymes allows the methylation level of eccDNA to be inferred. Such information cannot be obtained by previously recorded methods. In addition, the absence of CCGG fragment ends (or other recognition sequences specific to other types of restriction enzymes, i.e., other types of cleavage tags) in eccDNA fragments can provide insights into pre-existing eccDNA damage, which refers to the linearization of eccDNA before restriction enzyme cleavage. Such linearization may be caused by mechanical shearing during DNA processing, nuclease attack in the bloodstream, etc. Such eccDNA molecules, although detected with a junction site, usually lack restriction enzyme cleavage motifs at one or both ends of the fragment. Such a situation can be referred to as "pre-existing eccDNA damage". Such information cannot be obtained by previously recorded methods. Such information can provide valuable knowledge for the biological mechanisms of eccDNA generation and processing in vivo.

[0094] The use of restriction enzyme digestion has been used to produce recombinant plasmids for molecular cloning. However, there are obvious differences between such applications and the present disclosure. First, when mapped to the genome, eccDNA molecules are produced from the genome of an organism with clear start and end positions, but such a concept does not exist in bacterial plasmids. Secondly, the restriction enzyme method used for eccDNA research can provide insights into the host genome sequence. But for bacterial plasmid DNA, the restriction enzyme digestion method only allows a peek into the plasmid DNA information, but does not allow a peek into the host genome itself (Shintani et al., Front Microbiol. 2015; 31; 6: 242).

[0095] The restriction enzyme method uses the presence of specific recognition sites on eccDNA for its digestion and linearization. The labeling method using transposase to randomly cut DNA does not require a specific DNA sequence. Therefore, the labeling method can potentially provide a higher number of linearized eccDNA for library construction and sequencing. In previous reports, the use of labeling for eccDNA analysis in tissues was described (Shoura et al., G3 (Bethesda). 2017; 7 (10): 3295-3303). Shoura et al. enriched eccDNA from tissue genomic DNA using cesium chloride-ethidium bromide density gradient centrifugation. In contrast, such a step does not need to be performed. Therefore, the labeling method of the present disclosure may be more suitable for plasma DNA and other body fluids or feces including circulating DNA.

[0096] A. Principles and bioinformatics methods used for eccDNA identification

[0097] Figure 1 Shown is an example technique for eccDNA identification according to an embodiment of the present disclosure. The "blue" bar 102 and the "red" bar 106 in the genome 100 represent two regions, which are assumed to be connected together to form extrachromosomal circular DNA (eccDNA). The "cyan" bar represents a restriction enzyme recognition site 104, which serves as a cutting tag. For example, the MspI restriction enzyme can recognize and cut the CCGG site. Such specific cutting will linearize the original circular DNA molecule. The resulting linearized molecule will carry staggered ends, which can be repaired by an end repair step to form a flat-end molecule. Such flat DNA ends will carry cutting tags (i.e., 5'CGG and 3'CGG motifs). Subsequently, different sequencing techniques (including but not limited to Illumina platform, Ion Torrent sequencing, etc.) can be used to sequence the flat-end DNA.

[0098] The eccDNA 110 is shown to have a circular junction locus 112, which includes two regions 102 and 106 from the genome 100. The ends of regions 102 and 106 include nucleotides at two separate genomic locations that are directly adjacent to each other in the eccDNA 110 to form the circular junction locus 112. At step 120, digestion is performed at site 104 to produce a linearized DNA molecule 125. At step 130, end repair is performed, for example, as described above. At step 140, sequencing (e.g., paired end sequencing or single molecule sequencing) is performed to obtain a sequence 145, which includes the circular junction locus 112. As shown, the sequence 145 can include read 1 and read 2.

[0099] If we sequence read 1 and read 2 with sufficient read length, there is a high probability that the sequence read will cross the circular junction locus 112 (indicated by the mosaic arrow) in the step of paired end sequencing. Read 1 extends from the left end of the linearized DNA molecule 125, where read 1 is blue on the left side of the circular junction locus 112 and red on the right side of the circular junction locus 112. Read 2 extends from the right end of the linearized DNA molecule 125, where read 2 is red on the right side of the circular junction locus 112 and blue on the left side of the circular junction locus 112.

[0100] In step 150, an alignment is performed with the reference genome. When read 1 and / or read 2 covers the circular junction locus 112, in the alignment result, we will observe that the read 1 and read 2 sequences of the linearized molecule (e.g., cut by MspI) are mapped to the reference genome with a unique mapping directionality. For illustrative purposes, we define an unmapped segment 152 in read 1 (red arrow after the alignment step, "b->a" segment), which will correspond to sequences from other genomic regions that span the junction and are connected to form circular DNA. Similarly, we define an unmapped segment 154 in read 2 (blue arrow after the alignment step, "e->f" segment), which will correspond to sequences from other genomic regions that span the junction and are connected to form circular DNA molecules.

[0101] The following two cases cover such unique mapping directionality, involving the reverse orientation between the reads and the reference genome:

[0102] a. When the minimum mapping coordinate of read 1 (i.e., b) of segment "b->c" is equal to or less than the minimum mapping coordinate of read 2 (i.e., d) of segment "d->e", read 1 will be aligned in the reverse strand and read 2 will be aligned in the forward strand.

[0103] b. When the minimum mapping coordinate of read 2 is equal to or less than the minimum mapping coordinate of read 1, read 1 will be aligned in the forward strand and read 2 will be aligned in the reverse strand ( Figure 1 not shown).

[0104] Such unique mapping directionality is different from the conventional mapping direction of a pair of paired end reads derived from the original linear DNA.Therefore, such a standard can be used to identify circular molecules.For example, when the minimum mapping coordinate of reading 1 is equal to or less than the minimum mapping coordinate of reading 2, reading 1 is fully aligned in the forward strand, and reading 2 is fully aligned in the reverse strand; or when the minimum mapping coordinate of reading 2 is equal to or less than the minimum mapping coordinate of reading 1, reading 1 is fully aligned in the reverse strand, and reading 2 is fully aligned in the forward strand.In bioinformatics, searching for mapping sites in the reference genome in the unmapped segments present in reading 1 and / or reading 2 will allow the junction to be depicted.The distance between the junction sites inferred from the unmapped segments of the fragment will indicate the size of the circular DNA.For example, the distance between regions 102 and 104 provides the size of the circular DNA.

[0105] Another feature is that if the circular DNA is cut only once, there are two nucleotide overlaps between the mapped read 1 and read 2. This two nucleotide overlapping sequence between read 1 and read 2 is introduced by the staggered ends (i.e., jagged ends) produced by MspI or HpaII or other digestive enzymes. MspI or HpaII will produce two staggered single-strand breaks, and the distance between the two breaks will be 2bp. This 5' protruding 2-nt single-stranded end (complementary to each other) will be filled in the end repair step to form a flat end. Therefore, the resulting DNA sequence will carry a 2bp overlap between the ends of the read 1 and read 2 sequences. In other words, in the library preparation step, there will be an "end repair" step, which will complete the jagged end to a flat end by adding two nucleotides to each end. Therefore, the resulting DNA sequence will have two flat ends instead of two jagged ends. When two sequencing reads are aligned with the genome, the two nucleotides added in the end repair step will appear as two additional base pairs overlapping between the two reads, which can be used additionally or alternatively to identify circular DNA molecules.

[0106] In summary, in an example eccDNA identification method, there may be four "diagnostic features" including:

[0107] a. Circular DNA-specific mapping direction (directionality) as provided in (a) and (b) above;

[0108] b. Junction-identifiable reads (only a portion of the termination sequence maps to the reference genome);

[0109] c. Restriction enzyme cleavage tag;

[0110] d. Two overlapping bases at the 5' end of the read 1 and read 2 sequences.

[0111] Such diagnostic features can greatly improve the specificity of identifying whole genome eccDNA molecules in plasma DNA. In some embodiments, a sequencing read that satisfies at least one of these "diagnostic features" can be defined as a candidate circular DNA. For circular DNA cut multiple times by restriction enzymes, read 1 and read 2 will not have repeated sequences (overlapping bases) between each other. In other embodiments, only one read in a pair may span the junction site, and the other will not carry the junction. As another example, both reads from a pair will not carry the junction, but a unique mapping direction is shown, suggesting circular DNA. In another example, even if a complete restriction enzyme cutting tag cannot be directly observed in the sequencing read, a reference sequence can be retrieved from a reference genome between these inferred junction sites of a circular DNA. Then it can be studied by bioinformatics whether there are any restriction enzyme cutting tags (motifs) in the reference sequence retrieved in this way. The restriction enzyme cutting motif inferred in this way will increase the confidence that the circular DNA species identified are indeed correct.

[0112] Thus, a method can use restriction enzymes as part of analyzing eccDNA. Such techniques can be used in combination with other methods described herein, for example, for analyzing eccDNA as well as mtDNA. Downstream analysis can include measuring characteristics of a sample using detection of circular DNA.

[0113] In step 1, a biological sample of an organism can be received. Examples of biological samples are provided herein, such as plasma and serum. The biological sample includes multiple extrachromosomal circular DNA (eccDNA) molecules. The eccDNA can be from any number of chromosomes, including autosomes and / or sex chromosomes. Each of the multiple eccDNA molecules includes a junction at which the nucleotides at two separate genomic positions are directly adjacent to each other. Junction 112 is an example of such a junction, wherein regions 102 and 106 include two such separate genomic positions, which are directly adjacent to each other.

[0114] In step 2, digestion is performed using restriction enzymes. In some embodiments, more than one type of restriction enzyme can be used. Digestion of multiple eccDNA molecules can form a set of linearized DNA molecules, each of which includes a junction. Each restriction enzyme can cut at a different motif, and the resulting linearized DNA fragments have different cleavage tags. The term "linearized DNA fragment" is different from "linear DNA fragment", which is already linear before any digestion.

[0115] In step 3, for each of the linearized DNA molecules, at least two ends of the linearized DNA molecule can be sequenced to obtain one or more sequence reads. One or more sequence reads may or may not include a junction. If the read does not include a junction, the directionality of the mapping can still be used to identify the eccDNA molecule, as described in Table 1. In some embodiments, two sequence reads (one at each end) can be obtained. In other embodiments, a single sequence read of the entire linearized DNA molecule can include both ends, as described herein.

[0116] After obtaining the sequence reads, the sequence reads can be mapped (aligned) to a reference genome, for example, to see if they map in the reverse direction. If they do map in the reverse direction (an example criterion), the corresponding linearized DNA molecules can be identified as originally circular. Thus, for each of the linearized DNA molecules, a pair of end sequences of the linearized DNA molecule can be selected from one or more sequence reads. The pair of end sequences does not include the junction. Examples of such end sequences are Figure 1 The end sequence 146 and the end sequence 148 in the pair of end sequences are reversed to obtain a pair of reverse end sequences. Examples of such reverse end sequences are reverse end sequence 156 and reverse end sequence 158. The pair of reverse end sequences can then be mapped to a reference genome.

[0117] The reverse end sequence of the mapping can be analyzed to measure the characteristics of the biological sample. The example of such measurement is provided herein. Such analysis can use the set value (e.g., count, size or methylation) of the eccDNA detected. Therefore, the method can also include based on a pair of reverse end sequences mapped to a reference genome, identifying the linearized DNA molecule as being derived from an eccDNA molecule (other standards provided in Table 1 below), and determining the set value of the eccDNA molecule identified, wherein the reverse end sequence of the mapping is analyzed to measure the characteristics of the biological sample using the set value.

[0118] B. Identification Technology

[0119] As explained above, various criteria can be used to identify circular DNA molecules. In addition, the original sequence reads (e.g., from Figure 1 Various procedures are used in the analysis of read 1 and read 2) to identify one or more characteristics of circular DNA.

[0120] The raw sequence reads can be preprocessed. For example, duplicate reads, sequencing adapters, and low-quality bases at the 3' end of the sequencing reads can be removed. In addition, a specified number of bases of paired-end reads (or the ends from a single molecule read) can be selected for alignment.

[0121] 1. Identification of putative eccDNA

[0122] Bioinformatics truncated reads 1 and 2 consisting of the first 50 bp of reads 1 and 2 in preprocessed paired-end reads can be used to align with the human reference genome in paired-end mode using an alignment program, such as Bowtie 2 (Langmead et al., Nat Methods. 2012; 9: 357-9). Other alignment techniques can also be used. In addition to 50 bp, other lengths of each read can also be used, for example, at least 20, 25, 30, 35, 40 or 45 bp. The first pass in the alignment can try a standard direction, for example, aligning read 1 with the left end at a genomic position lower than the last base in the read. For those paired-end reads that are normally aligned (i.e., in the forward direction), the mapping directionality of reads 1 and 2 will be determined in the first pass. Compared to conventional correctly mapped paired ends, if the reads 1 and 2 of the fragment correspond to circular DNA, the forward direction will not provide the correct alignment of the pair, because such reads have a circular DNA-specific mapping direction ( Figure 1 ).

[0123] If the pair of reads does not align in the forward direction, the reverse direction can be tried in a second pass of the alignment. Figure 1 As shown, read 1 and read 2 are in reverse orientation. If the truncated read can be aligned in reverse orientation, the corresponding read before truncation can be realigned with the reference genome. The untruncated reads may be needed so that they cover the junction. If the read does cover the junction, it will not be fully aligned to the reference genome, even in the reverse orientation, e.g. Figure 1 As shown. Paired-end reads with at least one read that cannot be aligned to a reference genome over its entire length can be used for downstream detailed analysis of the "diagnostic features" of eccDNA (e.g., the 4 above), because such reads that cannot be aligned to a reference genome in an end-to-end mode indicate that there are junctions. These paired-end reads can be considered putative reads originating from circular DNA molecules.

[0124] 2. Detecting the junctions of eccDNA molecules

[0125] In order to accurately locate the genomic position of eccDNA with single-base resolution, some embodiments fine-tune the re-alignment of the putative reads respectively. Taking read 1 as an example, the first 20bp and the last 20bp from the read 1 sequence are used as seeds (seed A and seed B, respectively) to determine the candidate genomic regions that may carry the junction. The shortened reads for searching candidate positions help to minimize the possibility of reading containing the junction, which will affect the comparison accuracy and accurate determination of the junction site. In this step, multiple hits (for example, no more than 10 hits for each seed) can be allowed to maximize the sensitivity of the detection junction. If the seed B sequence is not placed downstream of the seed A mapping position in the same direction, it is shown that such a read 1 will carry the junction.

[0126] We then used a search method to probe the junction of read 1 (which was identified as a possible junction carrier) at single-base resolution.

[0127] Figure 2A and 2B A schematic method for a junction search method according to an embodiment of the present disclosure is shown. After alignment with a reference genome, a search is performed within the read, e.g. Figure 1 207 after step 150 in FIG. The read 207 carrying the junction contains two segments (red and blue) with opposite mapping directions, for example, Figure 1 shown.

[0128] exist Figure 2A and 2B In the embodiment of the present invention, the search is performed in a "split and match" manner. We use a "split site" 205 (as shown by a black dotted line) to split the original read 1 sequence into two parts, namely, part A and part B. We repeatedly slide the "split site" 205 along the entire read except for the seed regions 202 and 204 (e.g., 20 bp in length) to exhaust all combinations of part A and part B. The sequence to the left of the "split site" 205 (but not including the seed region 202) is part A. The sequence to the right of the "split site" 205 (but not including the seed region 204) is part B. The minimum length of each of part A and part B can be constrained, for example, not less than 18 bp.

[0129] Figure 2A An example is shown when the "split site" 205 does not overlap with the actual junction 212. After splitting the reads, the seed regions 202 and 204 can be realigned as shown. Then, part A and part B can be joined separately as shown. When the "split site" 205 does not overlap with the actual junction 212, if we compare part A and part B to the reference genome after part A and part B are pasted to seed A and seed B respectively, part A and part B will show many mismatches.

[0130] Figure 2B Shown is an example when the "split site" completely overlaps with the actual junction 212. If we compare part A and part B with the reference genome after pasting part A and part B to seed A and seed B, respectively, part A and part B will theoretically show zero mismatches. Therefore, the "split site" 250 in the read 1 sequence that gives the minimum mismatch in all combinations of part A and part B is identified as a junction. Such a minimum value can meet the mismatch condition. In other embodiments, the seed can be extended until a specified number (e.g., two or more) of consecutive positions do not match the reference.

[0131] Such a search is also applied independently to the read 2 sequence. The read 2 sequence will be used to further improve specificity. For example, the read 2 sequence will have two situations: (1) Like read 1, the read 2 sequence carries a junction. Such connection information should be compatible with the results inferred from the read 1 sequence. (2) The read 2 sequence does not carry a junction. In this case, the read 2 sequence should be fully aligned within the region demarcated by the sequences at either end of the junction site, which is inferred from the read 1 sequence (i.e., part A and part B). The processing order of read 1 and read 2 is interchangeable. In another embodiment, the total number of mismatches along the entire read carrying the inferred junction is required not to exceed a specified number (e.g., 2).

[0132] 3. Classification of eccDNA

[0133] eccDNA can be classified into different groups based on diagnostic features of the eccDNA, including the directionality of each query fragment, the integrity of the cleavage tags (e.g., for restriction enzymes and transposases), and the distance between the 5' end of read 1 and the 5' end of read 2.

[0134] (Table 1).

[0135]

[0136]

[0137] Table 1

[0138] Table 1 shows the classification of eccDNA sequencing reads. Using the restriction enzyme digestion method, we identified eccDNA reads with different characteristics and classified them into 3 types and 3 subtypes. We used four classification criteria, namely: (i) junction (the junction position of the start and end of the original genomic sequence that forms the eccDNA), (ii) directionality of the two sequencing reads, (iii) CCGG tags, and (iv) no gaps between the two sequencing reads. All types meet the directionality. Such typing for labeling can also occur in the case where there is a cutting tag (e.g., a 9-bp repeat sequence or a sequence of other sizes) at the end of the fragment as a marker for single / multiple cutting.

[0139] Type 1 has intact fragments and junctions are detected. The status of intact fragments indicates that the cleavage tag is intact, i.e., the ends are not damaged, so the cleavage tag can be identified.

[0140] Type 2 has incomplete fragments and junctions are detected. The status of incomplete fragments indicates that the cleavage tag is incomplete, i.e., the end is damaged, so the cleavage tag cannot be identified. For type 2-A, gaps may occur, for example, when the DNA is further broken after restriction enzyme cleavage, which will leave a gap between the actual fragment end and the original enzyme cleavage site, resulting in incomplete fragments. The cleavage tag can be inferred as follows. For type 2, since the fragment is incomplete, the 5' end of the read may not carry a complete enzyme cleavage pattern. After identifying the junction sites, we can retrieve a reference sequence from a reference genome between these junction sites of a circular DNA. The reference sequence can then be analyzed to determine whether there are any CCGG motifs in the reference sequence retrieved in this way. If CCGG motifs are present, we refer to them as CCGG.

[0141] Types 2-A and 2-C can be distinguished in the following way. For some eccDNA fragments, the attack on the molecule (physical or chemical) may have occurred more than once, which will not produce one break on the DNA molecule, but two or more breaks. In this case, the small fragment may have fallen out of the original molecule, leaving a gap between the two ends of the original fragment. The eccDNA sequence of the complete molecule can be viewed. We looked at the sequence between the start and end positions of the molecule generated from the genome. If there is a CCGG site within a 100bp radius (or other specified distance) of the end of the fragment, but we do not see it in the sequencing read, we assume that the molecule may be first cut by a restriction enzyme and then further destroyed by other unknown mechanisms, and we classify it as a 2-A read. On the other hand, if there is no CCGG site within a 100bp distance from the end of the fragment, we assume that the eccDNA is unlikely to be linearized by restriction enzyme cutting.

[0142] Type 3 does not detect junctions. Each type has subtypes: (A) opening due to one enzyme cut, (B) opening due to multiple enzyme cuts, and (C) opening due to other mechanisms. The symbol "---" means that both "yes" and "no" are possible. For type 3 reads, we do not distinguish between complete fragments and incomplete fragments. Type 3-A fragments can be similar to type 1-A or type 2-A, except that we cannot detect the junction site. If the paired end 250bp sequencing length is not long enough to reach the junction site, the junction may not be detected. If the molecule is too long and the junction site is somewhere in the middle of the fragment, it may also occur that the junction is not detected. Such an instance will occur when there is a 1000bp eccDNA fragment and the junction site is at the 400bp position. In this case, neither read 1 nor read 2 can reach the junction site. However, there is still reason to believe that this is indeed an eccDNA molecule because the mapping direction of read 1 and read 2 will be different from the mapping direction of the linear molecule.

[0143] Figure 1 An example of type 1-A is shown. For type 1-B, there is a gap between the two ends due to multiple cuts. For type 1-C, no cleavage tag is identified, but the directionality will lead to the inferred read being identified as possibly including a junction and having a common overhang at the two ends. And, any of the above techniques can be used to identify the junction. Analysis of type 2 can be performed in a similar manner to type 1-C, because the cleavage tag cannot be identified, but it can be inferred. For type 3, analysis can be performed in a similar manner to the corresponding type 1, but no junction is detected in this step. In such a case, DNA fragments can be excluded from downstream analysis. Other embodiments can quantify the amount of eccDNA in plasma DNA, which will use knowledge of the number of type C molecules.

[0144] C. Uses of circular DNA

[0145] The circular nature of eccDNA means that they are more resistant to exonuclease clearance in vivo. In addition, exonuclease digestion naturally shortens the length of the digestion target. On the other hand, the circular form of eccDNA molecules exempts them from exonuclease digestion and the resulting shortening. Therefore, eccDNA is expected to have a longer size distribution than linear DNA and therefore contains more genetic information per molecule. Therefore, for the detection of genomic abnormalities and non-invasive diagnosis of genetic diseases, this special type of DNA may be more stable than linear DNA and a more informative biomarker.

[0146] One application of this technology is for non-invasive prenatal testing (NIPT). However, it was previously unknown whether eccDNA of fetal origin is present in maternal plasma. By using the methods described in this disclosure, we have demonstrated for the first time that fetal eccDNA is present in maternal plasma. By using such fetal eccDNA, NIPT can be performed. Such eccDNA-based NIPT has the potential advantage that the amount of fetal genetic information available per molecule of fetal eccDNA may be higher than fetal non-eccDNA in maternal plasma.

[0147] It is known that fetal non-eccDNA in maternal plasma has a very short size distribution, with a modal size of approximately 143 bp (Lo et al., Sci Transl Med 2010; 2: 61ra91). Using the methods described in the present disclosure, we have shown that fetal eccDNA has a longer size distribution than fetal non-eccDNA. This feature is very desirable for NIPT. As an example, fragile X syndrome is a genetic disorder caused by a CGG triplet repeat expansion in the fragile X mental retardation 1 gene region. Such repeats can occur more than 200 times in fragile X patients, while for normal people, this number is less than 40 (Garber et al., Eur J Hum Genet. 2008; 16: 666-72). Because fetal non-eccDNA molecules in plasma are mainly DNA fragments shorter than 200bp (Yu et al., Proc Natl Acad Sci US A. 2014; 111: 8583–8588), it will be difficult to find molecules containing sequences of more than 200 CGG triplet repeats (more than 600bp). However, the larger size of eccDNA will allow such molecules to have a higher chance of containing 600bp long DNA sequences. Therefore, eccDNA represents a valuable resource for NIPT (and indeed other applications outside the pregnancy setting).

[0148] In our current study, we sequenced five pregnancies from the third trimester. Upon successful implementation of our novel approach, we believe that such a design can also be applied to other trimesters (e.g., first and second trimesters).

[0149] Table 2 shows the number of eccDNA molecules detected from different types of readings in a pregnancy case (13007). In Table 2, the number is in CPM, which is a circular DNA read per million mappings. Untreated plasma and plasma treated with MspI and HpaII were measured. Plasma DNA without exonuclease V or restriction enzyme treatment showed that very small amounts of eccDNA were detected (not detected in some types), but plasma DNA from the same case treated with exonuclease V and then digested with MspI or HpaII greatly increased the count of eccDNA molecules detected. The data show that eccDNA can be detected in plasma.

[0150]

[0151] Table 2

[0152] The samples were further analyzed to determine various properties of the eccDNA, such as size, genomic location, and methylation. Such an analysis is provided below.

[0153] 1. Dimensional analysis

[0154] Figures 3A-5B Size analysis of one pregnancy case (MspI treated) according to an embodiment of the present disclosure is shown. The horizontal axis is the size of the fragment and the vertical axis is the frequency percentage at a given size. The frequency is measured within a given type and subtype.

[0155] Plasma DNA from these cases was treated with exonuclease V to eliminate linear DNA. The resulting DNA molecules were digested with MspI or HpaII, and then libraries were constructed and paired end sequencing was performed. These figures show the size analysis of eccDNA using read types 1-A, 1-B, 1-C, 2-A, 2-B and 2-C. For type 3, since no junction was detected, size determination was unavailable. The size profile of eccDNA shows obvious clustering at about 200bp and 340bp. When amplified to 100-500bp range (see illustration), clear 10bp periodicity was also shown.

[0156] Figure 3A The size profile of type 1-A eccDNA is shown. The inset shows the same figure, but is magnified to the 100-500bp range. In addition to the clusters at about 200bp and 340bp, additional peaks can be seen at about 580bp and 750bp. The distance between such peaks can be related to the nucleosome size. Figure 3BThe size profiles of type 1-A and 1-B eccDNA are shown. The inset shows the same figure, but is magnified to the 100-500bp range. For IB, the peak at 340bp is shorter, but other peaks increase, especially at 200bp. The peak of about 200bp may be caused by this length being easily bent into circular DNA.

[0157] Figure 4A Size profiles of type 1-A and 1-C eccDNA are shown. Type 1-C eccDNA is much less abundant, and therefore, those plots show sharp peaks at specific numbers, as opposed to a smoother statistical distribution. Figure 4B The size profiles of type 1-A and 2-A eccDNA are shown. The inset shows the same graph but zoomed in to the 100-500 bp range. For 2-A, the peak at 200 bp is significantly increased.

[0158] Figure 5A Size profiles of type 1-A and 2-B eccDNA are shown. The size profile of 2-B is more dispersed compared to 1-A. The dispersion may be due to sampling variation of a small number of molecules in 2-B. The inset shows the same graph but zoomed in to the 100-500 bp range. Figure 5B The size profiles of type 1-A and 2-C eccDNA are shown. The inset shows the same graph but zoomed in to the 100-500 bp range. For 2-C, the peak at 200 bp is significantly increased.

[0159] Figures 6A-7B Size analysis of maternal and fetal specific eccDNA (5 cases combined) for MspI and HpaII treatment, respectively, is shown. Size data spans types 1 and 2. Genotyping of maternal and fetal DNA was done by Illumina iScan technology, which provides information on maternal and fetal specific SNPs. Fig. 6A and 7A In Figure 2, the frequency plot of maternal and fetal eccDNA showed clusters around 200 bp and 340 bp, with fetal eccDNA being more abundant at the two peaks.

[0160] exist Figure 6B and 7BIn, the cumulative frequency graph shows that fetal specific eccDNA molecules are relatively shorter than maternal specific eccDNA molecules. This phenomenon is observed for both MspI and HpaII processing. As an example, this difference in fragment size can be used to detect sequence imbalances (such as amplification and deletions) in the chromosome region of a fetus or tumor, for example, as described in U.S. Patent Publication Nos. 2011 / 0276277, 2016 / 0217251, 2013 / 0040824, 2016 / 0201142, and 2016 / 0217251 (incorporated by reference in their entirety). For example, due to the dosage effect, the amplification region in the fetus or tumor will increase the release of eccDNA into the plasma, resulting in more eccDNA with shorter size compared to the unaffected region. Due to the dosage effect, the region of the deletion will reduce the release of eccDNA into the plasma, resulting in less eccDNA with shorter size compared to the unaffected region.

[0161] The dosage effect can also be used directly to detect sequence imbalances (e.g., amplifications and deletions) in chromosomal regions of fetuses or tumors, for example, as described in U.S. Patent Publication Nos. 2009 / 0087847, 2009 / 0029377, 2011 / 0105353, 2013 / 0040824, 2016 / 0201142, and 2016 / 0217251 (incorporated by reference in their entirety). For cancer, published data on tumor cells show that more eccDNA is released from genomic regions with oncogenes due to amplification. (Verhaak, RGW, Bafna, V. & Mischel, PS Extrachromosomal oncogene amplification in tumour pathogenesis and evolution. Nat Rev Cancer 19, 283–288 (2019)). Once such cells undergo apoptosis or necrosis, such eccDNA will become cell-free in plasma or other body fluids.

[0162] 2. Genomic location of circular DNA

[0163] For five pregnancy cases, the genomic location of circular DNA was determined by alignment with the reference genome. The genomic location was determined for different types of locations.

[0164] Figure 8Annotations of the genomic locations of eccDNA are shown. The genomic locations of eccDNA from 5 pregnancy cases show similar patterns. Compared with other regions, eccDNA molecules up to 2kb in size are relatively abundant in 3' untranslated regions (UTRs) and CpG islands. EccDNA molecules longer than 2kb in size are relatively abundant in Alu regions.

[0165] Fig. 9 The genomic location of eccDNA from one pregnancy plasma sample treated with MspI and HpaII is shown. From outside to inside: 1) all eccDNA; 2) eccDNA ≤ 2kb; 3) eccDNA > 2kb; 4) maternal-specific eccDNA; 5) fetal-specific eccDNA. Red: > 95% confidence interval; gray: between confidence intervals. eccDNA treated with MspI and HpaII showed a similar distribution pattern. eccDNA was generated from a wide range of genomic regions with specific hotspots on different chromosomes. Fig. 9 It was shown that both MspI and HpaII treatments can detect eccDNA generated from similar regions spanning the genome, providing mutual validation for the two treatments.

[0166] As described in the previous section, a certain amount of eccDNA in a region (e.g., combined with linear DNA) can be used to identify copy number aberrations. Using eccDNA (e.g., in size or count analysis) can improve accuracy, for example, because more DNA is analyzed. For cancer, the amount of aberrant regions can be used to detect cancer, and aberrations can be tracked to monitor cancer over time, as described in 2013 / 0040824.

[0167] 3. Methylation

[0168] The methylation status of linear DNA molecules in human plasma has been used to detect and diagnose various diseases, such as cancer (Chan et al., Proc Natl Acad Sci US A. 2013; 110: 18761–18768; Liu et al., Ann Oncol 2018; 29: 1445-1453). Considering the biological application value of methylation information of linear cell-free DNA, the characterization of eccDNA methylation status in human plasma may provide new possibilities for clinical applications. However, no reports have been published for eccDNA methylation analysis.

[0169] As explained above, restriction enzymes MspI (methylation insensitive) and HpaII (methylation sensitive) can provide methylation information. The complementary information provided by the number of eccDNA detected from methylation-insensitive and methylation-sensitive restriction enzymes allows the methylation level of eccDNA to be inferred. For example, using the number of eccDNA detected in samples treated with MspI and HpaII from the same case, the methylation level of eccDNA can be inferred. The percentage of methylated CpG sites in the CCGG sequence (M) is inferred by the following formula:

[0170]

[0171] Where E M Represents the eccDNA counts in the MspI-treated samples and E H Represents the eccDNA counts in HpaII-treated samples.

[0172] Table 3 shows the methylation levels of eccDNA. The eccDNA counts from plasma DNA samples treated with MspI and HpaII were used to infer the methylation levels of those DNAs. Data from five cases showed that the DNA methylation levels were comparable to those previously documented for linear DNA (Lun et al., Clin Chem. 2013; 59: 1583-94).

[0173]

[0174] Table 3

[0175] As shown in the next section, the eccDNA methylation pattern of a specific tissue (such as a fetus or a tumor) can reflect the methylation pattern in the linear DNA of the same tissue. Therefore, the technology for linear DNA can also be used for eccDNA. For example, the methylation level of linear DNA and eccDNA can be used to detect cancer by comparing one or more methylation levels with reference values, for example, as described in U.S. Patent Publication No. 2014 / 0080715. As another example, the methylation difference of fetal DNA can be used to determine the haplotype of inheritance, for example, as described in U.S. Patent Publication No. 2017 / 0029900. As another example, the methylation pattern can be used to determine the percentage of eccDNA from a specific tissue type (such as a fetus or a tumor), for example, as described in U.S. Patent Publication No. 2016 / 0017419 and 2017 / 0349948. These publications are incorporated by reference in their entirety.

[0176] 4. Fetal fraction

[0177] Chromosomal fetal DNA is often hypomethylated relative to maternal DNA. This behavior persists for cell-free linear DNA. We analyzed the relationship between fetal DNA fractions measured using linear DNA and using eccDNA using methylation-sensitive and -insensitive restriction enzymes. The results suggest that fetal eccDNA is hypomethylated relative to maternal eccDNA.

[0178] Table 4 shows the fetal DNA portion inferred by linear and eccDNA. As previously recorded, when compared with maternal linear DNA in plasma, fetal linear DNA molecules are relatively hypomethylated (Tong et al., Clin Chem. 2007; 53: 1906-14). In our study, 4 of the 5 cases of HpaII-treated samples detected a higher portion of fetal linear DNA, and this difference can be explained by the hypomethylation state of fetal DNA. When we used eccDNA to analyze the fetal part, similar results were obtained. The fetal eccDNA portion detected in the sample treated with HpaII was higher than the fetal eccDNA portion detected in the sample treated with MspI.

[0179]

[0180]

[0181] Table 4

[0182] Thus, fetal eccDNA exhibits hypomethylation relative to maternal eccDNA. This difference between tissue types enables the use of analytical techniques developed for cell-free linear DNA for eccDNA.

[0183] D. Labeling

[0184] As mentioned above, after exonuclease treatment removes linear DNA, transposase can be used to cut circular DNA. Transposase (e.g., Tn5) can be used to mediate the fragmentation of circular DNA, and in a single reaction, the synthetic oligonucleotides at both ends are connected to the linearized fragments. Workflow is achieved by utilizing the fact that transposase has a "cut and paste" property. Such a "cut and paste" function can cut double-stranded DNA and paste the synthesized oligonucleotide into the target sequence. The complex of Tn5 dimer and synthetic aptamer can make the aptamer sequence be connected to the 5' end of the target DNA by the catalytic activity of transposase 'cut and paste' (Adey et al., Genome Biol.2010; 11: R119). The target DNA connected with the adaptor can promote the incorporation of the full sequencing adaptor for a specific sequencing platform, such as but not limited to Illumina synthetic sequencing platform, Pacific Biosciences single molecule real-time (SMRT) system, nanopore sequencing and semiconductor sequencing (such as Ion Proton and the GenapSys Gene Electronic Nano-Integrated Ultra-Sensitive (GENIUS)). It is believed that the Tn5-mediated fragmentation of double-stranded DNA is usually random. Therefore, such Tn5-mediated fragmentation will make it possible to sequence the circular DNA molecule. The mode of transposase catalyzing the incorporation of in vitro DNA fragmentation and adaptor can improve the efficiency of sequencing library construction.

[0185] 1. Overview

[0186] Fig. 10A An example of the principle of using transposase-based tagmentation for eccDNA identification according to an embodiment of the present disclosure is shown. Fig. 10A As shown, we have developed a scheme for analyzing extracellular circular DNA using transposase. The "blue" bar 1001 and the "red" bar 1006 in the genome represent two regions, which are assumed to be connected together to form a section of extrachromosomal circular DNA (eccDNA). The "yellow oval" represents a transposase dimer 1010. The "green" bar 1012 and the "cyan" bar 1014 connected to the transposase dimer represent synthetic oligonucleotides (adapter sequences) that contribute to sequencing. For example, such synthetic oligonucleotides can be hybridized by Illumina sequencing adapters (e.g., P7 and P5 adapters). Tn5 can trigger nonspecific cutting on plasma DNA molecules including linear and circular DNA, and connect synthetic oligonucleotides to the two ends of the cut fragment. Such Tn5 cutting can introduce a 9bp nick near the cleavage site. Before sequencing, such a nick and a single-stranded probe on the adapter can be filled to form double-stranded DNA.

[0187] Then you can Figure 1 Sequencing and alignment were performed as described. Fig. 10A As shown, if only one cleavage occurs, a 9 bp nick may result in a 9 bp overlap. If more than one cleavage occurs, there may be no overlap because the sequences at the two cleavages may be different.

[0188] Because most spontaneously occurring linear DNA molecules in plasma will be approximately 166 bp, we can adjust the concentration of Tn5 and duration of treatment to produce reaction conditions whereby most spontaneously occurring linear plasma DNA will be cleaved only once. Fig. 10B As shown, we will not be able to sequence the cut fragments derived from the linear DNA cut once (using paired end sequencing). In contrast, if the cut fragments are derived from a circular DNA, they will be sequenceable because the two ends of the linearized fragment will be connected to synthetic adapters for sequencing. This strategy allows skipping the step of removing the linear DNA (for example, without the need for exonuclease) or the step of enriching the circular DNA. Using Tn5 or similar enzymes has the advantage of being superior to using a restriction enzyme with a specific recognition site, because the former will make it possible to sequence any circular DNA without the need for a specific sequence motif for cutting.

[0189] Subsequently, the blunt-ended DNA can be sequenced using different sequencing technologies, including but not limited to the Illumina platform, Ion Torrent sequencing, etc. In one embodiment, if we sequence Read 1 and Read 2 with sufficient read length, we will have the opportunity to have the sequence reads span the junction (indicated by the mosaic arrows) in the paired-end sequencing step. Therefore, in the alignment results, we observe Fig. 10A Unique mapping directionality shown for read 1 and read 2 sequences of linearized molecules mapped to the reference genome, as targeted Figure 1 For the purpose of illustration, we defined an unmapped segment in read 1 (red arrow after the alignment step, "b->a" segment), which corresponds to the sequence spanning the junction from other genomic regions that are connected to form circular DNA. Similarly, we defined an unmapped segment in read 2 (blue arrow after the alignment step, "e->f" segment), which corresponds to the sequence spanning the junction from other genomic regions that are connected to form circular DNA molecules. Such unique mapping directionality can include the following two situations:

[0190] (a) When the minimum mapping coordinate of read 1 (i.e., b) of segment “b->c” is equal to or smaller than the minimum mapping coordinate of read 2 (i.e., d) of segment “d->e”, read 1 will be aligned in the reverse strand and read 2 will be aligned in the forward strand.

[0191] (b) When the read 2 minimum mapping coordinate is equal to or less than the read 1 minimum mapping coordinate, read 1 will be aligned in the forward strand and read 2 will be aligned in the reverse strand (not shown in the schematic).

[0192] Such unique mapping directionality is different from the conventional mapping direction of a pair of paired-end reads derived from an initially linear DNA. For example, when the minimum mapping coordinate of read 1 is equal to or less than the minimum mapping coordinate of read 2, read 1 is fully aligned in the forward strand and read 2 is fully aligned in the reverse strand; or when the minimum mapping coordinate of read 2 is equal to or less than the minimum mapping coordinate of read 1, read 1 is fully aligned in the reverse strand and read 2 is fully aligned in the forward strand. Bioinformatically, searching for mapping sites in the reference genome in unmapped segments (e.g., "b->a" and "e->f") present in read 1 and / or read 2 will allow the junction to be delineated. The distance between the junction sites inferred from the unmapped segments of the fragment will indicate the size of the circular DNA.

[0193] Another feature is that if the circular DNA is cut only once, it is possible to observe about 9 nucleotide overlap between the mapped reads 1 and 2. The staggered ends generated by Tn5 introduce such a 9 bp overlap sequence between reads 1 and 2. Tn5 will generate two staggered single-strand breaks, and the distance between the two breaks will be 9 bp ( Fig. 10A Each break will be ligated with read 1 and read 2 adapters respectively. After repair, the 9 bp gap will be filled by DNA polymerase, which will generate duplicate sequences in read 1 and read 2 ( Fig. 10A In summary, there are four “diagnostic features” in this tagging-based eccDNA identification method, including:

[0194] i. Circular DNA-specific mapping direction (directivity);

[0195] ii. Junction-aware reads;

[0196] iii. The cleavage tag corresponds to the adapter sequence;

[0197] iv. When the circular DNA is cut only once, the 5' ends of the read 1 and read 2 sequences have nine overlapping bases.

[0198] In other embodiments, if we use a sequencing platform that can generate long reads, such as Pacific Biosciences SMRT sequencing, nanopore sequencing, etc., we may only see the connection information from one read.

[0199] Sequencing reads that meet at least one diagnostic feature can be defined as candidate circular DNA. For circular DNA that is cut multiple times by Tn5, read 1 and read 2 will not have repeated sequences (overlapping bases) between each other. In some cases, only one read in a pair may span the junction site, while the other will not carry the junction. Even if both reads from a pair do not carry the junction, the unique mapping direction also means circular DNA.

[0200] Using transposase-based tagmentation, bioinformatics methods for detecting eccDNA can be mimicked after their use in restriction enzyme-based methods. For example, we can skip the steps involving analysis of the presence of cleavage motifs introduced by restriction enzymes. Mapping directionality and the presence of junctions in the reads will be the two main "diagnostic features" used to determine whether a particular fragment can be classified as circular DNA.

[0201] 2. Plasma results

[0202] Table 5 shows the detection of eccDNA by the tagging method. CPM corresponds to circular DNA per million mapped reads. Plasma samples from late pregnancy were treated with exonuclease V, followed by tagging and library preparation using the Illumina XTDNA library preparation kit. The sequencing results showed that a large amount of eccDNA was detected using this method. Moreover, the amount of eccDNA detected by this method was much higher than the amount of eccDNA detected by the restriction enzyme treatment method.

[0203]

[0204]

[0205] Table 5

[0206] Fig.11 The size analysis of eccDNA using the tagging scheme according to the embodiments of the present disclosure is shown. The size of plasma eccDNA detected using the tagging scheme shows obvious clustering at about 200bp and 340bp. When zoomed in to the range of 100bp to 500bp, a sharp 10bp periodicity is also shown. Fig.11 The size profiles in are similar to those of techniques using restriction enzymes. Therefore, tagmentation can be used in the same way as restriction enzyme techniques to analyze various properties of eccDNA as well as for diagnostic applications.

[0207] 3. Cell Results

[0208] In addition, Tn5 transposase has been used to develop an assay for transposase-accessible chromatin using sequencing (ATAC-seq) (Buenrostro et al., Nat Methods. 2013; 10: 1213-8). Such a method is based on the direct in vitro transposition of sequencing adapters into natural chromatin. The chromatin compression state will affect the efficiency of inserting sequencing adapters into chromatin. Therefore, the resulting sequencing coverage will reflect chromatin accessibility. Because ATAC-seq involves the use of Tn5 transposase, we predict that sequences derived from circular DNA species (such as eccDNA and mitochondrial DNA) in samples that have undergone ATAC-seq will be sequenced and present in ATAC-seq sequencing data sets. There are many publicly available ATAC-seq data sets in ENCODE (Encyclopedia of DNA Elements, www.encodeproject.org / ), including data from various organizations. We downloaded ATAC-seq FASTQ files of 9 tissues from the ENCODE database, including sigmoid colon, transverse colon, mammary epithelium, gastroesophageal sphincter, stomach, omental fat pad, spleen, subcutaneous adipose tissue, and tibial artery. All samples were analyzed using non-strand-specific ATAC-seq on the Illumina HiSeq 4000 platform. Such experimental data passed the strict quality metrics established by the ENCODE Consortium (median: 92.6 million reads; range: 76.8-103 million reads).

[0209] We used the Tn5-based bioinformatics approach developed in this disclosure to analyze ATAC-seq sequencing results. Table 6 shows the number of eccDNA molecules identified in different tissues. It was found that some organs such as the spleen were relatively rich in eccDNA molecules. These results show that eccDNA in different organs or tissues has a certain degree of variation. We predict that such changes can be used for tissue typing and for detection or monitoring of diseases. Because ATAC-seq can be performed at the single cell level (Chen et al., Nat Commun. 2018; 9: 5345), the analysis of eccDNA can be achieved at the single cell level using the method invented in this disclosure.

[0210]

[0211] Table 6: Identification of eccDNA in different tissues using ATAC-seq datasets. CPM is circular DNA per million mapped reads.

[0212] For tissue typing of one or more cells, the DNA from one or more cells can be analyzed using the above-mentioned techniques to count the number of eccDNA molecules. Based on the count of each cell, one or more tissue types can be identified. For example, a measurement of 2,900 eccDNAs per cell can indicate that the tissue type is spleen. However, a measurement of 787 eccDNAs per cell can indicate that the tissue type is mammary epithelium or transverse colon.

[0213] If the tissue type of one or more cells is known, a significant deviation of the measured count of eccDNA per cell from a normal reference value (e.g., as provided above) can indicate a disease. Such a cutoff value between disease and health can be determined by measuring the statistical distribution of eccDNA measured per cell of healthy cells and the statistical distribution of eccDNA measured per cell of diseased cells, which may have different diseases. Such a detection method can detect whether a cell is healthy based on the measured value of the number of eccDNA molecules in the cell. For example, cancer cells will have more eccDNA molecules.

[0214] 4. Methylation analysis using tagging

[0215] We further describe a novel method that can simultaneously achieve the identification and methylation analysis of eccDNA. As explained above, efficient eccDNA identification can be achieved by combining exonuclease V (exoV) digestion with restriction enzyme or Tn5 transposase treatment. In order to simultaneously enable such a method to lead the identification and methylation analysis of eccDNA, we combined exo V and Tn5, and then enzymatically converted unmethylated cytosine to uracil.

[0216] In this embodiment, plasma DNA is first extracted from human plasma. 50ng DNA is then treated with exo V to eliminate plasma DNA in linear form in the sample in large quantities. The remaining DNA is then incubated with an internally constructed modified Tn5 transposome [all cytosine bases on the adapter sequence are replaced with methylated cytosine (5-mC) to protect the adapter from the conversion of downstream cytosine to uracil]. After 5-mC-Tn5 treatment, the circular DNA molecule is cut open and the adapter is connected to the end of the fragment. Enzymatic conversion and PCR amplification will convert unmethylated cytosine (C) bases into thymine (T) bases, and then Illumina sequencing is performed. Bioinformatics approaches are developed to identify eccDNA molecules and obtain information on their sequence, size profiles, and methylation status.

[0217] Fig.12An example workflow for identifying and methylating eccDNA by enzymatic conversion according to an embodiment of the present disclosure is shown. Plasma DNA is extracted and then incubated with exo V to digest the linear DNA in the sample. The remaining DNA is then treated with 5-mC-Tn5 to open the ring and connect the sequencing adapter (all C is replaced by 5-mC) to the fragment end. Enzymatic conversion and library construction are then performed using the NEBNext Enzymatic Methy-seq kit from New England Biolabs. Bioinformatics approaches for identification, size analysis, and methylation analysis of eccDNA molecules were developed based on sequencing results.

[0218] To demonstrate that this method can effectively capture eccDNA, we also performed parallel experiments on plasma DNA samples treated with 5-mC-Tn5 alone (no exo V treatment). Our results showed that for the exoV+Tn5 sample, the number of eccDNA molecules identified was 40,599 per million mappable reads (EPM) of eccDNA and 12,807 for the Tn5 only sample. These data indicate that the 5-mC-Tn5 transposomes we generated can effectively linearize eccDNA molecules for downstream sequencing analysis, and that exo V treatment can significantly increase the amount of captured eccDNA. In other embodiments, alternative measures for eccDNA linearization, such as restriction enzyme treatment, can also be applied.

[0219] To further demonstrate that this technique can be applied to eccDNA methylation analysis, we compared the methylation levels between small (≤450 bp) and large (>450 bp) eccDNA molecules. Figures 13A-13B As shown in our size analysis data, the majority of the eccDNA molecules we identified were smaller than 450 bp, therefore a cutoff of 450 bp was applied.

[0220] Figures 13A-13B The size analysis and cumulative frequency of eccDNA in human plasma according to the embodiments of the present disclosure are shown. Based on the sequencing data, we identified the eccDNA molecules and obtained their size information. The size of eccDNA peaked at about 202bp and 338bp, with a sharp 10bp periodicity, which is similar to other size profiles herein. Most eccDNA molecules are less than 450bp.

[0221] In the process of calculating eccDNA methylation levels, we applied the following formula to adjust the results according to the C to T conversion rate in the lambda DNA control genome:

[0222]

[0223] Among them, M adj is the adjusted methylation level; M is the methylation level before adjustment and is calculated as: α is the conversion rate of λ DNA spiked-in into the control genome.

[0224] The methylation level of λDNA is an internal control. When processing genomic DNA, we add a small amount of λDNA to the sample. This added DNA will be processed with the same experimental process as the genomic DNA. The reason for using λDNA as an internal control is that the λ genome is theoretically completely unmethylated. If we see that there are residual levels of C that are not converted to T in λDNA, this means that the conversion of λDNA, and genomic DNA for that matter, is incomplete. The incomplete conversion of unmethylated C to T in genomic DNA will lead to an overestimation of methylation levels. By taking into account the conversion rate of λDNA, the above formula can be used to adjust for this overestimation of methylation levels.

[0225] The methylation level of small eccDNA molecules (65.2%) was relatively higher than that of large eccDNA molecules (61.7%). Our data also showed that the residual linear DNA in this sample exhibited a higher methylation level (68.2%) than the eccDNA molecules (65%).

[0226] Fig.14 The methylation levels of linear and eccDNA molecules in different chromosomes were compared. Fig.14 In the study, the methylation level of eccDNA was comparable to that of linear DNA. In addition, by observing the methylation levels in different chromosomes, we can see that most of the time, eccDNA has a lower methylation level than linear DNA.

[0227] We also tested whether bisulfite treatment could also facilitate the identification and methylation analysis of eccDNA. Exo V treatment was combined with bisulfite treatment to identify eccDNA.

[0228] Fig.15An example workflow for identifying and methylating eccDNA by bisulfite conversion according to an embodiment of the present disclosure is shown. Plasma DNA is extracted from a pregnant subject and then incubated with exo V to digest the linear DNA in the sample. The remaining DNA is then treated with bisulfite to convert unmethylated cytosine (C) to uracil while opening the DNA ring. Single-stranded DNA library construction and Illumina sequencing are then performed, in which the uracil converted by unmethylated C will be represented as T, keeping the methylated C base unchanged. Bioinformatics approaches for identification, size analysis, and methylation analysis of eccDNA molecules have been developed.

[0229] Given the harsh nature of bisulfite treatment and the DNA damage it causes, we hypothesized that bisulfite treatment alone might simultaneously achieve conversion of unmethylated cytosine to uracil and open circular DNA molecules. To test this hypothesis, we first investigated whether circular DNA species such as mitochondrial DNA (mtDNA) and plasmid DNA could be linearized by bisulfite treatment. In one experiment, we treated plasma DNA samples with exo V alone or with exo V + bisulfite. In this experiment, the percentage of mtDNA reads in the library was 0.139% for the exo V + bisulfite sample and 0.034% for the exo V alone sample, a 3.1-fold difference. In another experiment, we added the same amount of pBR322 plasmid DNA to the plasma DNA samples before and after bisulfite treatment. When pBR322 was added before bisulfite treatment, the percentage of pBR322 reads in the library was 31.4%; when it was added after bisulfite treatment, this percentage dropped to 0.43%, a 72-fold difference. Thus, bisulfite treatment can linearize circular DNA species. Our data also suggest that such an approach could potentially be developed to simultaneously provide sequence, size, and methylation information for eccDNA molecules.

[0230] These novel methods we developed could potentially be applied to tissue and cell-free eccDNA.

[0231] 5. Methods using transposase

[0232] Thus, a method can use transposase as part of analyzing eccDNA. Such a technique can be used in combination with other methods described herein, for example, for analysis of eccDNA as well as mtDNA. Downstream analysis can include measuring characteristics of a sample using detection of circular DNA.

[0233] In step 1, a biological sample of an organism can be received. Examples of biological samples are provided herein, such as plasma and serum. The biological sample includes multiple extrachromosomal circular DNA (eccDNA) molecules. The eccDNA can be from any number of chromosomes, including autosomes and / or sex chromosomes. Each of the multiple eccDNA molecules includes a junction at which the nucleotides at two separate genomic positions are directly adjacent to each other. Junction 1016 is an example of such a junction, wherein regions 1002 and 1006 include two such separate genomic positions, which are directly adjacent to each other.

[0234] In step 2, cleavage is performed using a transposase. In some embodiments, more than one type of transposase can be used. The transposase can have two adapter sequences attached, for example, Fig. 10A 1012 and 1014.

[0235] In step 3, a transposase is used to ligate an adapter sequence to the two cut ends of each of the plurality of eccDNA molecules, thereby forming a set of linearized DNA molecules, each of which includes a junction and an adapter sequence.

[0236] In step 4, for each of the linearized DNA molecules, at least two ends of the linearized DNA molecule can be sequenced to obtain one or more sequence reads. One or more sequence reads may or may not include a junction. If the read does not include a junction, the directionality of the mapping can still be used to identify the eccDNA molecule, as described in Table 1. In some embodiments, two sequence reads (one at each end) can be obtained. In other embodiments, a single sequence read of the entire linearized DNA molecule can include both ends, as described herein. The sequence read can also include an adapter sequence.

[0237] After obtaining the sequence reads, the sequence reads can be mapped to a reference genome, for example, to see if they map in the reverse direction. If they do map in the reverse direction (an example criterion), the corresponding linearized DNA molecules can be identified as originally circular. Thus, for each of the linearized DNA molecules, a pair of end sequences of the linearized DNA molecule can be selected from one or more sequence reads. The pair of end sequences does not include the junction. Examples of such end sequences are Figure 1 The end sequence 1046 and the end sequence 1048 in the pair of end sequences are reversed to obtain a pair of reverse end sequences. Examples of such reverse end sequences are reverse end sequence 1056 and reverse end sequence 1058. The pair of reverse end sequences can then be mapped to a reference genome.

[0238] The reverse end sequence of the mapping can be analyzed to measure the characteristics of the biological sample. The example of such measurement is provided herein. Such analysis can use the set value (e.g., count, size or methylation) of the eccDNA detected. Therefore, the method can also include based on a pair of reverse end sequences mapped to a reference genome, identifying (detecting) the linearized DNA molecule as being derived from an eccDNA molecule (other standards provided in Table 1 below), and determining the set value of the eccDNA molecule identified, wherein the reverse end sequence of the mapping is analyzed to measure the characteristics of the biological sample using the set value.

[0239] E. Methods for Analyzing eccDNA

[0240] Fig.16 1 is a flow chart illustrating a method 1600 for analyzing circular nuclear DNA according to an embodiment of the present disclosure. Although method 1600 includes physical steps, such steps can be performed using a machine (e.g., a robot) controlled by a computer system. In some embodiments, circular nuclear DNA can be analyzed simultaneously with linear nuclear DNA. Method 1600 can identify circular nuclear DNA according to one or more criteria, and use such circular DNA in determining the characteristics of a sample.

[0241] At block 1610, a biological sample from an organism is received. The biological sample includes a plurality of extrachromosomal circular DNA (eccDNA) molecules. The eccDNA can be from any number of chromosomes, including autosomes and / or sex chromosomes. Each of the plurality of eccDNA molecules includes a junction where nucleotides at two separate genomic locations are directly adjacent to each other, e.g., Figure 1 and 10A If two positions appear at consecutive positions in the sequence read, they are directly adjacent. In any embodiment, the eccDNA can be cell-free, such as in plasma, serum, and similar samples.

[0242] At block 1620, the plurality of eccDNA molecules are cut to form a set of linearized DNA molecules, each of which includes a junction. As an example, a restriction enzyme or a transposase may be used for the cutting. Figure 1 and 10A An example procedure is described in .

[0243] At block 1630, each of the linearized DNA molecules can be sequenced at at least two ends of the linearized DNA molecule to obtain one or more sequence reads. One or more sequence reads may or may not include a junction. If this is done, when paired-end sequencing is performed, only one or two reads may include a junction, while a single read of the entire molecule will include a junction. If the read does not include a junction, the directionality of the mapping can still be used to identify the eccDNA molecule, as described in Table 1. For paired-end sequencing, sequencing of various lengths can be performed at the ends.

[0244] At block 1640, a pair of end sequences of the linearized DNA molecule is selected from one or more sequence reads. The pair of end sequences can be selected from a single sequence read of the entire DNA molecule. In some embodiments, the pair of end sequences does not include a junction. For example, the pair of end sequences can correspond to the sequence described above (e.g., Figure 2A and 2B ). The length of the end sequence can vary, and multiple end sequences can be selected and attempted in subsequent mapping steps. If the initially selected sequence read cannot be mapped, a smaller end sequence can be used.

[0245] At block 1650, the direction of each of the pair of end sequences is reversed to obtain a pair of reverse end sequences. The reversal can be done relative to one strand of the reference genome, for example by selecting a particular strand of the reference genome to map. Figure 1 and 10A Different scenarios describe this reversal, Figure 1 and 10A One situation is shown in . Inversion allows detection of circular DNA, which occurs due to cleavage at locations other than junctions.

[0246] At block 1660 , the pair of reverse end sequences is mapped to a reference genome. Figure 2A and 2B as well as Figure 1 and 10A Examples of such mappings are provided. If the pair of reverse end sequences are indeed successfully mapped, the DNA molecule can be identified as circular. Other criteria can be used to identify circular DNA, for example as shown in Table 1. The presence of a cleavage tag (possibly at both ends of the DNA fragment) can be used as a separate or additional criterion.

[0247] As part of or after mapping, the junction can be detected, for example, for embodiments in which the size of a circular DNA molecule is to be determined. The ends of the original linear fragments can be identified based on the nucleotides at either end of the junction. Analyzing the mapped reverse end sequences can include comparing the bases in one or more sequence reads extending beyond each of the mapped reverse end sequences to the reference genome until a mismatch condition is identified. Figure 2B An example of a mismatch condition is provided. The end positions of the linear DNA fragments that form the eccDNA molecule can be identified based on the position of the mismatch condition in the reference genome. The end positions can be used to determine the size of the linear DNA fragments using the end positions.

[0248] In box 1670, the reverse end sequence of the mapping is analyzed to measure the characteristics of the biological sample. The reverse end sequence of the mapping can be analyzed to measure the characteristics of the biological sample. Examples of such measurements are provided herein. Such analysis can use the set value (e.g., count, size, or methylation) of the eccDNA detected. Therefore, the method can also include identifying the linearized DNA molecule as being derived from the eccDNA molecule (other standards provided in Table 1) based on a pair of reverse end sequences mapped to the reference genome, and determining the set value of the eccDNA molecule identified, wherein the reverse end sequence of the mapping is analyzed to measure the characteristics of the biological sample using the set value.

[0249] Thus, multiple eccDNA molecules can be detected based on a pair of reverse end sequences mapped to a reference genome. A collection value of the detected eccDNA molecules can be determined, wherein the mapped reverse end sequences are analyzed to measure the characteristics of the biological sample using the collection value. Examples of collection values ​​include counts (e.g., aligned to a specific region), sizes, or methylation levels determined using the detected eccDNA molecules.

[0250] Regarding the use of size, a size distribution of sizes measured for multiple eccDNA molecules can be determined and used to measure properties of a biological sample. For example, if the sample is from a pregnant woman or from a subject with cancer, the size distribution will change for regions with copy number aberrations in the region because fetal and tumor fragments are statistically shorter DNA than fragments from maternal / healthy cells. Examples of such size analysis are provided in Section IC1.

[0251] Regarding the use of counting, analyzing the mapped reverse end sequence can include counting the number of multiple eccDNA molecules mapped to the chromosome region, where the characteristic of the biological sample is the characteristic of the chromosome region. The number can then be used to measure the characteristics of the chromosome region. In such an example, the characteristic can be a copy number aberration in the chromosome region, for example, as described in Section 1C above.

[0252] Analyzing the mapped reverse end sequence may include measuring the methylation level in the chromosome region using a DNA molecule in a biological sample. The methylation level can be measured using a linear DNA molecule (i.e., initially linear) and / or a circular DNA molecule (i.e., becoming linearized). In various embodiments, the methylation level may be determined to exhibit high methylation by comparison with a cutoff value. Copy number aberrations and methylation density may be used to detect the condition of an organism. As an example, the condition may be fragile X syndrome or triplet repeat amplification.

[0253] As an example, characteristic is the genotype information of sex or region.Such genotype information can be human leukocyte antigen status or blood type.Therefore, fragment can be analyzed to detect markers (such as Y chromosome, genotype alleles etc.).It can be considered that such example is that the region carries information about biological sample characteristics.As other examples, characteristic is the distortion that comprises sequence change, doubling, expansion, disappearance or amplification in biological sample.

[0254] The biological sample can be obtained from the object being screened for cancer. The level of cancer in the organism can be determined based on the chromosomal region with aberration or at least a specific number (threshold) of aberration regions. As described above, including eccDNA in the assay can improve detection accuracy, for example, because the chromosome can repair itself after releasing eccDNA, so that the chromosome copy does not show aberrations, but the total genetic material will show aberrations.

[0255] Biological samples can be obtained from women with fetuses. Aberrations or sequence imbalances in the fetus can be detected. For example, eccDNA can be used to detect amplified / deleted regions. As another example, genotype information can be detected by identifying variations that occur in linear DNA (nuclear or mitochondrial).

[0256] When a restriction enzyme is used to cleave a specific sequence, it can be used to cleave the tag, e.g. Figure 1 In the example of , a specific sequence is identified at the CCGC that spans a pair of end sequences of at least a portion of the linearized DNA molecule. When a transposase is used, a specific adapter sequence can be used as a cleavage tag.

[0257] The sample may include a first tissue type (e.g., maternal / healthy) and a second tissue type (e.g., fetal / tumor). The first tissue type may be homozygous for the first allele at the locus, and the second tissue type may be heterozygous for the first allele and the second allele at the locus. The first number of reverse end sequences (proxy for circular DNA) with a mapping of the first allele at the locus may be determined. The second number of reverse end sequences with a mapping of the second allele at the locus may be determined. The fractional concentration of eccDNA molecules from the second tissue type may be determined using the first number and the second number. The number of cell-free linear DNA may also be determined, and the pooled concentration of the two may be determined.

[0258] As another example of the analysis in box 1670, sequence variants relative to a reference genome (or relative to the constituent genome of the object, a consensus sequence of healthy cells) can be identified in eccDNA. Multiple sequence variants relative to a reference can be determined in the mapped reverse end sequence. The number of fragments with sequence variants can be used, for example, by comparing the number with a threshold, to determine the level of cancer. Before the contribution of a specific number of sequence variants is added to the total number of eccDNAs with sequence variants, the specific number of sequence variants may be required at any given locus. In addition, eccDNA can be specific to tumor cells (e.g., having variants relative to a reference genome or a constituent genome) or preferentially released from tumor cells (e.g., from an amplification region). After using, for example, exo V to deplete linear DNA, tumor eccDNA will be greatly enriched and improve detection sensitivity.

[0259] II. Simultaneous dissection of circular linear and circular mitochondrial DNA

[0260] Chiu et al. revealed the presence of particle-associated (e.g., identified via a physical filtration step, including intact mitochondria) and free mitochondrial DNA (mtDNA) (e.g., including DNA released during DNA extraction) in plasma using real-time quantitative PCR (qPCR) and filtration with different pore sizes (Chiu et al., Clin Chem. 2003; 49: 719-26). However, the method of Chiu et al. cannot accurately quantify the relative ratio of linear to circular forms of cell-free mtDNA, because qPCR will amplify linear and circular DNA with primer annealing sites. Recently, Newell et al. reported the presence of full-length mtDNA in cell-free human plasma by using two long overlapping amplicons (9,239 bp and 11,216 bp, respectively) (Newell et al., Mol Genet Metab. 2018; doi: 10.1016 / j.ymgme.2018.10.002). Given the prior knowledge of the coexistence of particle-associated and free mitochondrial DNA in plasma, it is not surprising to see that larger amplicons can be amplified (Chiu et al., Clin Chem. 2003; 49: 719-26). However, such PCR assays targeting amplicons of several thousand kilobases cannot amplify short cell-free mtDNA molecules, e.g., <600 bp. Thus, PCR-based assays lack the ability to simultaneously infer the form (circular vs. linear) of mitochondrial DNA in plasma DNA.

[0261] In the present disclosure, some embodiments can provide new methods for simultaneously analyzing short linear and circular mtDNA molecules. The present disclosure allows (1) quantification of the relative amounts between linear and circular forms of cell-free mtDNA molecules in a plasma DNA pool, for example, to determine disease levels; and (2) inference of the tissue origin of linear and circular mtDNA molecules in a plasma DNA pool, for example, as part of determining whether a non-hematopoietic or hematopoietic tissue has a sequence variant.

[0262] A. Technical Principle

[0263] Linear and circular cell-free mtDNA coexist in plasma DNA molecules. In order to enable circular cell-free mtDNA to be sequenced, we linearize the circular form of such molecules. There are a variety of ways to linearize circular cell-free mtDNA molecules, such as but not limited to ultrasonic treatment and enzyme-mediated cutting, such as described above. In one example, enzyme-mediated cutting involves restriction enzymes. In another example, the enzyme is a nuclease (e.g., DNASE1L3 or DNASE1) other than a restriction enzyme. In another example, enzyme-mediated cutting involves a transposase.

[0264] Fig.17An exemplary technique for distinguishing between cell-free circular and linear mtDNA molecules in plasma according to an embodiment of the present disclosure is shown. Plasma DNA molecules include a mixture of short linear DNA and circular DNA. As shown, plasma DNA 1700 includes nuclear DNA 1702, short linear mtDNA 1704, and circular mtDNA 1706.

[0265] In step 1710, enzymatic cleavage is performed. Scissors 1712 (red) represent restriction enzyme cleavage sites. Typically, circular DNA in plasma cannot be sequenced directly, for example using an Illumina sequencing platform. In order to enable circular DNA in plasma to be sequenced effectively, we have created nicks in the circular DNA. On the basis of artificial cleavage of circular DNA molecules (via restriction enzymes, other nucleases, transposases, or physical methods such as ultrasonic treatment), we can linearize those circular DNA molecules to form relatively short linear DNA fragments.

[0266] Therefore, if Fig.17 As shown, enzymatic cleavage is used to distinguish short linear DNA from circular mtDNA. Because most short linear mtDNA is expected to be less than 200bp (Lo et al., Sci Transl Med. 2010; 2: 61ra91; Jiang et al., Proc Natl Acad Sci USA. 2015; 112: E1317-E1325) and circular mtDNA is expected to be approximately 16.5kb, the restriction enzyme selected intentionally will allow most cleavage events to occur preferentially in circular mtDNA. For example, if we choose a 4-bp cutter (e.g., a restriction enzyme that cuts at a specific 4-base sequence), each fragment with <200bp will have less than one such enzymatic cleavage site on average. This is because such a 4-base recognition site is expected to be present every 256bp (where 256 is 4 4 ). The probability of having two cleavage sites due to random chance is only 1 / 65536, which is much lower than the probability of intact circular mtDNA being cleaved. Therefore, if a DNA fragment has two cleavage tags corresponding to the 4-base sequence of the restriction enzyme, the DNA fragment can be identified as originating from circular mtDNA. Specific cleavage tags can be selected to appear at least at a specified interval between the cleavage sites of the reference mitochondrial genome.

[0267] Since we know the sequence of the mitochondrial genome, we can choose a restriction enzyme (e.g., BfaI) that has a sufficient number of cleavage sites in the proposed circular mtDNA molecule. When we choose to use a restriction enzyme, we also consider the distribution of restriction enzyme sites in the mitochondrial genome so that most of the fragments cut have a length suitable for the sequencing platform used. Such sequencing platforms include, but are not limited to, the Illumina synthetic sequencing platform, the Pacific Biosciences single molecule real-time (SMRT) system, nanopore sequencing, and semiconductor sequencing (e.g., Ion Proton and the GenapSysGene Electronic Nano-Integrated Ultra-Sensitive (GENIUS)), etc.

[0268] Therefore, the resulting short linear DNA molecules can then be analyzed by various sequencing platforms. Sequencing 1720 (e.g., massively parallel sequencing) can provide sequence reading for mtDNA molecules 1724 of linear origin and mtDNA molecules 1726 of circular origin. If restriction enzymes are used to cut circular DNA, larger circular DNA 1716 (e.g., a complete circular mtDNA with a size of about 16.5 kb) will have more cutting sites compared to the spontaneously generated short linear DNA 1714. Therefore, by treating plasma DNA with restriction enzymes, we can take advantage of the principle that plasma DNA molecules carrying two cutting ends (represented by black lines with two red ends 1728 corresponding to cutting tags) will have a much higher probability of being artificially derived from circular DNA molecules when compared to those molecules without cutting ends or with one cutting end (represented by black lines with no red ends or with a single red end).

[0269] In some embodiments, after restriction enzyme treatment of circular mtDNA, each of the resulting linear DNA molecules derived from such enzymatic cleavage must have two enzyme-cleaved ends. Such terminal characteristics can allow such artificially produced molecules to be distinguished from linear mtDNA produced spontaneously in plasma. In summary, mtDNA of circular origin is expected to have two cleaved ends, while mtDNA of spontaneously produced linear origin is expected to have no cleaved ends or occasionally have one cleaved end. We call this new technology topological analysis of mitochondrial DNA (T OP M).

[0270] Based on this analysis, the relative amounts of linear and circular DNA forms can be determined simultaneously. In this cleavage-based assay, the presence of "super-long" cell-free linear DNA (>2000 bp) cannot be completely excluded, which may have a considerable probability of generating fragments carrying two cleavage ends. However, based on prior knowledge of cell-free plasma DNA, the probability of having such "super-long" cell-free linear DNA is very low, because the major cell-free DNA molecules in linear form are reported to be <200 bp (Jiang and Lo. Trends Genet 2016; 32: 360-371).

[0271] In some embodiments, restriction enzymes can be selected to meet one or more of the following criteria:

[0272] 1) The restriction enzyme recognition site should be at least 4 bp;

[0273] 2) no star activity (star activity is a relaxation or change in the specificity of restriction enzyme-mediated DNA cleavage, which can occur under reaction conditions that differ significantly from the optimal conditions for the enzyme);

[0274] 3) methylation-insensitive;

[0275] 4) at least 70% of the distance between two cutters in the circular mtDNA should be within a specific range, such as 50 to 600 bp, 100 bp to 600 bp, 150 bp to 600 bp, 200 bp to 600 bp, 300 bp to 1000 bp, 500 bp to 2000 bp, 100 bp to 4000 bp, or other combinations suitable for sequencing, for example;

[0276] 5) The expected sequencing coverage of the mitochondrial genome will be greater than 80%. In other words, after the restriction enzyme digestion and sequencing steps, even with restrictions on the length of the sequence reads, 80% of the circular mitochondrial genome can be covered at least once.

[0277] "Methylation-insensitive" means that the enzyme can cut both methylated and unmethylated DNA. Many restriction enzymes are sensitive to DNA methylation status. When the recognition site is methylated, cutting can be blocked. Therefore, methylation-insensitive enzymes may be preferred.

[0278] Fig.18 The use of the BfaI restriction enzyme to demonstrate multiple principles is illustrated. BfaI can cleave DNA with a CTAG motif within the sequence. The cleavage sites are C^TA G and G AT^C. Fig.18Shown is an example of in silico analysis for selecting restriction enzymes. The x-axis represents the length of the in silico prediction of the DNA molecule after the circular mitochondrial genome is treated with BfaI. The y-axis represents the frequency of specific fragment lengths. The length of the DNA molecule depends on the position of the cleavage tag present in the mitochondrial genome, and the length is the distance between the two cleavage tags. This simulation is for the complete circular mtDNA by in silico cutting of the mtDNA genome at the enzyme recognition site.

[0279] B. Percentage of mtDNA between enzymatically cleaved and untreated plasma DNA

[0280] To demonstrate the ability to detect circular DNA and to show that circular mtDNA is present in appreciable amounts, we analyzed enzymatically treated and untreated samples. We also demonstrated the ability to determine the ratio between circular and linear DNA to represent the abundance of circular mtDNA for further applications such as inference of tissue origin and diagnosis of disease.

[0281] 1. Pregnancy

[0282] For sample collection and plasma DNA preparation, we collected 15 plasma DNA samples from 15 pregnant women. Each plasma DNA sample was divided into two aliquots for experimental determination with and without restriction enzyme digestion. Women with singleton pregnancy were recruited from the Department of Obstetrics and Gynecology, Prince of Wales Hospital, Hong Kong, with written informed consent and approval by the Institutional Ethics Committee. Maternal peripheral blood was collected in tubes containing EDTA, which were subsequently centrifuged at 1,600g for 10 minutes at 4°C. The plasma portion was centrifuged again at 16,000g for 10 minutes at 4°C to obtain cell-free plasma, which was stored at -80°C until further analysis. Plasma DNA was extracted from 4 to 10 mL of plasma using the QIAamp DSP DNA blood micro kit (Qiagen). Plasma DNA was concentrated to a final volume of 75 μL per sample using a SpeedVac concentrator (Savant DNA120; Thermo Scientific). Although the following analysis uses pregnancy samples, the analysis is also applicable to samples from non-pregnant subjects.

[0283] For sequencing and comparison, we use massively parallel paired-end sequencing for two determinations with and without restriction enzyme digestion. Using SOAP2 (Li et al., Bioinformatics.2009; 25: 1966-7), the paired-end reads of sequencing are compared with a reference genome database including all autosomes (chr1, chr2, ... and chr22), sex chromosomes and mitochondrial genomes. Various comparison techniques can be used. Such a compiled reference database including two genomes can reduce the influence of homologous regions existing between the nuclear genome and the mitochondrial genome. For example, some embodiments can only retain the results of unique mappings. Those reads with the same mapping quality aligned with multiple regions or nuclear genomes and mitochondrial genomes can be discarded. In one embodiment, although other numbers of mismatches can be used, for each read, we allow mismatches of up to two nucleotides. As a further example standard, only paired-end reads are used for downstream analysis, and the two ends of the paired-end reads are aligned with the same chromosome in the correct direction, spanning an insert size that does not exceed a specified size (e.g., 600bp). After comparison, the chromosome origin and the length of each sequencing fragment can be determined.

[0284] We sequenced 15 pregnancy samples that were not treated with a restriction enzyme (BfaI). We obtained a median of 26 million (range: 18-40 million) uniquely mapped paired-end reads using massively parallel sequencing (Illumina). The median mapping rate was 82% (range: 67-83%). The median number of sequenced mtDNA fragments was 660 / sample (range: 320-1477).

[0285] For the treated samples, we used a restriction enzyme (BfaI) to digest 15 matched plasma DNA samples before sequencing. Plasma DNA was digested with BfaI (New England Biolabs) under the following conditions: 10 units of BfaI (1 μl), 1X CutSmart buffer (5 μl) and DNA in a 50 μl reaction volume. The reaction was incubated at 37°C for 2 hours and then heat inactivated at 80°C for 20 minutes. The enzymatically cut plasma DNA samples were end-repaired, A-tailed and sequencing adapter-ligated. The adapter-ligated plasma DNA samples were then sequenced using the Illumina platform. Those skilled in the art will understand the various sequencing platforms and preparation techniques that can be used.

[0286] Fig.19A and 19BShown is a comparison of plasma mtDNA metrics of plasma DNA with and without restriction enzyme (BfaI) digestion according to embodiments of the present disclosure. The bars represent measurements of different fractions (treated and untreated) obtained from the same sample.

[0287] Fig.19A The proportion of mtDNA fragments in total plasma DNA molecules is shown. Total plasma DNA molecules include mtDNA molecules and nuclear DNA molecules. In the absence of treatment, all samples provide about the same percentage of mtDNA, which is close to zero. However, the percentage of mtDNA in the treated samples undergoing enzymatic cleavage increases significantly, and different samples have different percentages of mtDNA fragments. This result shows that most of the mtDNA in the plasma samples is circular mtDNA, and enzymatic cleavage enables the detection of such circular mtDNA molecules.

[0288] Fig.19B The ratio of mtDNA fragments with two enzymatically cleaved ends in the total plasma mtDNA molecules is shown. The horizontal axis shows: simulation: in silico cleavage of plasma DNA; untreated: plasma DNA without enzymatic cleavage; enzymatic cleavage: plasma DNA with enzymatic cleavage.

[0289] For the group of plasma DNA samples treated with BfaI, we obtained a median of 24 million (range: 14-30 million) uniquely mapped paired-end fragments. The median mapping rate was 79% (range: 73-81%). The median number of sequenced mtDNA fragments was 9,777 (range: 1,725-37,650). The proportion of mtDNA fragments among total plasma DNA molecules in plasma DNA samples treated with BfaI was, on average, 17-fold higher than that in plasma DNA samples not treated with BfaI ( Fig.11 A). The proportion of mtDNA fragments with two cleavage ends in all mtDNAs was above 80%, which was much higher than that in samples without restriction enzyme treatment (<1%) and the prediction of in silico cleavage using a computer program (<20%). These results indicate that fragments with two cleavage ends are most likely derived from intact circular mtDNA molecules, which cannot be sequenced in plasma DNA samples without restriction enzyme treatment.

[0290] Fig. 20 The size profiles of mtDNA fragments sequenced under different treatments are shown. Untreated 2010 (black) corresponds to plasma DNA not treated with restriction enzymes. Enzymatic cleavage 2020 (red) corresponds to plasma DNA treated with restriction enzymes. Fig. 20It was shown that plasma DNA treated by enzymatic cleavage contained longer mtDNA fragments than plasma DNA not treated with restriction enzymes. Fig. 20 We also provide evidence that fragments with two cleaved ends originate from intact circular mtDNA molecules that cannot be sequenced in plasma DNA samples that have not been treated with restriction enzymes.

[0291] Fig.21 A and 21B show the proportion of mtDNA>200 bp including autosomal DNA and total mtDNA. The bars represent different fractions (treated and untreated) obtained from the same sample. Fig.21 A shows that the proportion of long mtDNA (>200bp) in all DNA in plasma DNA samples with enzymatic cleavage (median: 36.49%; range: 33.17-42.58) is 5.9 times higher than that in plasma samples without restriction enzyme treatment (median: 6.21%; range: 1.76-18.44). This increase is expected based on those cleavage tags being more than 200bp apart. Moreover, the proportion of long mtDNA (>200bp) in all mtDNA also increased, which again shows that DNA fragments from the cleavage of circular mtDNA are detected.

[0292] Fig. 22 Shows the Fig. 20 Enzymatic cleavage in 2020 and Fig.18 The size profile of mtDNA fragments with two cleaved ends from the simulation results in Figure 2. The measured results of enzymatic cleavage 2020 are similar to those of Fig.18 Thus, the size profile of the mtDNA fragments with two cleaved ends correlates well with the size profile of the fragments predicted by in silico cleavage of the circular mitochondrial genome.

[0293] There is a theoretical possibility of "ultra-long" linear mtDNA (e.g. >2000 bp), which, if present, could contribute cleaved molecules to the population of mtDNA fragments with two restriction enzyme-associated cleavage ends. However, the likelihood of such a scenario is relatively low based on the many pieces of evidence listed below:

[0294] 1) The size distribution of plasma mtDNA molecules not treated with restriction enzymes does not have an obvious long tail distribution ( Fig. 20 ).

[0295] 2) In a sequencing experiment using Pacific Biosciences SMRT technology to sequence plasma DNA, we obtained approximately 850,977 plasma DNA sequences. This sequencing technology was chosen for this experiment because it can produce much longer sequence reads than those produced by technologies such as synthetic sequencing technology from Illumina. In the Pacific Biosciences reads, we identified 24 mtDNA fragments. However, none of these were longer than 1000 bp ( Fig.23 ).

[0296] Fig.23 The mtDNA fragments in plasma DNA that was not treated with restriction enzymes sequenced by the Pacific Biosciences SMRT sequencing platform are shown. As described above, 24 mtDNA fragments were identified. The length of each DNA fragment is provided in Fig.23 Since the length of linear mtDNA fragments is limited, long linear mtDNAs do not significantly contribute to the population of mtDNA fragments with two restriction enzyme-associated cleavage ends.

[0297] Fig.24A Shown is a correlation between the proportions of mtDNA in plasma with or without enzymatic cleavage treatment according to an embodiment of the present disclosure. In order to obtain two data points containing a single point on the graph, we divided the sample into at least two aliquots before performing the assay. Different aliquots were treated and not treated to provide different measurements. The vertical axis corresponds to the proportion of mtDNA in all DNA (nuclear and mitochondrial) of the treated sample, and the horizontal axis corresponds to the proportion of mtDNA in all DNA (nuclear and mitochondrial) of the untreated sample. This correlation shows that the amount of linear mtDNA can still be reflected in the new assay with enzymatic digestion.

[0298] Fig. 24B The correlation of the proportion of linear mtDNA in all DNA (nuclear and mitochondrial) of plasma DNA treated with enzymatics to the proportion of all mtDNA in all plasma DNA (nuclear and mitochondrial) without treatment is shown. Although the proportion of the treated samples is much higher, the higher proportion in the treated samples leads to the possibility that the corresponding untreated samples have a higher proportion. These results indicate that the dosage of spontaneously generated linear mtDNA fragments (i.e., fragments that occur naturally or spontaneously and are not generated by artificial enzymatic digestion) in plasma DNA samples treated with restriction enzymes still allows monitoring the relative amount of original mtDNA present in the blood circulation.

[0299] In the above description, we have used BfaI as an example of a restriction enzyme that can be used. Other examples of restriction enzymes include BsaJI, BseDI, BssECI, SecI, Aci, AluI, MaeI and XspI. In other embodiments, a restriction enzyme that is not a 4-base cutter, such as a 6-base cutter, can be used. In other embodiments, a combination of restriction enzymes can be used. In other embodiments, nucleases other than restriction endonucleases can be used, such as DNASE1L3 (Serpas et al., Proc Natl Acad Sci USA 2019; 116: 641-649). In other embodiments, transposases such as Tn5, Mu, Tn7 or Ty1 can also be used. In other embodiments, non-enzymatic methods can be used to open circular mitochondrial DNA, such as physical methods such as ultrasonic treatment, repeated freezing and thawing, repeated heating / cooling cycles, etc.

[0300] 2. Cancer analysis (e.g., HCC)

[0301] We previously reported that an increase in cell-free mtDNA molecules was associated with hepatocellular carcinoma (HCC) patients compared to healthy controls, HBV carriers, and subjects with cirrhosis (Jiang et al., Proc Natl Acad Sci USA. 2015; 112: E1317-E1325). Therefore, we reasoned that TopM would find its utility in the case of patients with hepatocellular carcinoma (HCC).

[0302] We sequenced plasma samples from five HBV carriers and five HCC patients. We prepared three aliquots of plasma DNA for each case for three types of assays: (1) sequencing without restriction enzyme treatment; (2) sonication prior to sequencing (sonication); and (3) restriction enzyme digestion prior to sequencing (enzymatic cleavage). For assays involving sonication, plasma DNA (60 μl) was sheared with a focused ultrasound generator (Covaris S220). The following settings were used: peak incident power 175 W, duty cycle 10%, 200 cycles per burst, and 120 s of treatment time, although other settings may also be used.

[0303] Fig.25A and 25B Shown is a comparison of plasma mtDNA metrics for plasma DNA of HBV and HCC samples with and without restriction enzyme digestion. Fig.25AThe proportion of mtDNA fragments in the total plasma DNA molecules is shown. The total plasma DNA molecules include mtDNA molecules and nuclear DNA molecules. Different columns correspond to different aliquots of the same sample, wherein the left column corresponds to an aliquot of shearing based on ultrasound treatment of plasma DNA before sequencing (ultrasound treatment), the center column corresponds to plasma DNA in the absence of enzymatic cleavage (untreated), and the right column corresponds to plasma DNA treated with restriction enzymes (enzymatic cleavage). Lines connect the data points of the three aliquots.

[0304] Fig.25B The proportion of mtDNA fragments with two enzymatically cleaved ends in total plasma mtDNA molecules is shown. The left column corresponds to the simulation of in silico cleavage of plasma DNA. The other columns are similar to Fig.25A . The simulations were performed by in silico cleavage of mtDNA at the BfaI enzyme recognition site in the plasma sample. This simulation retained linear DNA, thus simulating results where, for example, treatment with an exonuclease would not reduce linear DNA. The enzymatic cleavage data did use such an enzyme treatment, thus providing a baseline of mtDNA carrying two enzymatic cleavage ends in the plasma DNA. In other words, further in silico cleavage of the linear plasma DNA was performed to determine the resulting size profile, illustrating how linear fragments carrying one enzyme site would affect the size profile.

[0305] Compared with the fraction of mtDNA in plasma DNA not treated with restriction enzymes (median: 2.5x10 -3 %; Range: 7.4x10 -4 % to 8.8x10 -3 %), and found that the mtDNA fraction of plasma DNA samples sonicated before sequencing (median: 7.3x10 -3 %; Range: 2.3x10 -3 % to 0.017%) and the mtDNA fraction of plasma DNA samples treated with enzymatic cleavage (median: 9.8x10 -3 %; Range: 7.8x10 -4 % to 0.037%) (p values ​​were 0.02 and 0.03, respectively) ( Fig.25A Such an increase in mtDNA in plasma DNA samples after cleavage of circular mtDNA molecules was observed in both HBV and HCC subjects.

[0306] like Fig.25B As shown, compared with samples without restriction enzyme treatment (median: 0%; range: 0%-0.25% (p value: 4.4×10 -5The proportion of mtDNA fragments with two cleaved ends was greatly increased in enzymatically cleaved plasma DNA samples (median: 75.1%; range: 2.3%-92.0%) compared with the mock case (median: 14.0%; range: 11.6%-19.7% (p-value: 0.0003).

[0307] On the other hand, HBV has a circular genome. In fact, enzymatic or ultrasonic shearing makes HBV fragments more easily detected in HCC patients with HBV infection positive than in the absence of treatment (Table 7). These results show that physical shearing or enzymatic cleavage of plasma DNA will be a universal tool for assessing the contribution of circular DNA present in plasma DNA samples, even for subjects with cancer or viral infection (which may have a circular genome).

[0308]

[0309] Table 7: Number of HBV DNA fragments detected in different library preparations.

[0310] 3. Labeling

[0311] As mentioned above, transposases can also be used instead of restriction enzymes to cleave circular DNA.

[0312] Fig.26A The percentage of mtDNA between samples with and without Tn5 tagmentation is shown. Fig.26A It was shown that the percentage of mtDNA in samples with Tn5 tagmentation was significantly increased (mean: 16.1%; range: 13.5%-18.9%) compared to those without Tn5 tagmentation (mean: 0.0012%; range: 0.001-0.002% (P value = 0.003). This data is consistent with the restriction enzyme data, indicating that tagmentation can be used in a similar manner.

[0313] Fig.26B Shown are the size profiles of mtDNA fragments between samples with and without Tn5 tagmentation. Fig.26B The peak of the size profile of mtDNA molecules in the case of Tn5 treatment is shown to the right of the peak of the size profile of mtDNA molecules in the case of no Tn5 treatment, indicating that there are more long mtDNA molecules in the mtDNA molecules in the case of Tn5 treatment. In summary, the increase in the amount of mtDNA in the sample and the lengthening of mtDNA in the case of Tn5 tagmentation may indicate that Tn5 opens the circular mtDNA and enables those linearized mtDNA molecules to be sequenced.

[0314] C. Targeted capture assay and molecular barcoding strategy for plasma mtDNA analysis

[0315] Although artificial shearing of plasma DNA can increase the proportion of mtDNA fragments in the plasma DNA pool, mtDNA reads are still a minority in the total sequencing reads due to the much smaller size of the mitochondrial genome compared to the nuclear genome. When the fractional concentration of nucleotide variants (including single nucleotide variants, small insertions / deletions, structural variations such as rearrangements) of cell-free mtDNA in the mitochondrial genome is extremely low, detecting such nucleotide variants (i.e., variant analysis) will be suboptimal.

[0316] One implementation to significantly increase the number of mtDNA in sequencing results is to use target capture enrichment. Therefore, we designed a set of hybridization probes targeting the mitochondrial genome.

[0317] Fig. 27 A targeted capture-based TopM analysis according to an embodiment of the present disclosure is shown. Short linear nuclear DNA 2702, short linear mtDNA 2704, and circular mtDNA undergo enzymatic cleavage. Scissors 2701 (red) represent restriction enzyme cleavage sites.

[0318] At step 2710, adapters and unique molecular identifiers (UMIs) can be attached to the DNA molecules. Adapters can be attached to both ends for paired end sequencing. The UMI can be added to only one end and can be a part of the adapter. Thus, the UMI can be a unique sequence of a specified number of bases that can distinguish different molecules from each other.

[0319] In step 2720, target capture enrichment can be performed, for example, using hybridization probes for mtDNA. In some embodiments, after the probe hybridizes with the mtDNA, the mtDNA can be amplified, for example, using a probe or other primers. Amplification is performed using PCR or other suitable amplification procedures. In other embodiments, the probe can be used only to capture mtDNA, thereby increasing the mtDNA concentration in the sample. As shown in the figure, the number of mtDNA molecules increases relative to the amount of nuclear DNA molecules.

[0320] At step 2730, sequencing is performed. Using reads with the same UMI, a consensus sequence can be determined for each template DNA molecule, indicating that the reads originate from the same template molecule. UMIs can be used to remove amplification (e.g., PCR) duplicates, correct errors in late cycles of amplification, and correct sequencing errors. The consensus sequence can provide such corrections.

[0321] like Fig. 27As shown, the additional step of target capture is introduced before sequencing, so that the abundance of mtDNA in the sequencing library will be greatly increased. By amplifying mitochondrial DNA in cell-free DNA, the ratio of mtDNA molecules to nuclear DNA molecules is increased. The number of DNA molecules carrying two cut ends and the number of long DNA molecules (e.g., greater than 200 bases) are also increased.

[0322] Another feature of the capture design is the incorporation of a unique molecular identifier (UMI). Each DNA molecule before capture will be tagged with an adaptor with a UMI. In the sequencing results, sequenced fragments with the same UMI are considered PCR duplicates. All sequenced fragments carrying the same UMI are folded into a single consensus sequence to represent the original molecule present in the plasma DNA pool. This process will allow us to remove PCR bias and minimize sequencing errors as well as PCR errors introduced in late cycles. Therefore, this approach will improve the accuracy of TopM analysis and make it suitable for calling variants with very low mutation scores.

[0323] Thus, embodiments can attach molecular identifiers to linearized mitochondrial DNA molecules and multiple linear mitochondrial DNA molecules. A common sequence for a group of mitochondrial DNA molecules having the same molecular identifier can be determined, and the common sequence can be used as a single sequence read for downstream analysis, for example, as described in the following flow chart.

[0324] Fig.28 It is shown that enzymatic cleavage and target enrichment improve the detection of mtDNA fragments according to embodiments of the present disclosure. Fig.28 As shown, the TopM analysis based on target enrichment indeed improved the detection capacity of mtDNA fragments (i.e., 80-fold more mtDNA than non-target forms).

[0325] D. Example of liver transplantation to determine the level of disease

[0326] Using a sex-mismatched transplantation model, Lo et al. demonstrated the presence of donor-specific DNA in the plasma of kidney and liver transplant recipients (Lo et al., Lancet. 1998; 2: 1329-30). Liver transplantation is an attractive model for studying the biology of tissue-specific cell-free DNA molecules by utilizing genetic features including single nucleotide polymorphism (SNP) differences. We applied TopM analysis based on targeted capture to the plasma DNA of patients receiving liver transplants.

[0327] Fig.29An example technique is shown using a liver transplant model for analyzing linear and circular mtDNA, according to an embodiment of the present disclosure. Donor-specific alleles can be used with certain measurements to confirm the results, e.g., the presence of circular mtDNA and the accuracy of the measurements. However, the techniques for measuring linear and circular mtDNA can be used outside of transplantation applications, such as for other diseases (e.g., cancer) in addition to transplant failure. The results show that subjects with organs (including transplanted organs) with disorders / conditions have reduced amounts of circular mtDNA relative to linear mtDNA due to increased cell death compared to healthy controls.

[0328] For nuclear DNA and mtDNA, "A" and "B" represent two different nucleotide variants. Fig.29 As shown, for plasma nuclear DNA analysis, the liver DNA fraction in plasma can be inferred by using informative SNP sites for which the recipient is homozygous (AA) and the donor is heterozygous (AB), as shown in scenario 2902, rather than scenario 2904. In scenario 2902, the liver DNA fraction can be defined as the proportion of plasma nuclear DNA with the donor-specific allele (B) multiplied by a constant factor of 2.

[0329] For plasma mtDNA analysis, we utilize genomic sites in the mitochondrial genome, wherein acceptor and donor seem to be homogeneous (i.e., seem to be usually the same), but they have different nucleotides at such genomic sites. Sites showing donor-specific and acceptor-specific variants are referred to as mitochondrial information variants, which will be particularly interesting and important. Based on the plasma mtDNA fragments covering these mitochondrial information variants, plasma mtDNA fragments can be divided into donor-specific mtDNA fragments 2914 and acceptor-specific mtDNA fragments 2912. Donor-specific mtDNA fragments 2914 are further classified into mtDNA 2946 from circular sources and 2944 (i.e., spontaneously occurring linear DNA) mtDNA from linear sources, according to whether such fragments carry two enzymatically cut ends. Similarly, acceptor-specific mtDNA 2912 fragments are further classified into mtDNA 2926 from circular sources and mtDNA 2924 from linear sources, according to whether such fragments carry two enzymatically cut ends.

[0330] In this analysis, we attempt to address the following points:

[0331] a) Relative contribution of mtDNA from liver tissue and non-liver tissues (mainly of hematopoietic origin).

[0332] b) Relative amounts of linear (fragments with no cleavage ends or a single cleavage end) and circular (with two cleavage ends) DNA molecules of liver origin and hematopoietic origin.

[0333] c) The use of linear and circular forms of plasma mtDNA mutations can be used as biomarkers for cancer detection.

[0334] In some embodiments of (c), linear mtDNA may be derived primarily from the liver, for example, if the liver has a tumor. Tumor cells disproportionately undergo apoptosis, which can cause circular mtDNA to become linear, thereby enriching linear DNA with tumor-derived DNA molecules. Circular mtDNA may be derived primarily from blood cells (perhaps associated with aging).

[0335] As another example, when the sample is plasma or serum, circular mtDNA mutations can be used to correct background noise. For example, variants identified from circular mtDNA are more likely to be associated with blood lineage cells. Therefore, we can use these variants from circular mtDNA as baseline mutations. Any mutations that do not overlap with those baseline mutations will increase the possibility of originating from other organs (e.g., organs with tumors, such as the liver in this example). As another example, we can use circular mtDNA to filter potential false positives (mutations derived from hematopoietic cells) or focus on mutations that are only present in linear mtDNA and not in circular mtDNA. Such an embodiment of (c) does not require transplanted organs for analysis.

[0336] As an example to test this model, we sequenced plasma DNA samples from two liver transplant recipients using a median of 226 million paired-end reads (range: 130-267 million). In the capture design, we also designed hybridization probes targeting 1000 SNPs on the autosomes to estimate the contribution of donor DNA in the recipient's plasma DNA. We prepared two aliquots of plasma DNA from each liver transplant recipient. One aliquot of plasma DNA was used for target capture enrichment of mtDNA (no restriction enzyme treatment before capture). Another aliquot of plasma DNA was used for enzymatic cleavage treatment, followed by target capture enrichment of mtDNA. DNA samples extracted from archived formalin-fixed paraffin-embedded (FFPE) liver tissue of the donor and leukocytes of the recipient were also enzymatically cleaved and then subjected to target capture enrichment of mtDNA. By sequencing the leukocytes of the donor and recipient, we obtained the genotypes in the nuclear DNA and also identified mitochondrial informative variants. For liver transplant cases TBR1453 and TBR1574, there were 563 and 297 informative SNPs (nuclear DNA), respectively. For liver transplant cases TBR1453 and TBR1574, there were 41 and 29 informative mitochondrial variants, respectively.

[0337] Fig.30 A table 3000 illustrating statistics of liver transplant cases according to an embodiment of the present disclosure is shown. Capture only indicates target capture enrichment of mtDNA. Ez+ capture indicates enzymatic cleavage treatment followed by target capture enrichment of mtDNA.

[0338] For liver transplant cases TBR1453 and TBR1574, the liver DNA fraction (nuclear DNA) was found to be 38.9% and 15.8%, respectively. It is noteworthy that for plasma DNA samples treated with target capture enrichment of mtDNA only (labeled "Capture"), 94.5% and 87.47% of the mtDNA fragments were found to originate from the donor. Since this process does not involve enzymatic cleavage, the results indicate that the liver is the main source of spontaneously linearized mtDNA fragments shed into the plasma. Tumor cells would similarly be expected to be the main source of spontaneously linearized mtDNA fragments shed into the plasma. On the other hand, because circular mtDNA molecules cannot be sequenced by the Illumina sequencing protocol, we do not present data for mitochondrial DNA in circular form in this analysis ("NA" in Table 3000).

[0339] For plasma DNA samples treated with enzymatic cleavage followed by target capture enrichment of mtDNA, the proportion of mtDNA in total plasma DNA molecules was found to be 3.63% for TBR1453 and 4.57% for TBR1574. These figures were higher than those of matched plasma DNA samples that were not treated with enzymatic cleavage (i.e., TBR1453: 1.52% and TBR1574: 0.49%). By enzymatic cleavage of plasma DNA of TBR1453, the recipient's mtDNA contribution increased significantly from a level of 5.5% in matched samples that were not treated with restriction enzymes to as high as 51.03%. These results suggest that a significant proportion of circular mtDNA is of hematopoietic origin. A similar pattern was also observed in liver transplant case TBR1574. The recipient's mtDNA contribution increased from a level of 0.49% in matched samples that were not treated with restriction enzymes to as high as 80.50%. In summary, we conclude that most mtDNA molecules of hematopoietic origin are in circular configuration, while most of the spontaneously occurring linear mtDNA fragments in plasma originate from the liver.

[0340] according to Fig.29 Based on the principle shown, we can further analyze the relative amounts of linear and circular forms of the recipient mtDNA and the donor mtDNA using the presence of enzymatically cleaved ends.

[0341] Fig.31 Analysis of linear and circular mtDNA molecules in the plasma of liver transplant cases according to an embodiment of the present disclosure is shown. For the recipient mtDNA of case TBR1453, the linear and circular mtDNA fractions were 10.0% and 41.03%, respectively. For the donor mtDNA of case TBR1453, the linear and circular mtDNA fractions were 47.54% and 1.43%, respectively. These results indicate that most hematopoietic mtDNA will be circular, while most liver-derived mtDNA will be linear. Similarly, a similar distribution of the amount of linear and circular mtDNA was observed in liver transplant case TBR1574. For the recipient mtDNA of liver transplant case TBR1574, the linear and circular mtDNA fractions were 29.37% and 51.13%, respectively. For the donor mtDNA of liver transplant case TBR1574, the linear and circular mtDNA fractions were 19.0% and 0.5%, respectively.

[0342] Based on these results, we assume that conventional whole genome random sequencing will mainly quantify linear mtDNA, and it is expected that the mtDNA score will be correlated with the liver DNA score. We analyzed plasma DNA samples from 14 liver transplant cases with shallow sequencing depth (median: 21 million paired end reads; range: 16-26 million). The liver contribution in the x-axis was quantified by liver-specific methylation markers via droplet digital PCR (ddPCR) assay (Gai et al., Clin Chem. 2018; 64: 1239-1249).

[0343] Fig.32 The correlation of mtDNA fraction (determined by sequencing without enzyme treatment) with liver DNA fraction (determined by methylation analysis using ddPCR) according to an embodiment of the present disclosure is shown. The mtDNA fraction on the vertical axis is a percentage of all DNA (i.e., linear mtDNA and linear nuclear DNA). The liver DNA fraction on the horizontal axis is measured using liver-specific methylation markers in linear nuclear DNA fragments (Gai et al., Clin Chem. 2018; 64: 1239-1249). In one embodiment, the liver DNA fraction in a liver transplant patient will be determined by donor-specific alleles, for example, as directed against Fig.29 Indeed, the mtDNA fraction correlates with the liver DNA contribution ( Fig.32 , r=0.7, p-value=0.005). Therefore, linear mtDNA abundance can be used to reflect the contribution of liver (or other transplanted or diseased organ) DNA to the plasma DNA pool in a non-invasive manner.

[0344] Since cell death phenomena such as apoptosis and necrosis have been postulated as key mechanisms for the release of plasma DNA, the concentration of linear mtDNA in plasma (relative or absolute concentration) will represent a rapid method for monitoring liver cell death, which will be associated with a variety of conditions, such as cancer (such as liver metastases of hepatocellular carcinoma or cancer originating from other sites), inflammation (such as hepatitis or cirrhosis caused by viruses (such as HBV or hepatitis C virus) or non-viral (such as alcohol-related or fatty liver disease or autoimmune or drug-related (such as acetaminophen) reasons). The higher the amount of linear mtDNA in plasma DNA, the higher the severity of liver cell death and the lower the amount of circular mtDNA. Although the data shown herein are expressed as percentages, the amount of linear and circular mtDNA can be provided in various ways. For example, such percentage values ​​can be converted to absolute concentrations (such as ng / ml) by multiplying by the total concentration of plasma DNA. This analysis of the relative abundance between linear mtDNA and circular mtDNA to determine the level of disease / condition / disease can also be used for such detection in other organs, because such organs undergoing cell death will have similar behavior.

[0345] In some embodiments, the proportion of circular mtDNA will reflect other organ DNA damage. Because the proportion of circular mtDNA is mainly derived from hematopoietic cells, a decrease in the proportion of circular mtDNA will indicate that some other organ DNA is released into plasma DNA.

[0346] Fig.33 The difference in the amount of linear and circular mtDNA between healthy controls and liver transplant patients according to an embodiment of the present disclosure is shown. As shown in Table 3300, the percentage of mtDNA with two cut ends (i.e., circular mtDNA) in liver transplant patients (52%) is shown to be lower than that in healthy controls (73%). In addition, linear mtDNA in liver transplant patients is higher than that in healthy controls. These results indicate increased liver DNA damage, resulting in more liver-derived DNA being released into the blood circulation. Such dynamic tracking of linear and circular mtDNA molecules in plasma DNA allows us to assess the health status of patients, including organ rejection, inflammation, metabolic changes or disorders, immune-related injuries, oncology, and the like.

[0347] E. Linear and circular mtDNA in HCC patients and non-HCC subjects

[0348] We tested the diagnostic potential of the relative abundance of linear and circular mtDNA in distinguishing HCC patients from non-HCC subjects. In these example experiments, we used an enzymatic cleavage followed by a target enrichment protocol, although other techniques, such as non-target enrichment or cleavage using transposases or other cleavage techniques, may also be used. We sequenced five plasma DNA samples from HCC patients and five plasma DNA samples from HBV carriers, with a median of 28.6 million sequenced fragments (range: 14.10-57.10 million). The median depth of the mitochondrial genome was 2,690X (range: 1,018-5,336X), which was on average 5 times higher than the sequencing results without enzymatic cleavage (median: 512X; range: 243-3,022X). Plasma mtDNA molecules without enzymatic cleavage ends are considered linear mtDNA, while plasma DNA molecules with two enzymatic cleavage ends are considered circular mtDNA. Those mtDNAs with one cleaved end are also considered as linear mtDNAs, which are derived from fragments carrying enzyme recognition sites or those that accidentally carry one cleaved end at the DNA end.

[0349] Fig.34 The proportion of linear and circular mtDNA molecules among all mtDNA molecules according to an embodiment of the present disclosure is shown. The vertical axis is the percentage of circular mtDNA molecules (determined as two enzymatically cleaved ends) among all mtDNA molecules. The horizontal axis is the percentage of linear DNA (determined as ends without enzymatic cleavage). The data points correspond to these two values. Different classifications of objects are marked with different colors: healthy controls 3402 are light green, liver transplants 3406 are dark green, HBV carriers 304 are blue, and HCC subjects are red.

[0350] like Fig.34 As shown, it was found that the proportion of linear mtDNA in the plasma DNA of HCC patients (mean: 27.49%; range: 10.64%-36.18%) was significantly higher (p value: 0.03) than the proportion of linear mtDNA in the plasma DNA of non-HCC patients (including healthy controls, liver transplant cases and HBV carriers) (mean 13.53%; range: 3.28-41.52%). In contrast, it was found that the proportion of circular mtDNA in the plasma DNA of HCC patients (mean: 47.99%; range: 28.46%-66.96%) was lower than the proportion of circular mtDNA in the plasma DNA of non-HCC patients (including healthy controls, liver transplant cases and HBV carriers) (mean: 64.33%; range: 36.89-79.02%). Fig.34The combined analysis of circular and linear mtDNA shown allows to distinguish HCC from non-HCC subjects. If we use the following criteria for determining cancer, the sensitivity and specificity are 80% and 92%, respectively:

[0351] a. Circular mtDNA% < 50% (where 50% is an example of a cutoff value);

[0352] b. Linear mtDNA%>21% (where 21% is an example of a cutoff value).

[0353] Figures 35A-35C The differences between HCC and non-HCC subjects are shown for the percentage of mtDNA in all molecules (35A), the percentage of linear mtDNA in all mtDNA molecules (35B), and the percentage of circular mtDNA in all mtDNA molecules (35C). The cleavage technique (e.g., using restriction enzymes or transposases) was used in all three figures. Fig.35A Shown is the distinction between non-HCC and HCC using the difference in the percentage of all mtDNA in all DNA molecules.

[0354] However, the percentage of linear or circular mtDNA in the mtDNA molecules provides a good distinction. Fig.35B The percentage of linear mtDNA (defined as the absence of enzymatically cleaved ends) among all mtDNA (i.e., linear and circular) in non-HCC and HCC subjects is shown. Such values ​​use the measurement of circular DNA, allowing the percentage to be determined among all mtDNA. Fig.35B The percentage of circular mtDNA (defined as ends without enzymatic cleavage) among all mtDNA (ie, linear and circular) in non-HCC and HCC subjects is shown.

[0355] F. Pregnancy

[0356] In addition to the example of only organ disorders, the analytical method for distinguishing between linear and circular mtDNA molecules in plasma can also be applied to pregnancy. To illustrate the concepts involved, we used a pregnancy model involving a special pregnancy. The special pregnancy is a form of assisted reproductive treatment (ART) in which a woman carries a baby in her uterus on behalf of another person. If the leukocytes of the egg donor and the recipient mother are obtained, we can perform genetic analysis for the nuclear DNA and mtDNA of the egg donor and the recipient mother.

[0357] Fig.36An example technique for analyzing linear and circular mtDNA using a pregnancy model according to an embodiment of the present disclosure is shown. Donor-specific alleles can be used with certain measurements to confirm results, e.g., the presence of circular mtDNA and the accuracy of the measurements. However, the techniques for measuring linear and circular mtDNA can be used outside of transplantation applications, such as for fetal disease. For example, a fetus with a disorder / condition will have a reduced amount of circular mtDNA relative to linear mtDNA due to increased cell death compared to a healthy control.

[0358] For nuclear DNA and mtDNA, "A" and "B" represent two different nucleotide variants. Fig.36 As shown, using such genotype information, we can infer the fetal DNA fraction using the nuclear plasma DNA and nuclear DNA genotype information of the egg donor and the recipient mother, as shown in case 3602, but not in case 3604. On the other hand, by utilizing egg donor-specific variants in mtDNA, we can further detect egg donor-related mtDNA molecules 3614 carrying such specific variants, which represent placental / fetal mtDNA, because mtDNA is inherited from the biological mother. Similarly, we can also use those mtDNA fragments covering the recipient mother-specific variants in mtDNA to further detect the recipient mother-specific mtDNA molecules 3612.

[0359] According to whether the two ends of the mtDNA fragment carry restriction enzyme cutting ends, the molecules 3626 and 2646 of circular origin and the mtDNA molecules 3624 and 2644 of linear origin can be identified. TopM will allow us to reveal the form of mtDNA in the plasma DNA of pregnant women and infer its relative proportion, which was not possible before. If the relative proportion of linear and circular mtDNA molecules will be changed by pregnancy-related disorders, such technology can be used to monitor such mitochondrial disorders. We will predict that the spontaneously occurring linear mtDNA (e.g., due to cell death) derived from placental tissue will be added to the population of linear mtDNA (e.g., derived from the liver and other organs) of the pregnant recipient female object. Compared with healthy pregnancies, problematic pregnancies will have a relatively high amount of linear mtDNA (relative to circular mtDNA). For example, the ratio of linear mtDNA to circular mtDNA can be compared with the cutoff value for distinguishing healthy pregnancies from problematic pregnancies.

[0360] In a similar manner, detection of fetal mitochondrial DNA in the plasma of pregnant women can also be used to monitor the success of mitochondrial gene therapy or mitochondrial replacement therapy (Zhang et al., Reprod Biomed Online 2017; 34: 361-368). The increased levels of linear mtDNA relative to circular mtDNA can reflect problems in such therapies. Since the fetal-derived mtDNA in maternal plasma is mainly linear, linear mtDNA can be monitored to reflect the status of fetal mtDNA. This can be done as part of mutation loading, which will be described below.

[0361] In various embodiments, using the ratio of linear mtDNA to circular mtDNA as a surrogate, the ratio of linear mtDNA molecules to circular mtDNA molecules can be used to detect changes in the amount of fetal mitochondrial DNA, monitor the success of the particular pregnancy (e.g., as described above), or monitor the success of mitochondrial gene therapy or mitochondrial replacement therapy.

[0362] G. Variant Analysis

[0363] The mtDNA analysis of the liver transplant model revealed that linear and circular mtDNA molecules can have markers of different tissue origins (e.g., donor-specific alleles), and that linear and circular mtDNA molecules are more likely to come from diseased tissues and healthy tissues, respectively. Therefore, the identification of mtDNA variants of linear mtDNA and circular mtDNA can reveal variants associated with different tissues. For example, variants identified from linear mtDNA will be more likely to be associated with liver cells or other diseased cells (e.g., for detecting and / or monitoring liver cancer, or disorders involving liver cells), whereas variants identified from circular mtDNA will be more likely to be associated with blood lineage cells, which are dominant in plasma or serum (e.g., for assessing clonal hematopoiesis associated with aging (Greaves et al., PLoS Genet. 2014; 10: e1004620). Specific variants (or the amount of variants) in linear mtDNA can be used to identify the presence of diseased tissues.

[0364] Fig.37An example of a mutation load calculation for a linear mtDNA molecule according to an embodiment of the present disclosure is shown. As shown, there are loci 3710 in the mitochondrial genome 3720 at which variants are present in the linear-derived mtDNA 3704 but not in the circular-derived mtDNA 3706. Thus, these variants appear at different rates in the linear-derived mtDNA 3704 compared to the circular-derived mtDNA 3706. In other cases, variants may be present at a higher rate in the circular-derived mtDNA 3706 than in the linear-derived mtDNA 3704.

[0365] The ratio of variants in the linear-derived mtDNA 3704 can reflect the mutation load from diseased tissue (e.g., liver in the above example), because diseased tissues primarily release mtDNA in linear form. In this example, the mutation load is defined by the total number of reads carrying the variant divided by the total number of linear mtDNA analyzed. Fig.37 As shown, by analyzing the loci with lower variant rates in the mtDNA 3706 of circular origin, variants present in linear mtDNA but not present in the mtDNA of circular origin can be identified. In this specific example, the variant rate is zero, that is, it is homozygous (all identical alleles) in the mtDNA 3706 of circular origin. Therefore, the mutation load can be limited to the loci where the mtDNA 3706 of circular origin is homozygous. However, for the mtDNA 3706 of circular origin, the variant rate does not need to be zero.

[0366] Compared to circular mtDNA, the unique variants present in linear mtDNA will reflect the de novo synthesis variants or tissue-specific haplogroups of mtDNA (Grandhi et al., Hum Mol Genet. 2017; 26: 2912-2922; Samuels et al., PLoS Genet. 2013; 9: e1003929). The term "haplotype" can refer to mutations that occur in one mtDNA genome. Different mutations in different haplotypes can occur in the same circular mitochondrial genome or in different circular mitochondrial genomes. The combination of mutations in the circular mitochondrial genome can reflect tissue specificity. Therefore, the embodiment opens up new possibilities for molecular diagnostic applications based on cell-free mtDNA by being able to identify mutations in linear mtDNA fragments from diseased tissues.

[0367] Variants in linear mtDNA are usually of non-hematopoietic origin, because most of the linear mtDNA fragments that occur spontaneously in plasma are from non-hematopoietic tissue. If a variant appears in the mtDNA 3706 of a circular origin, it can be determined that the hematopoietic tissue includes the variant. In either case, the number of loci with variants or the number of sequence reads with variants can be used to determine mutation load. Mutation load can be compared with a threshold value to determine whether there is a disease (such as cancer). In some embodiments, only loci with more than N (such as 1, 2, 3, etc.) variants read are used to determine mutation load, such as to avoid the situation of sequencing errors. Then, whether linear or circular mtDNA has a variant can be used to determine whether the disease has a non-hematopoietic origin or a hematopoietic origin. In addition, tissue-specific variants in biological samples can be used to infer mtDNA sources.

[0368] In some embodiments, TopM analysis will be useful for monitoring immune or autoimmune responses to mitochondrial transplantation (e.g., transplantation of autologous mitochondria to improve myonecrosis (Masuzawa et al., Am J Physiol Heart CircPhysiol. 2013; 304: H966-82)). For example, myonecrosis will release more mtDNA from muscle tissue, leading to an increase in mtDNA in plasma DNA. The specific pattern is that the relative amount of linear DNA will increase and the relative amount of circular DNA will decrease. For variant analysis, muscle-specific mtDNA variants will increase in plasma DNA. Such analysis will be applied to other tissues.

[0369] In the context of cell-free molecular studies, tracking the dynamics of the relative ratio between linear and circular mtDNA molecules and the mutation profiles in linear and circular mtDNA molecules as well as the differential mutation profiles between linear and circular mtDNA molecules will provide new diagnostic avenues for mitochondrial-related diseases (including but not limited to cancer, autoimmunity, myonecrosis, cardioprotection, organ damage and aging).

[0370] H. Methods using circular mtDNA and linear mtDNA

[0371] Measurement of circular mtDNA and linear mtDNA can be used for various purposes. For example, the presence of diseased tissue can be identified from an increase in the amount of linear mtDNA relative to circular mtDNA. As another example, sequence variants that occur primarily in linear mtDNA or circular mtDNA can be used to identify sequence variants in non-hematopoietic tissue or hematopoietic tissue. Various preparation protocols can be used, for example, Fig.17 and Fig. 27 described.

[0372] 1. Determine the level of disease

[0373] Measurement of the relative amounts of circular mtDNA and linear mtDNA can be used for various purposes, such as described above. For example, due to increased cell death, diseased tissue will release more linear mtDNA, resulting in an increase / decrease in parameters including the ratio of linear mtDNA to circular mtDNA relative to subjects without diseased tissue (depending on the formulation of the parameters). In contrast, circular mtDNA is typically produced by blood cells and therefore appears at a relatively constant background rate. Such diseased tissues can include, for example, transplanted organs, cancers, and inflammations, as well as other tissues described herein. In some embodiments, blood cells can be identified as diseased if the amount of circular DNA increases.

[0374] Figure 3800 is a flow chart illustrating a method 3800 for analyzing circular mitochondrial DNA (mtDNA) and linear mtDNA according to an embodiment of the present disclosure. Method 3800 can determine the level of disease based on the analysis. As with other methods, a machine (e.g., a robot) controlled by a computer system can be used to perform the physical steps. In some embodiments, circular mtDNA and linear mtDNA can be analyzed simultaneously. Method 3800 can identify circular nuclear DNA according to one or more criteria and use the relative abundance of circular DNA relative to linear mtDNA when determining the level of disease.

[0375] At block 3810, a biological sample of an organism is received. The biological sample includes cell-free DNA, including linear mitochondrial DNA and circular mitochondrial DNA. The biological sample can be purified, for example, to separate a predominantly cell-free fraction, such as plasma. Other pre-processing steps can also be performed.

[0376] At block 3820, multiple circular mitochondrial DNA molecules are cut to form a set of linearized mitochondrial DNA molecules having predetermined sequences at the ends. The circular mitochondrial DNA molecules can be complete, thus including the entire mitochondrial genome. In other cases, the circular mitochondrial DNA molecules can be only a portion of the mitochondrial genome, thus having junctions, for example, as described in Section I for nuclear DNA.

[0377] As described herein, cutting can be performed in various ways, such as using a combination of a restriction enzyme or a transposase with an aptamer sequence. Restriction enzymes can preferably cut DNA at a specific motif, producing linearized DNA molecules with a predetermined sequence at the end. When using a transposase, after cutting multiple eccDNA molecules, the aptamer sequence can be connected to the two cut ends of each of the multiple eccDNA molecules, thereby forming a group of linearized DNA molecules with a predetermined sequence (aptamer sequence) at the end.

[0378] At block 3830, at least two ends of the set of linearized DNA molecules and the plurality of linear mitochondrial DNA molecules are sequenced to obtain sequence reads. For paired end sequencing, two reads can be obtained: one for each end of the DNA molecule. For single molecule sequencing, the sequence read can be for the entire DNA molecule. The distinction between which molecules are linear or linearized (i.e., initially circular) can be determined by identifying whether a predetermined sequence (e.g., a cleavage tag) is at the end of the DNA molecule.

[0379] Sequence reads corresponding to mtDNA can be identified by alignment with a reference mitochondrial genome, for example, to distinguish from nuclear DNA. This alignment can be performed in stages (e.g., initially identifying those that align with the reference mitochondrial genome, and then determining whether that subset aligns with the reference nuclear genome). As described above, those reads that align with multiple regions or the nuclear genome and mitochondrial genome with the same mapping quality can be discarded.

[0380] At block 3840, a first number of sequence reads having zero or one predetermined sequence at the end is identified. In some embodiments, any of the criteria for having zero or one predetermined sequence at the end can be used. For example, the number of sequence reads having only zero predetermined sequence at the end can be counted to determine the first number. As another example, the number of sequence reads having only one predetermined sequence at the end can be counted to determine the first number. In yet another example, the two numbers can be added to obtain the total number of sequence reads that match either standard. Such sequence reads can be identified as corresponding to linear mtDNA.

[0381] At block 3850, a second number of sequence reads having a predetermined sequence at both ends is identified. Fig. 27 , 29 As described in 36, the linearized DNA molecule will have predetermined sequences at both ends (eg, as a result of restriction enzymes or transposases). Thus, the second number of sequence reads can be identified as corresponding to circular mtDNA.

[0382] At block 3860, a separation value (e.g., a ratio) between the first number and the second number is determined. Various separation values ​​can be used, such as the percentage of mtDNA fragments from linear mtDNA molecules or the percentage from circular mtDNA. As a further example, the separation value can be the first number divided by the second number, or the second number divided by the first number. Thus, the separation ratio can include the ratio of the first number to the second number.

[0383] At block 3870, a disease level associated with the organism is determined based on the ratio. The level can be considered a classification as described herein. The disease level can be a level of cancer, for example, in the case of screening an organism for cancer. The disease level can be a disease level for a specific organ, such as the liver. For example, the level of liver disease is determined as cancer, HBV, or no disease. In another example, the disease level is whether a transplanted organ is rejected.

[0384] In another example, the organism is a woman with a fetus, wherein the disease level is fetal or gestational. If the fetus suffers from a disease, the cell death that increases in the fetus will increase the number of linear mtDNA molecules. An example disease is Kearns Sayre syndrome, which is a rare inborn error of metabolism characterized by progressive external ophthalmoplegia (PEO), retinitis pigmentosa, and onset before the age of 20. Common additional features include deafness, cerebellar ataxia, and heart block. Another example is maternal inherited diabetes with deafness (MIDD). Such a disease is mainly caused by mutations in the mitochondrial genome.

[0385] Determining the disease level may include comparing the ratio to a reference value and determining the disease level based on the comparison. The reference value may be determined based on a group of subjects with known disease levels, e.g., Fig.34 and 35. A reference value (e.g., a cutoff value) can be selected to optimize the specificity and sensitivity of predicting disease levels. Therefore, a training set of samples of all disease-free, disease-free, or a combination of both can be used to determine a reference value. Therefore, a reference value can be determined based on a reference separation value determined from a sample of an object with a known disease level.

[0386] In addition to the separation value between the first number and the second number, other features can also be used to perform block 3870. For example, multiple separation values ​​can be used. As described above, the first number can be defined in various ways, for example, depending on the number of predetermined sequences at the end. A separation value can be determined, where the first number is defined using zero predetermined sequences. A second separation value can be determined, where a third number is defined using zero predetermined sequences, where a separation value between the second number and the first number is determined. This feature can be used to train a machine learning model.

[0387] 2. Detection of mutations in mtDNA from non-hematopoietic tissues

[0388] The measurement of sequence variants in circular mtDNA and linear mtDNA can be used for various purposes, such as described above. For example, sequence variants that mainly appear in linear mtDNA or circular mtDNA can be used to identify sequence variants in non-hematopoietic tissue or hematopoietic tissue, respectively. Such information can be used in various ways. For example, mutations in linear mtDNA in non-hematopoietic tumors may occur more than in healthy cells, and relative to circular mtDNA, such non-hematopoietic tumors will mainly contribute linear mtDNA to plasma. Therefore, if the number of sequence variants in linear mtDNA is higher than a threshold value (e.g., indicating cancer), the tumor can be identified as a non-hematopoietic tissue. And, if the number of sequence variants in circular mtDNA is higher than a threshold value, it can be determined that the tumor is of hematopoietic origin.

[0389] Fig.39 is a flow chart illustrating a method 3900 for analyzing circular mitochondrial DNA according to an embodiment of the present disclosure. The method 3900 can use the techniques described in Section II.G. The method 3900 can determine the tissue origin of sequence variants in mitochondrial DNA.

[0390] At block 3910, a biological sample of an organism is received. The biological sample includes cell-free DNA, which includes linear mitochondrial DNA and circular mitochondrial DNA. Block 3910 may be performed in a manner similar to Fig.38 Box 3810 is performed in a similar manner.

[0391] At block 3920, the plurality of circular mitochondrial DNA molecules are cut to form a set of linearized mitochondrial DNA molecules having predetermined sequences at the ends. Block 3920 may be performed in the same manner as Fig.38 Block 3820 is performed in a similar manner.

[0392] At block 3930, at least two ends of the set of linearized DNA molecules and the plurality of linear mitochondrial DNA molecules are sequenced to obtain sequence reads. Block 3930 may be performed in the same manner as Fig.38 Block 3830 is performed in a similar manner.

[0393] At block 3940, a first set of sequence reads having zero or one predetermined sequence at the end is identified. Block 3940 may be performed with Fig.38 Block 3840 is performed in a similar manner.

[0394] At block 3950, a second set of sequence reads having predetermined sequences at both ends are identified. Block 3950 may be performed with Fig.38 Block 3850 is performed in a similar manner.

[0395] At block 3960, the first set of sequence reads and the second set of sequence reads are compared to a reference mitochondrial genome to identify sequence variants that occur in one group at a different ratio than another group. In some embodiments, the first set of sequence reads can be compared to the reference mitochondrial genome to identify the first set of sequence variants at the first set of loci. The second set of sequence reads can be compared to the reference mitochondrial genome to determine whether the first set of sequence variants appear in the second set of sequence reads. In some embodiments, the reference mitochondrial genome can correspond to the object, for example, as determined from a healthy cell. Such a reference mitochondrial genome can be referred to as a constituent mitochondrial genome.

[0396] At block 3970, it is determined whether the non-hematopoietic tissue or the hematopoietic tissue has a sequence variant based on the different ratios. For example, the different ratios can be a percentage between a first ratio (amount) of a first set of sequence reads (which may be from a non-hematopoietic tissue) and a second ratio (amount) of a second set of sequence reads. Variants in non-hematopoietic tissue can occur in transplanted tissue, for example, in the liver example described above, or in an egg donor, such that the non-hematopoietic tissue can be fetal.

[0397] In some embodiments, a determination can be made whether a non-hematopoietic tissue has a sequence variant based on a first amount of a first set of sequence reads (i.e., for linear mtDNA) having a sequence variant and a second amount of a second set of sequence reads (i.e., for circular mtDNA) having a sequence variant, e.g., the first amount is greater than the second amount. The second amount can be zero, indicating that the circular mitochondrial DNA is homozygous in the biological sample. The loci in which the circular mitochondrial DNA is homozygous can be a criterion for identifying loci that may have variants in linear mitochondrial DNA.

[0398] In order to determine that non-hematopoietic tissue has sequence variants, it may be necessary that the first amount is greater than the first threshold value and the second amount is less than the second threshold value (for example, if homozygosity is required, one is required). The first threshold value may be the same or different, and may be the percentage of sequence reading. In various embodiments, it is necessary to cover at least but not limited to 5, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 200, 300, 400, 500, 1000, 2000, 3000, 4000, 5000, 10000 sequence reads of the analyzed genomic sites. The percentage of sequence reads carrying the variant allele in the linear mtDNA molecule can be, for example, but not limited to, greater than 1%, 2%, 3%, 4%, 5%, 10%, 20%, 30%, 40%, and 50%, while the percentage of sequence reads carrying the variant allele in the circular mtDNA molecule can be, for example, but not limited to, less than 1%, 2%, 3%, 4%, 5%, 10%, 20%, 30%, 40%, and 50%. A threshold value can be used to determine that a variant is more prevalent in the linear mtDNA than in the circular mtDNA, for example, as opposed to just a first amount being higher than a second amount. Variants identified from the linear mtDNA will be more likely to be associated with liver cells (e.g., for detecting and / or monitoring liver cancer or a condition involving liver cells), while variants identified from the circular mtDNA will be more likely to be associated with blood lineage cells.

[0399] In some embodiments, the amount of mutation can be compared to a threshold to detect a disease (e.g., cancer). The amount of mutation may need to be in linear mtDNA and / or circular mtDNA. For example, a non-hematopoietic tissue may be determined to have a disease (e.g., cancer) based on a first amount being greater than a threshold (cutoff value). The threshold may be determined based on measurements in a healthy subject group and / or based on measurements in a subject group with a disease.

[0400] In other embodiments, it is possible to determine that hematopoietic tissue has sequence variants based on the second amount being greater than the first amount. A similar threshold value can be used to require that the second amount is fully greater than the first amount, for example, in order to obtain the desired statistical accuracy. In some embodiments, it is possible to determine that hematopoietic tissue is diseased (for example, hematopoietic cancer) based on the second amount being greater than a threshold value. The threshold value can be determined based on measurements in a healthy subject group and / or based on measurements in a subject group suffering from a disease.

[0401] I. Disease Level and Treatment

[0402] Some embodiments may also include treating the disease / condition of the subject in response to the classification of the subject's condition, thereby improving the condition (e.g., to remove the condition or reduce the severity). Treatment may be provided based on the determined level of the disease / condition, the identified variant, and / or tissue source (e.g., hematopoietic or non-hematopoietic). For example, the identified variant may be targeted with a specific drug or chemotherapy. The tissue source may be used to guide surgery or any other form of treatment. Also, the level of the condition may be used to determine how aggressive any type of treatment is.

[0403] Various treatments can be performed. Treatment can include any suitable therapy, including drugs, chemotherapy, radiation, immunotherapy, hormone therapy, stem cell transplantation or surgery, including any treatment described in the references mentioned herein. Information about treatment in the references is incorporated herein by reference. Treatment can be targeted, for example, using precise drugs customized for specific characteristics of the disease (e.g., the specific genetic composition of the tumor). Based on the determined level of the condition, a treatment plan can be formulated to reduce the risk of damage to the object. The method can also include treating the subject according to the treatment plan.

[0404] Biological samples can be obtained at different time points and analyzed independently at those time points, or combined with measurements and classifications at other time points. Examples of such time points include before and after cancer treatment (e.g., targeted therapy, immunotherapy, chemotherapy, surgery), different time points after cancer diagnosis, before and after cancer progression, before and after metastasis development, before and after increased disease severity, or before and after complication development.

[0405] As described above, machine learning models can be used to determine disease levels. Example models can include, but are not limited to, linear regression, logistic regression, neural networks such as deep recurrent neural networks, Bayesian classifiers, hidden Markov models (HMMs), linear discriminant analysis (LDA), k-means clustering, density-based spatial clustering of applications with noise (DBSCAN), decision trees (e.g., random forests), and support vector machines (SVMs).

[0406] The model may include a supervised learning model. The supervised learning model may include different methods and algorithms, including analytical learning, artificial neural networks, back propagation, boosting (meta-algorithms), Bayesian statistics, case-based reasoning, decision tree learning, inductive logic programming, Gaussian process regression, genetic programming, grouping methods for data processing, kernel estimators, automatic learning machines, learning classifier systems, minimum message length (decision trees, decision diagrams, etc.), multilinear subspace learning, naive Bayes classifiers, maximum entropy classifiers, conditional random fields, nearest neighbor algorithms, probabilistic approximately correct learning (PAC) learning, ripple descent rules, knowledge acquisition methods, symbolic machine learning algorithms, sub-symbolic machine learning algorithms, support vector machines, minimum complexity machines (MCM), random forests, ensembles of classifiers, ordered classification, data preprocessing, handling unbalanced data sets, statistical relational learning or Proaftn (a multi-criteria classification algorithm).

[0407] III. Example System

[0408] Fig.40 A measurement system 4000 according to an embodiment of the present disclosure is described. The system shown includes a sample 4005, such as a cell-free DNA molecule in a sample holder 4010, wherein the sample 4005 can be contacted with an analyzer 4008 to provide a signal of a physical feature 4015. An example of a sample holder can be a flow cell, which includes a probe and / or primer of the analyzer or a tube through which a droplet moves (the droplet is included in the analyzer). The physical feature 4015 (such as fluorescence intensity, voltage or current) from the sample is detected by a detector 4020. The detector 4020 can be measured at intervals (e.g., periodic intervals) to obtain data points that constitute a data signal. In one embodiment, an analog-to-digital converter converts the analog signal from the detector into a digital form many times. The sample holder 4010 and the detector 4220 can form an assay device, such as a sequencing device for sequencing according to the embodiments described herein. The data signal 4025 is sent from the detector 4020 to a logic system 4030. The data signal 4025 may be stored in a local memory 4035 , an external memory 4040 , or a storage device 4045 .

[0409] The logic system 4030 may be or may include a computer system, an ASIC, a microprocessor, etc. It may also include a display (e.g., a monitor, an LED display, etc.) and a user input device (e.g., a mouse, a keyboard, a button, etc.) or be coupled thereto. The logic system 4030 and other components may be part of a computer system that is independent or network-connected, or it may be directly connected or integrated in a device (e.g., a sequencing device) including a detector 4020 and / or a sample holder 4010. The logic system 4030 may also include software executed in the processor 4050. The logic system 4030 may include a computer-readable medium storing instructions for controlling the system 4000 to perform any method described herein. For example, the logic system 4030 may provide a command to a system including the sample holder 4010 to perform sequencing or other physical operations. Such physical operations may be performed in a specific order, for example, adding and removing reagents in a specific order. Such physical operations may be performed by a robotic system, for example, including a robotic arm, which may be used to obtain a sample and perform analysis.

[0410] Any computer system mentioned herein may utilize any suitable number of subsystems. Examples of such subsystems are shown in FIG. Fig.41 In some embodiments, the computer system includes a single computer device, wherein the subsystem can be a component of the computer device. In other embodiments, the computer system can include multiple computer devices with internal components, each of which is a subsystem. The computer system can include desktop computers and laptop computers, tablet computers, mobile phones, and other mobile devices.

[0411] Fig.41 The subsystems shown in FIG. 7 are interconnected via a system bus 75. Additional subsystems are shown, such as a printer 74, a keyboard 78, a storage device 79, a monitor 76 (e.g., a display screen such as an LED) coupled to a display adapter 82, etc. Peripheral devices and input / output (I / O) devices coupled to the I / O controller 71 may be connected via any number of means known in the art, such as an input / output (I / O) port 77 (e.g., USB, ) is connected to the computer system. For example, I / O port 77 or external interface 81 (e.g., Ethernet, Wi-Fi, etc.) can be used to connect the computer system 10 to a wide area network, such as the Internet, a mouse input device, or a scanner. The interconnection via the system bus 75 allows the central processor 73 to communicate with each subsystem and control the execution of multiple instructions from the system memory 72 or storage device 79 (e.g., a fixed disk, such as a hard drive or optical disk), as well as the exchange of information between subsystems. The system memory 72 and / or storage device 79 can be embodied as a computer readable medium. Another subsystem is a data collection device 85, such as a camera, microphone, accelerometer, etc. Any data mentioned herein can be output from one component to another and can be output to a user.

[0412] The computer system may include multiple identical components or subsystems connected together, for example, via an external interface 81, via an internal interface, or via a removable storage device (which may be connected from one component to another and removed from it). In some embodiments, the computer system, subsystem, or device may communicate via a network. In such cases, one computer may be considered a client and another computer may be considered a server, each of which may be part of the same computer system. The client and server may each include multiple systems, subsystems, or components.

[0413] Aspects of the embodiments can be implemented in the form of control logic (e.g., ASIC or field programmable gate array) using hardware circuits, and / or implemented in a modular or integrated manner using computer software with the aid of a general-purpose programmable processor. As used herein, the processor may include a single-core processor, a multi-core processor on the same integrated chip, or multiple processing units and dedicated hardware on a single circuit board or networked. Based on the disclosure and teachings provided herein, those skilled in the art will know and understand other ways and / or methods of implementing the embodiments of the present disclosure using hardware and a combination of hardware and software.

[0414] Any software components or functions described in this application can be implemented as software code executed by a processor using any suitable computer language (e.g., Java, C, C++, C#, Objective-C, Swift) or scripting language (e.g., Perl or Python) using, for example, traditional or object-oriented techniques. The software code can be stored as a series of instructions or commands for storage and / or transmission on a computer-readable medium. Suitable non-transitory computer-readable media may include random access memory (RAM), read-only memory (ROM), magnetic media such as hard drives or floppy disks, or optical media such as compact disks (CDs) or DVDs (digital versatile disks) or Blu-ray disks, flash memory, etc. The computer-readable medium can be any combination of such storage or transmission devices.

[0415] Such programs can also be encoded and transmitted using carrier signals suitable for transmission by wired, optical and / or wireless networks (including the Internet) that meet multiple protocols. Therefore, computer-readable media can be generated using data signals encoded with such programs. Computer-readable media encoded with program code can be packaged with compatible devices or provided separately from other devices (for example, downloaded via the Internet). Any such computer-readable media can reside on or within a single computer product (for example, a hard drive, a CD or an entire computer system), and can be present on or within different computer products in a system or network. A computer system can include a monitor, a printer, or other suitable displays for providing any result mentioned herein to a user.

[0416] Any method described herein can be performed in whole or in part with a computer system, and the computer system includes one or more processors, and the processor can be configured to perform the steps. Therefore, the embodiment can be related to a computer system configured to perform the steps of any method described herein, and the computer system may have different components for performing each step or each step group. Although presented in numbered steps, the steps of the method herein can be performed simultaneously or at different times or in different orders. In addition, a part of these steps can be used together with a part of other steps of other methods. And all or a part of the steps can be optional. In addition, any step in any method can be performed by means of a module, unit, circuit or other method for performing these steps.

[0417] Without departing from the spirit and scope of the embodiments of the present disclosure, the specific details of a particular embodiment may be combined in any suitable manner. However, other embodiments of the present disclosure may be directed to specific embodiments related to each individual aspect or specific combinations of these individual aspects.

[0418] The above description of example embodiments of the present disclosure has been presented for purposes of illustration and description. It is not intended to be exhaustive or to limit the disclosure to the precise form described, and many modifications and variations are possible in light of the above teaching.

[0419] Unless specifically indicated to the contrary, the use of "a", "an", or "the" is intended to mean "one or more". Unless specifically indicated to the contrary, the use of "or" is intended to mean "logical or", not "mutually exclusive or". Reference to a "first" component does not necessarily require the provision of a second component. Furthermore, unless explicitly stated, reference to a "first" or "second" component does not limit the referenced components to specific locations. The term "based on" is intended to mean "based at least in part on".

[0420] All patents, patent applications, publications, and descriptions mentioned herein are incorporated by reference for all purposes. No one is admitted to be prior art.

Claims

1. A method comprising: receiving a biological sample from an organism, the biological sample comprising a plurality of extrachromosomal circular DNA (eccDNA) molecules, wherein each of the plurality of eccDNA molecules comprises a junction at which nucleotides at two separate genomic locations are directly adjacent to each other; cleaving the plurality of eccDNA molecules using the cleavage tag to form a set of linearized DNA molecules, each of which includes the junction; For each of the set of linearized DNA molecules: sequencing at least two ends of the linearized DNA molecule to obtain one or more sequence reads; selecting a pair of end sequences of the linearized DNA molecule from the one or more sequence reads, the pair of end sequences excluding the junction; reversing the direction of each of the pair of end sequences to obtain a pair of reverse end sequences; as well as mapping the pair of reverse end sequences to a reference genome; as well as The mapped reverse end sequences are analyzed to measure the properties of the biological sample.

2. The method of claim 1, further comprising: detecting the plurality of eccDNA molecules based on a pair of reverse end sequences mapped to the reference genome; as well as An aggregate value of the detected eccDNA molecules is determined, wherein the mapped reverse end sequences are analyzed to measure a property of the biological sample using the aggregate value.

3. The method of claim 2, wherein the aggregate value comprises a count, size, or methylation level determined using the detected eccDNA molecules.

4. The method of claim 1, wherein: For each of the linearized DNA molecules, the one or more sequence reads include the junction.

5. The method of claim 4, wherein analyzing the mapped reverse end sequences comprises: comparing bases in the one or more sequence reads extending beyond each of the mapped reverse end sequences to a reference genome until a mismatch condition is identified; as well as Based on the positions of mismatch conditions in the reference genome, the end positions of the linearized DNA molecules that form the eccDNA molecules are identified.

6. The method of claim 5, wherein analyzing the mapped reverse end sequences further comprises: The end positions are used to determine the size of the linearized DNA molecule.

7. The method of claim 6, wherein analyzing the mapped reverse end sequences further comprises: determining a size distribution of sizes measured for the plurality of eccDNA molecules; as well as The size distribution is used to measure a property of the biological sample.

8. The method of claim 1, wherein analyzing the mapped reverse end sequences comprises: counting the number of a plurality of eccDNA molecules mapped to a chromosome region, wherein the characteristic of the biological sample is a characteristic of the chromosome region; as well as The number is used to measure the identity of the chromosome region.

9. The method of claim 1, wherein the biological sample comprises a first tissue type and a second tissue type, wherein the first tissue type is homozygous for a first allele at a locus, and wherein the second tissue type is heterozygous for the first allele and the second allele at the locus, the method further comprising: determining a first number of mapped reverse end sequences having the first allele at the locus; determining a second number of mapped reverse end sequences having the second allele at the locus; as well as The fractional concentration of eccDNA molecules from the second tissue type is determined using the first number and the second number.

10. The method of claim 1, further comprising: Prior to cleaving the plurality of eccDNA molecules, linear DNA in the biological sample is reduced by exonuclease digestion.

11. A method comprising: receiving a biological sample from an organism, the biological sample comprising a plurality of extrachromosomal circular DNA (eccDNA) molecules, wherein each of the plurality of eccDNA molecules comprises a junction at which nucleotides at two separate genomic locations are directly adjacent to each other; digesting the plurality of eccDNA molecules with a restriction enzyme to form a set of linearized DNA molecules, each of which includes the junction; as well as For each of the set of linearized DNA molecules: sequencing at least two ends of the linearized DNA molecule to obtain one or more sequence reads; selecting a pair of end sequences of the linearized DNA molecule from the one or more sequence reads, the pair of end sequences excluding the junction; reversing the direction of each of the pair of end sequences to obtain a pair of reverse end sequences; mapping the pair of reverse end sequences to a reference genome; as well as The linearized DNA molecule is identified as being derived from an eccDNA molecule based on a pair of reverse end sequences mapped to the reference genome.

12. A method comprising: receiving a biological sample from an organism, the biological sample comprising a plurality of extrachromosomal circular DNA (eccDNA) molecules, wherein each of the plurality of eccDNA molecules comprises a junction at which nucleotides at two separate genomic locations are directly adjacent to each other; cleaving the plurality of eccDNA molecules using a transposase; Using the transposase, ligating an adapter sequence to the two cut ends of each of the plurality of eccDNA molecules, thereby forming a set of linearized DNA molecules, each of which includes the junction and the adapter sequence; as well as For each of the set of linearized DNA molecules: sequencing at least two ends of the linearized DNA molecule to obtain one or more sequence reads; Selecting a pair of end sequences of the linearized DNA molecule from the one or more sequence reads, the pair of end sequences not including the adapter sequence or the junction; reversing the direction of each of the pair of end sequences to obtain a pair of reverse end sequences; mapping the pair of reverse end sequences to a reference genome; as well as The linearized DNA molecule is identified as being derived from an eccDNA molecule based on a pair of reverse end sequences mapped to the reference genome.

13. The method of any one of claim 11 or claim 12, wherein: For each of the linearized DNA molecules, the one or more sequence reads include the junction.

14. The method of any one of claims 11 or 12, further comprising: The mapped reverse end sequences are analyzed to measure the properties of the biological sample.

15. The method of claim 14, further comprising: A collection value is determined for a set of a plurality of eccDNA molecules, wherein the mapped reverse end sequences are analyzed to measure a property of the biological sample using the collection value.

16. The method of any one of claims 1-15, wherein the eccDNA molecule is cell-free.

17. The method of any one of claims 1-10, wherein cleaving the plurality of eccDNA molecules comprises: The plurality of eccDNA molecules are digested with a restriction enzyme to form the set of linearized DNA molecules.

18. The method of claim 17, wherein the restriction enzyme cuts a specific sequence, the method further comprising: A specific sequence is identified that spans a pair of end sequences of at least a portion of the linearized DNA molecule.

19. The method of claim 17, wherein the restriction enzyme preferentially cuts a sequence of at least 4-bp.

20. The method of any one of claims 1-10, wherein cleaving the plurality of eccDNA molecules comprises: cleaving the plurality of eccDNA molecules using a transposase; as well as The adaptor sequence is ligated to the two cut ends of each of the plurality of eccDNA molecules using the transposase, thereby forming the set of linearized DNA molecules.

21. The method of any one of claims 1-20, wherein the biological sample is plasma, serum, blood, urine, saliva, cerebrospinal fluid, pleural fluid, peritoneal fluid, bronchoalveolar lavage fluid, ascites fluid, cervical lavage fluid, or sweat.

22. A system comprising means for performing the method of any one of claims 1-21.

23. A system comprising one or more processors configured to perform the method of any one of claims 1-21.

24. A system comprising modules respectively performing the method of any one of claims 1-21.

25. A computer product comprising a computer readable medium storing instructions for controlling a computer system to perform a method comprising: receiving a biological sample from an organism, the biological sample comprising a plurality of extrachromosomal circular DNA (eccDNA) molecules, wherein each of the plurality of eccDNA molecules comprises a junction at which nucleotides at two separate genomic locations are directly adjacent to each other; cleaving the plurality of eccDNA molecules using the cleavage tag to form a set of linearized DNA molecules, each of which includes the junction; For each of the set of linearized DNA molecules: sequencing at least two ends of the linearized DNA molecule to obtain one or more sequence reads; selecting a pair of end sequences of the linearized DNA molecule from the one or more sequence reads, the pair of end sequences excluding the junction; reversing the direction of each of the pair of end sequences to obtain a pair of reverse end sequences; as well as mapping the pair of reverse end sequences to a reference genome; as well as The mapped reverse end sequences are analyzed to measure the properties of the biological sample.

26. The computer product of claim 25, further comprising: detecting the plurality of eccDNA molecules based on a pair of reverse end sequences mapped to the reference genome; as well as An aggregate value of the detected eccDNA molecules is determined, wherein the mapped reverse end sequences are analyzed to measure a property of the biological sample using the aggregate value.

27. The computer product of claim 26, wherein the aggregate value comprises a count, size, or methylation level determined using the detected eccDNA molecules.

28. The computer product of claim 25, wherein: For each of the linearized DNA molecules, the one or more sequence reads include the junction.

29. The computer product of claim 28, wherein analyzing the mapped reverse end sequences comprises: comparing bases in the one or more sequence reads extending beyond each of the mapped reverse end sequences to a reference genome until a mismatch condition is identified; as well as Based on the positions of mismatch conditions in the reference genome, the end positions of the linearized DNA molecules that form the eccDNA molecules are identified.

30. The computer product of claim 29, wherein analyzing the mapped reverse end sequences further comprises: The end positions are used to determine the size of the linearized DNA molecule.

31. The computer product of claim 30, wherein analyzing the mapped reverse end sequences further comprises: determining a size distribution of sizes measured for the plurality of eccDNA molecules; as well as The size distribution is used to measure a property of the biological sample.

32. The computer product of claim 25, wherein analyzing the mapped reverse end sequences comprises: counting the number of a plurality of eccDNA molecules mapped to a chromosome region, wherein the characteristic of the biological sample is a characteristic of the chromosome region; as well as The number is used to measure the identity of the chromosome region.

33. The computer product of claim 32, wherein the characteristic is a copy number aberration in the chromosomal region, the method further comprising: measuring the methylation density in the chromosome region using the DNA molecules in the biological sample; as well as The copy number aberration and the methylation density are used to detect a condition of the organism.

34. The computer product of claim 33, wherein the methylation density is determined to exhibit hypermethylation by comparison to a cutoff value, and wherein the condition is fragile X syndrome or a triplet repeat expansion.

35. The computer product of claim 32, wherein the characteristic of the chromosome region is that the chromosome region carries information about a characteristic of the biological sample or has an aberration in the biological sample, which includes a sequence change, duplication, expansion, deletion, or amplification.

36. The computer product of claim 35, wherein the biological sample is obtained from a subject undergoing cancer screening, the method further comprising: A level of cancer in the organism is identified based on the chromosomal region having the aberration.

37. The computer product of claim 35, wherein the biological sample is obtained from a female carrying a fetus, and wherein the aberration is in the fetus.

38. The computer product of claim 35, wherein the characteristic is gender or genotype information.

39. The computer product of claim 25, wherein the biological sample comprises a first tissue type and a second tissue type, wherein the first tissue type is homozygous for a first allele at a locus, and wherein the second tissue type is heterozygous for the first allele and the second allele at the locus, the method further comprising: determining a first number of mapped reverse end sequences having the first allele at the locus; determining a second number of mapped reverse end sequences having the second allele at the locus; as well as The fractional concentration of eccDNA molecules from the second tissue type is determined using the first number and the second number.

40. The computer product of claim 25, further comprising: determining the number of sequence variants in the mapped reverse end sequences; as well as The number of sequence variants is used to determine the level of cancer.

41. The computer product of claim 25, further comprising: Prior to cleaving the plurality of eccDNA molecules, linear DNA in the biological sample is reduced by exonuclease digestion.

42. A computer product comprising a computer readable medium storing instructions for controlling a computer system to perform a method comprising: receiving a biological sample from an organism, the biological sample comprising a plurality of extrachromosomal circular DNA (eccDNA) molecules, wherein each of the plurality of eccDNA molecules comprises a junction at which nucleotides at two separate genomic locations are directly adjacent to each other; digesting the plurality of eccDNA molecules with a restriction enzyme to form a set of linearized DNA molecules, each of which includes the junction; as well as For each of the set of linearized DNA molecules: sequencing at least two ends of the linearized DNA molecule to obtain one or more sequence reads; selecting a pair of end sequences of the linearized DNA molecule from the one or more sequence reads, the pair of end sequences excluding the junction; reversing the direction of each of the pair of end sequences to obtain a pair of reverse end sequences; mapping the pair of reverse end sequences to a reference genome; as well as The linearized DNA molecule is identified as being derived from an eccDNA molecule based on a pair of reverse end sequences mapped to the reference genome.

43. A computer product comprising a computer readable medium storing instructions for controlling a computer system to perform a method comprising: receiving a biological sample from an organism, the biological sample comprising a plurality of extrachromosomal circular DNA (eccDNA) molecules, wherein each of the plurality of eccDNA molecules comprises a junction at which nucleotides at two separate genomic locations are directly adjacent to each other; cleaving the plurality of eccDNA molecules using a transposase; Using the transposase, ligating an adapter sequence to the two cut ends of each of the plurality of eccDNA molecules, thereby forming a set of linearized DNA molecules, each of which includes the junction and the adapter sequence; as well as For each of the set of linearized DNA molecules: sequencing at least two ends of the linearized DNA molecule to obtain one or more sequence reads; Selecting a pair of end sequences of the linearized DNA molecule from the one or more sequence reads, the pair of end sequences not including the adapter sequence or the junction; reversing the direction of each of the pair of end sequences to obtain a pair of reverse end sequences; mapping the pair of reverse end sequences to a reference genome; as well as The linearized DNA molecule is identified as being derived from an eccDNA molecule based on a pair of reverse end sequences mapped to the reference genome.

44. A computer product as claimed in any one of claims 42 or 43, wherein: For each of the linearized DNA molecules, the one or more sequence reads include the junction.

45. The computer product of any one of claims 42 or 43, further comprising: The mapped reverse end sequences are analyzed to measure the properties of the biological sample.

46. ​​The computer product of claim 45, further comprising: A collection value is determined for a set of a plurality of eccDNA molecules, wherein the mapped reverse end sequences are analyzed to measure a property of the biological sample using the collection value.

47. The computer product of any one of claims 25-46, wherein the eccDNA molecule is cell-free.

48. The computer product of any one of claims 25-41, wherein cleaving the plurality of eccDNA molecules comprises: The plurality of eccDNA molecules are digested with a restriction enzyme to form the set of linearized DNA molecules.

49. The computer product of claim 48, wherein the restriction enzyme cuts a specific sequence, the method further comprising: A specific sequence is identified that spans a pair of end sequences of at least a portion of the linearized DNA molecule.

50. The computer program product of claim 49, wherein the restriction enzyme preferentially cuts sequences of at least 4-bp.

51. The computer product of any one of claims 25-41, wherein cleaving the plurality of eccDNA molecules comprises: cleaving the plurality of eccDNA molecules using a transposase; as well as The adaptor sequence is ligated to the two cut ends of each of the plurality of eccDNA molecules using the transposase, thereby forming the set of linearized DNA molecules.

52. The computer product of any one of claims 25-51, wherein the biological sample is plasma, serum, blood, urine, saliva, cerebrospinal fluid, pleural fluid, peritoneal fluid, bronchoalveolar lavage fluid, ascites fluid, cervical lavage fluid, or sweat.

53. A system comprising: The computer product of any one of claims 25-52; as well as One or more processors for executing instructions stored on the computer-readable medium.

Images