Using cell-free DNA in blood to detect blood disorders

Abstract

The present disclosure provides techniques for detecting blood disorders using cell-free DNA in a blood sample, e.g., using plasma or serum. For example, an analysis can target one or more differentially methylated regions specific to a particular blood cell lineage (e.g., erythroblasts). Methylation levels can be quantified from the analysis to determine the amount of methylated or unmethylated DNA fragments in the cell-free mixture of the blood sample. The methylation levels can be compared to one or more cutoff values, e.g., corresponding to a normal range for the particular blood cell lineage, as part of determining a level of a blood disorder.

Description

[0001] Related Applications

[0002] This application is a divisional application of Chinese Patent Application No. 201780033380.0 and claims priority to and is a non-provisional of U.S. Provisional Application No. 62 / 343,050, filed May 30, 2016, entitled "Detection of Blood Disorders Using Cell-Free DNA in Blood," the entire contents of which are incorporated herein by reference for all purposes. BACKGROUND

[0003] To determine whether a blood disorder (e.g., anemia) is present in a human, conventional techniques perform histological examination of a bone marrow biopsy. However, a bone marrow biopsy is an invasive procedure, resulting in pain and anxiety for the patient undergoing the procedure. Thus, it is desirable to identify new techniques for detecting and characterizing blood disorders in a human.

[0004] Anemia can be caused by a number of clinical conditions, each with its own treatment. Thus, it would be clinically useful to determine the cause of an anemia case, and subsequently further investigate or treat accordingly. One cause of anemia is a deficiency in nutrients necessary for erythropoiesis (the process of producing red blood cells), such as (but not limited to) iron, B12, folate, etc. Another cause of anemia is blood loss, which can be acute or chronic. Blood loss can be caused by, for example, heavy menstrual bleeding or gastrointestinal bleeding. Anemia is also commonly seen in many chronic conditions, also known as anemia of chronic disease, which can be found in cancer and inflammatory bowel disease.

[0005] Thus, it is desirable to provide new techniques for screening a subject for a blood disorder, for determining the cause of a blood disorder, for monitoring a subject with a blood disorder, and / or for determining an appropriate treatment for a subject with a blood disorder. SUMMARY

[0006] Some embodiments provide systems, methods, and apparatus for detecting a blood disorder using cell-free DNA in a blood sample, for example, using plasma or serum. For example, the analysis can target one or more differentially methylated regions specific to a particular blood cell lineage (e.g., erythroblasts). The level of methylation can be quantified from the analysis to determine the amount of methylated or unmethylated DNA fragments in the cell-free mixture of the blood sample. The level of methylation can be compared to one or more cutoff values, for example, corresponding to a normal range for the particular blood cell lineage, as part of determining a level of a blood disorder. Some embodiments can measure the amount of DNA of a particular blood cell lineage (e.g., erythroblast DNA) in a blood sample in a similar manner using one or more levels of methylation.

[0007] Such analysis can provide detection of blood disorders without the need for invasive procedures such as bone marrow biopsy. For example, our results indicate that bone marrow cells contribute a large proportion of the circulating cell-free DNA. Analysis of methylation markers of hematopoietic cells in the circulating cell-free DNA can reflect the state of bone marrow cells. These embodiments can be particularly useful for monitoring the response of bone marrow to treatment, for example, the response of a patient with iron deficiency anemia to oral iron therapy. Embodiments can also be used to assign patients to different procedures, for example, bone marrow biopsy or less invasive studies.

[0008] Other embodiments relate to systems and computer readable media related to the methods described herein.

[0009] As non-limiting examples, the present application provides the following embodiments:

[0010] 1. A method of analyzing a blood sample of a mammal, the method comprising:

[0011] obtaining a cell-free mixture of the blood sample, the cell-free mixture comprising cell-free DNA from a plurality of cell lineages;

[0012] contacting DNA fragments in the cell-free mixture with an assay corresponding to one or more differentially methylated regions, each of the one or more differentially methylated regions having specificity for a particular blood cell lineage by being hypo- or hyper-methylated relative to other cell lineages;

[0013] detecting a first number of methylated or unmethylated DNA fragments in the cell-free mixture at the one or more differentially methylated regions based on signals obtained from the assay;

[0014] determining a methylation level using the first number; and

[0015] comparing the methylation level to one or more cutoff values as part of determining a classification of a blood disorder in the mammal.

[0016] 2. The method of embodiment 1, further comprising:

[0017] determining a total number of DNA fragments in the cell-free mixture at the one or more differentially methylated regions; and

[0018] determining the methylation level using the first number and the total number.

[0019] 3. The method of embodiment 1, further comprising:

[0020] determining a volume of the cell-free mixture, wherein the methylation level determines the first number of the cell-free mixture and the volume.

[0021] 4. The method of embodiment 1, wherein obtaining the cell-free mixture comprises:

[0022] separating the cell-free mixture from the blood sample, the cell-free mixture comprising plasma or serum.

[0023] 5. The method of embodiment 1, further comprising identifying the one or more differentially methylated regions by:

[0024] obtaining a methylation index for a plurality of sites for each of a plurality of cell lineages, the plurality of cell lineages including the particular blood cell lineage and the other cell lineages;

[0025] comparing the methylation index for each of the plurality of cell lineages at each of the plurality of sites;

[0026] identifying one or more of the plurality of sites that have a methylation index in the particular blood cell lineage that is below / above a first methylation threshold and in each of the other cell lineages that is above / below a second methylation threshold; and

[0027] identifying a differentially methylated region containing the one or more sites.

[0028] 6. The method of embodiment 1, further comprising determining the one or more cutoff values, comprising:

[0029] obtaining a plurality of samples, each sample known to have a particular classification of the blood disorder, the plurality of samples having at least two classifications of the blood disorder;

[0030] determining a methylation level of the one or more differentially methylated regions for each of the plurality of samples;

[0031] identifying a first group of samples having a first classification of the blood disorder;

[0032] identifying a second group of samples having a second classification of the blood disorder, the first group of samples collectively having a statistically higher methylation level than the second group of samples; and

[0033] determining a cutoff value that distinguishes the first group of samples and the second group of samples within a particular specificity and sensitivity.

[0034] 7. The method of embodiment 1, wherein determining the classification of the blood disorder comprises identifying a particular type of the blood disorder.

[0035] 8. The method of embodiment 1, further comprising:

[0036] treating the blood disorder in the mammal in response to determining that the classification of the blood disorder indicates that the mammal has the blood disorder;

[0037] after the treatment, repeating the analysis to determine updated methylation levels; and

[0038] determining whether to continue the treatment based on the updated methylation levels.

[0039] 9. The method of embodiment 8, wherein determining whether to continue the treatment comprises:

[0040] stopping the treatment, increasing a dose of the treatment, or performing a different treatment when the updated methylation levels do not change relative to the methylation levels to within a specified threshold.

[0041] 10. The method of embodiment 8, wherein determining whether to continue the treatment comprises:

[0042] continuing the treatment when the updated methylation levels change relative to the methylation levels to within a specified threshold.

[0043] 11. The method of embodiment 1, further comprising:

[0044] determining that a blood disorder is present based on comparing the methylation levels to one or more cutoff values; and

[0045] performing a bone marrow biopsy in response to determining that the blood disorder is present.

[0046] 12. The method of embodiment 1, wherein the analysis is a PCR analysis or a sequencing analysis.

[0047] 13. The method of embodiment 1, wherein the one or more differentially methylated regions comprise CpG sites.

[0048] 14. The method of embodiment 13, wherein a first region of the one or more differentially methylated regions comprises a plurality of CpG sites within 100 bp of each other, and wherein the plurality of CpG sites are all either hypo- or hyper-methylated.

[0049] 15. The method of embodiment 14, wherein the plurality of CpG sites span 100 bp or less on a reference genome corresponding to the mammal.

[0050] 16. The method of embodiment 1, wherein the particular blood cell lineage is red blood cells.

[0051] 17. The method of embodiment 16, further comprising:

[0052] measuring a hemoglobin level of the blood sample;

[0053] comparing the hemoglobin level to a hemoglobin threshold; and

[0054] determining the classification of the blood disorder further based on the comparison of the hemoglobin level to the hemoglobin threshold.

[0055] 18. The method of embodiment 16, wherein one of the one or more differentially methylated regions is in the FECH gene.

[0056] 19. The method of embodiment 16, wherein one of the one or more differentially methylated regions is in chromosome 12 at genomic coordinates 48227688-48227701.

[0057] 20. The method of embodiment 16, wherein one of the one or more differentially methylated regions is in chromosome 12 at genomic coordinates 48228144-48228154.

[0058] 21. The method of embodiment 16, wherein the blood disorder is anemia.

[0059] 22. The method of embodiment 21, wherein the classification of the blood disorder corresponds to increased erythropoietic activity, moderate erythropoietic activity, or decreased erythropoietic activity.

[0060] 23. The method of embodiment 22, wherein the classification of the blood disorder is increased erythropoietic activity of beta thalassemia.

[0061] 24. The method of embodiment 22, wherein the classification of the blood disorder is moderate erythropoietic activity from iron deficiency anemia.

[0062] 25. The method of embodiment 22, wherein the classification of the blood disorder is decreased erythropoietic activity of aplastic anemia or chronic kidney failure.

[0063] 26. The method of embodiment 1, wherein the one or more differentially methylated regions are hypomethylated.

[0064] 27. The method of embodiment 1, wherein the cell-free mixture is plasma.

[0065] 28. A method of measuring the amount of cells of a particular cell lineage in a biological sample, the method comprising:

[0066] obtaining a cell-free mixture of the biological sample, the cell-free mixture comprising cell-free DNA from a plurality of cell lineages;

[0067] contacting DNA fragments in the cell-free mixture with an assay corresponding to one or more differentially methylated regions, each of the one or more differentially methylated regions having specificity for a particular cell lineage by being hypo- or hyper-methylated relative to other cell lineages;

[0068] detecting a first number of methylated or unmethylated DNA fragments in the cell-free mixture at the one or more differentially methylated regions based on signals obtained from the assay;

[0069] determining a first methylation level using the first number;

[0070] obtaining one or more calibration data points, wherein each calibration data point specifies (1) an amount of cells of the particular cell lineage and (2) a calibration methylation level, and wherein the one or more calibration data points are determined from a plurality of calibration samples;

[0071] comparing the first methylation level to a calibration methylation level of at least one calibration data point; and

[0072] estimating the amount of cells of the particular cell lineage in the biological sample based on the comparison.

[0073] 29. The method of embodiment 28, wherein the particular cell lineage is a particular blood cell lineage.

[0074] 30. The method of embodiment 28, wherein the one or more calibration data points are a plurality of calibration data points, and wherein the calibration data points form a calibration curve.

[0075] 31. A computer product comprising a computer readable medium storing a plurality of instructions for controlling a system to perform the method of any one of embodiments 1-30.

[0076] 32. A system comprising:

[0077] the computer product of embodiment 31; and

[0078] one or more processors for executing the instructions stored on the computer readable medium.

[0079] 33. A system configured to perform the method of any one of embodiments 1-30.

[0080] 34. A computer product comprising a computer readable medium storing a plurality of instructions for controlling a system to analyze a blood sample of a mammal by performing the following:

[0081] detecting, based on signals obtained from the analysis, a first number of methylated or unmethylated DNA fragments in a cell-free mixture of the blood sample at one or more differentially methylated regions, each of the one or more differentially methylated regions having specificity for a particular blood cell lineage by being hypo- or hypermethylated relative to other cell lineages;

[0082] determining a methylation level using the first number; and

[0083] comparing the methylation level to one or more cutoff values as part of determining a classification of a blood disorder in the mammal.

[0084] 35. A computer product comprising a computer readable medium storing a plurality of instructions for controlling a system to measure a cell quantity of a particular cell lineage in a biological sample by performing the following:

[0085] detecting, based on signals obtained from the analysis, a first number of methylated or unmethylated DNA fragments in a cell-free mixture of the biological sample at one or more differentially methylated regions, each of the one or more differentially methylated regions having specificity for the particular cell lineage by being hypo- or hypermethylated relative to other cell lineages;

[0086] determining a first methylation level using the first number;

[0087] obtaining one or more calibration data points, wherein each calibration data point specifies (1) a cell quantity of the particular cell lineage and (2) a calibration methylation level, and wherein the one or more calibration data points are determined from a plurality of calibration samples;

[0088] comparing the first methylation level to the calibration methylation level of at least one calibration data point; and

[0089] estimating a cell quantity of the particular cell lineage in the biological sample based on the comparison.

[0090] The nature and advantages of embodiments of the application will become more apparent to those skilled in the art after consideration of the following detailed description and drawings. BRIEF DESCRIPTION OF DRAWINGS

[0091] Figure 1Figure showing the methylation density of CpG sites within the promoter of the ferrochelatase (FECH) gene according to embodiments of the application.

[0092] Figure 2A and 2B Figure showing the analysis of globally methylated and unmethylated DNA using digital PCR analysis designed to detect methylated and unmethylated DNA according to embodiments of the application.

[0093] Figure 3A Figure showing the correlation between E% in blood cells and the number of nucleated RBCs (erythroblasts) according to embodiments of the application. Figure 3B Figure is a flowchart illustrating a method 300 of determining the amount of cells of a specific cell lineage in a biological sample by analyzing cell-free DNA according to embodiments of the application.

[0094] Figure 4 Figure showing Unmeth% in buffy coat and plasma of healthy non-pregnant subjects and pregnant women at different gestational ages according to embodiments of the application.

[0095] Figure 5 Figure showing the lack of correlation between Unmeth% in buffy coat and plasma.

[0096] Figure 6A and 6B Figure showing the percentage of erythrocyte DNA (E% (FECH)) in healthy subjects according to embodiments of the application. E% can be defined as the same as Unmeth%.

[0097] Figure 7 Figure showing the lack of correlation between E% (FECH) results of plasma DNA and age of healthy subjects.

[0098] Figure 8 Figure is a plot of Unmeth% against hemoglobin concentration in patients with aplastic anemia, beta-thalassemia major and healthy control subjects according to embodiments of the application.

[0099] Figure 9 Figure is a plot of plasma Unmeth% in patients with iron (Fe) deficiency anemia and acute blood loss according to embodiments of the application.

[0100] Figure 10 Figure showing the relationship between the percentage of erythrocyte DNA (E% (FECH)) in plasma and hemoglobin levels in patients with aplastic anemia, chronic renal failure (CRF), beta-thalassemia major, iron deficiency anemia and healthy subjects according to embodiments of the application.

[0101] Figure 11A and11B shows the relationship between reticulocyte count / index and hemoglobin level in anemic patients with aplastic anemia, chronic renal failure (CRF), beta-thalassemia major, and iron deficiency anemia according to embodiments of the present application.

[0102] Figure 12 is a plot of plasma Unmeth% in patients with myelodysplastic syndrome and polycythemia vera according to embodiments of the present application.

[0103] Figure 13A shows the percentage of erythroid DNA in plasma (E%(FECH)) in patients with aplastic anemia (AA) and myelodysplastic syndrome (MDS) according to embodiments of the present application. FIG. 13B shows the percentage of erythroid DNA in plasma (E%(FECH)) between treatment responsive and non-responsive groups in aplastic anemia according to embodiments of the present application.

[0104] Figure 14 is a plot of plasma Unmeth% against hemoglobin concentration in a normal subject and two patients with leukemia according to embodiments of the present application.

[0105] Figure 15A and 15B shows the methylation density of CpG sites within the erythroid-specific DMR on chromosome 12 according to embodiments of the present application.

[0106] Figure 16 shows histone modifications (H3K4mel and H3K27Ac) for two other erythroid-specific DMRs (Ery-1 and Ery-2) from the ENCODE database.

[0107] Figure 17A and 17B shows the correlation between the percentage of erythroid DNA sequences (E%) in buffy coat DNA of beta-thalassemia major patients measured by digital PCR analysis targeting the Ery-1 marker Figure 17A ) and the Ery-2 marker (FIG. 17B) with the percentage of erythroid cells in all peripheral blood leukocytes measured using an automated hematology analyzer.

[0108] Figure 18A and 18B shows the correlation of E%(FECH) results with E%(Ery-1) and E%(Ery-2) for buffy coat DNA of beta-thalassemia major patients.

[0109] Figure 19Figure showing the percentage of erythroid DNA in plasma in healthy subjects and patients with aplastic anemia and beta-thalassemia major according to embodiments of the application using digital PCR analysis targeting three erythroid-specific DMRs.

[0110] Figure 20A and 20B Figure showing serial measurements of the percentage of erythroid DNA in plasma (E% (FECH)) and reticulocyte count in a patient with iron deficiency anemia before treatment and two days after receiving intravenous iron therapy according to embodiments of the application.

[0111] Figure 21A Figure showing serial changes in plasma E% at erythroid DMRs in a patient with iron deficiency anemia due to menorrhagia receiving oral iron treatment according to embodiments of the application. Figure 21B Figure showing changes in hemoglobin after treatment.

[0112] Figure 22 Figure showing serial changes in plasma Unmeth% at erythroid DMRs in patients with chronic kidney disease (CKD) receiving recombinant erythropoietin (EPO) or erythropoiesis stimulating agent (ESA) treatment according to embodiments of the application.

[0113] Figure 23A Figure showing serial changes in plasma Unmeth% at erythroid DMRs in patients with aplastic anemia receiving anti-thymocyte globulin (ATG) treatment or cyclosporine as immunosuppressive therapy according to embodiments of the application. Figure 23B Figure showing serial changes in hemoglobin in patients with aplastic anemia receiving treatment according to embodiments of the application.

[0114] Figure 24A and 24B Figure showing Unmeth% in plasma against hemoglobin concentration in four patients with aplastic anemia according to embodiments of the application.

[0115] Figure 25 Figure illustrating a box-whisker plot showing the absolute concentration of erythroid DNA at the FECH gene-associated DMR (copies per milliliter of plasma) in healthy subjects and anemic patients according to embodiments of the application.

[0116] Figure 26 is a flowchart illustrating a method of analyzing a blood sample of a mammal according to embodiments of the application.

[0117] Figure 27 Figure illustrating a system 2700 according to embodiments of the application.

[0118] Figure 28A block diagram showing an example computer system that can be used with the systems and methods according to embodiments of the application. DETAILED DESCRIPTION

[0119] TERMINOLOGY

[0120] A "methylome" provides a measure of the amount of DNA methylation at a plurality of sites or loci in a genome. The methylome can correspond to all of a genome, a substantial portion of a genome, or a relatively small portion of a genome.

[0121] A "cell lineage" represents the developmental history of a tissue or organ from a fertilized embryo. Different types of tissues (e.g., different types of blood cells) will have different cell lineages. Red blood cells (RBCs) are derived from proerythroblasts through a series of intermediate cells. Proerythroblasts, promegakaryocytes, and myeloblasts are derived from a common myeloid progenitor cell. Lymphocytes are derived from a common lymphoid progenitor cell. A nucleated RBC is a erythroblast, an immature anucleated RBC is a reticulocyte, and a mature anucleated RBC is an erythrocyte, which is a red blood cell that carries hemoglobin in the bloodstream.

[0122] A "cell-free mixture" corresponds to a sample that includes cell-free DNA fragments from various cells. For example, a cell-free mixture can include cell-free DNA fragments from various cell lineages. Plasma and serum are examples of cell-free mixtures obtained from a blood sample (e.g., by centrifugation). Other cell-free mixtures can be from other biological samples. A "biological sample" refers to any sample taken from a subject (e.g., a human, such as a pregnant woman, a person with cancer, or an individual suspected of having cancer, an organ transplant recipient, or a subject suspected of having a disease process involving an organ, such as the heart in myocardial infarction, the brain in stroke, or the hematopoietic system in anemia) and containing one or more relevant nucleic acid molecules. A biological sample can be a bodily fluid, such as blood, plasma, serum, urine, vaginal fluid, hydrosalpinx (e.g., testicular) fluid, or a vaginal lavage, pleural fluid, ascitic fluid, cerebrospinal fluid, saliva, sweat, tears, sputum, bronchoalveolar lavage, etc. Fecal 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 by a centrifugation protocol) can be cell-free (as opposed to cellular), such as greater than 50%, 60%, 70%, 80%, 90%, 95%, or 99%. A centrifugation protocol can include 3,000 g x 10 minutes, obtaining a fluid fraction, and re-centrifuging at 30,000 g for another 10 minutes to remove residual cells.

[0123] A "plasma methylome" is a methylome determined from plasma or serum of an animal (e.g., a human). A plasma methylome is an example of a cell-free methylome, as plasma and serum include cell-free DNA. A plasma methylome is also an example of a mixed methylome, as it is a mixture of DNA from different organs or tissues or cells within the body. In one embodiment, these cells are hematopoietic cells, including, but not limited to, cells of the erythroid (i.e., red blood cell) lineage, myeloid lineage (e.g., neutrophils and their precursors), and megakaryocyte lineage. During pregnancy, a plasma methylome can contain methylome information from both the fetus and the mother. In a patient with cancer, a plasma methylome can contain methylome information from tumor cells and other cells within the patient's body. A "cell methylome" corresponds to a methylome determined from a cell (e.g., a blood cell) of a patient. The methylome of a blood cell is referred to as a blood cell methylome (or blood methylome). Techniques for determining a methylome are further described in PCT Patent Application No. WO 2014 / 043763, entitled "Non-invasive Determination of a Fetal or Tumor Methylome from Plasma," the disclosure of which is incorporated by reference in its entirety for all purposes.

[0124] A "site" corresponds to a single site, which can be a single base position or a set of related base positions, such as a CpG site. 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.

[0125] A "methylation index" for each genomic site (e.g., a CpG site) can refer to the proportion of DNA fragments within the total number of reads covering the site that show methylation at the site (e.g., as determined from sequence reads or probes). A "read" can correspond to information obtained from a DNA fragment (e.g., the methylation status at a site). Reads can be obtained using reagents (e.g., primers or probes) that preferentially hybridize to DNA fragments of a particular methylation status. Typically, these reagents are applied after a method that differentially modifies DNA molecules according to their methylation status, such as bisulfite conversion or methylation-sensitive restriction enzymes. A read can be a sequence read. A "sequence read" refers to a string of nucleotides sequenced from any portion or the entirety of a nucleic acid molecule. For example, a sequence read can be a short string of nucleotides (e.g., 20-150) sequenced from a nucleic acid fragment, a short string of nucleotides at one or both ends of a nucleic acid fragment, or sequencing of the entire nucleic acid fragment present in a biological sample. Sequence reads can be obtained in a variety of ways, such as using sequencing technology or using probes (e.g., hybridization arrays or capture probes, or amplification techniques such as polymerase chain reaction (PCR) or linear amplification using a single primer or isothermal amplification).

[0126] The "methylation density" of a region can refer to the number of reads at sites within the region that show methylation divided by the total number of reads in the region that cover the sites. The sites can have specific characteristics, for example, be CpG sites. Thus, the "CpG methylation density" of a region can refer to the number of reads that show CpG methylation divided by the total number of reads in the region that cover CpG sites (e.g., a particular CpG site, CpG sites within a CpG island, or a larger region). For example, the methylation density per 100-kb bin in the human genome can be determined using the proportion of unconverted cytosines (which correspond to methylated cytosines) at CpG sites after bisulfite treatment out of all CpG sites covered by sequence reads mapping to a 100-kb region. This analysis can also be performed for other bin sizes, for example, 500 bp, 5 kb, 10 kb, 50-kb, or 1-Mb, etc. The region can be the whole genome or a chromosome or a portion of a chromosome (e.g., a chromosome arm). When a region includes only one CpG site, the methylation index of that CpG site is the same as the methylation density of the region. The "proportion of methylated cytosines" can refer to the number of cytosine sites "C's" that show methylation (e.g., unconverted after bisulfite conversion) out of the total number of cytosine residues analyzed (i.e., including cytosines outside of the CpG context in the region). Methylation index, methylation density, and proportion of methylated cytosines are examples of "methylation levels." In addition to bisulfite conversion, other methods known to those of skill in the art can be used to interrogate the methylation status of a DNA molecule, including (but not limited to) enzymes sensitive to methylation status (e.g., methylation-sensitive restriction enzymes), methylation-binding proteins, single molecule sequencing using platforms sensitive to methylation status (e.g., nanopore sequencing (Schreiber et al. Proc Natl Acad Sci 2013; 110: 18910-18915) and Pacific Biosciences' single molecule real-time analysis (Flusberg et al. Nat Methods 2010; 7: 461-465)).

[0127] A "methylation profile" (also referred to as a methylation state) includes information related to DNA methylation of a region. Information related to DNA methylation can include, but is not limited to, methylation index of CpG sites, methylation density of CpG sites in a region, distribution of CpG sites across a region, methylation pattern or level of each individual CpG site within a region containing more than one CpG site, and non-CpG methylation. Most of the methylation profile in a genome can be considered to be equivalent to the methylome. "DNA methylation" in a mammalian genome generally refers to the addition of a methyl group to the 5' carbon of a cytosine residue in a CpG dinucleotide (i.e., 5-methylcytosine). DNA methylation can occur in other contexts of cytosine, 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 methylation has also been reported, such as N 6 - methyladenine.

[0128] A "tissue" corresponds to a group of cells that are grouped together as a functional unit. More than one type of cell can be found in a single tissue. Different types of tissues can be composed of different types of cells (e.g., hepatocytes, alveolar cells, or blood cells), and can correspond to tissues from different organisms (mother versus fetus) or to healthy cells versus tumor cells. A "reference tissue" corresponds to a tissue used to determine tissue-specific methylation levels. Tissue-specific methylation levels for a tissue type can be determined using multiple samples of the same tissue type from different individuals. The same tissue from the same individual can exhibit differences at different times due to physiology (e.g., pregnancy) or pathology (e.g., cancer or anemia or infection or mutation). The same tissue from different individuals can exhibit differences due to physiology (e.g., age, sex) or pathology (e.g., cancer or anemia or infection or mutation).

[0129] The term "level of a condition," also referred to as "classification of a condition," can refer to a classification of the type of condition, the stage of the condition, and / or other measures of the severity of the condition, whether or not the condition is present. The level can be a number or other character. The level can be zero. The level of a condition can be used in a variety of ways. For example, screening can examine whether a person previously unknown to have a condition has the condition. Assessment can study a person already diagnosed with a condition to monitor the progression of the condition over time, to study the utility of a therapy, or to determine a prognosis. In one embodiment, a prognosis can be expressed as the likelihood that a patient will die from the condition or the likelihood that the condition will progress after a particular duration of time. Detection can mean'screening' or can mean examining whether a person with potential characteristics of a condition (e.g., symptoms or other positive tests) has the condition.

[0130] Anemia is a condition in which the number of red blood cells or their oxygen-carrying capacity is insufficient to meet physiological needs, which can vary by age, sex, altitude, smoking, and pregnancy status. According to the World Health Organization (WHO), anemia can be diagnosed when the hemoglobin concentration is less than 130 g / L in men and less than 110 g / L in women. The term "degree of anemia" can be reflected by the hemoglobin concentration of a subject. Lower hemoglobin levels indicate a more severe degree of anemia. According to the WHO, severe anemia is defined as a hemoglobin concentration <80 g / L in men and <70 g / L in women, moderate anemia is defined as a hemoglobin concentration of 80-109 g / L in men and 70-99 g / L in women, and mild anemia is defined as a hemoglobin concentration of 110-129 g / L in men and 100-109 g / L in women.

[0131] A "separation value" corresponds to a difference or ratio involving two values (e.g., two contribution rates or two methylation levels). A separation value can be a simple difference or ratio. A separation value can include other factors, such as a multiplication factor. As other examples, a difference or ratio of a function of the values can be used, such as a difference or ratio of the natural logarithm (ln) of the two values. A separation value can include both a difference and a ratio.

[0132] As used herein, the term "classification" refers to any number or other character that is related to a particular characteristic of a sample. For example, a "+" symbol (or the word "positive") can indicate that a sample is classified as having a deletion or amplification. A classification can be binary (e.g., positive or negative) or have more levels of classification (e.g., a scale from 1 to 10 or 0 to 1). The terms "cut-off value" and "threshold value" refer to a predetermined number used in an operation. A threshold value can be a value above or below which a particular classification applies. Either of these terms can be used in either of these contexts.

[0133] In some embodiments, the contribution of cell-free DNA from erythroblasts (also known as circulating DNA) is quantified using one or more methylation markers (e.g., one marker per marker) that are specific to erythroblasts relative to cell-free DNA from other tissues. A marker (e.g., a differentially methylated region, DMR) can include one site or a set of sites that contribute to the same marker.

[0134] The contribution of cell-free DNA from erythroblasts can be used to determine the level of a blood disorder, such as anemia. For example, embodiments can be used to analyze anemia in a fetus, a neonate, or a child. In the context of anemia, embodiments can be used to study a person suspected of having anemia or who has been diagnosed with anemia: (i) to elucidate the cause of the anemia; (ii) to monitor the progression of the clinical state over time; (iii) to study the effect of a therapy; or (iv) to determine the prognosis. Thus, embodiments have identified erythrocyte DNA as a major component of a hitherto unrecognized circulating DNA pool and as a non-invasive biomarker for differential diagnosis and monitoring of anemia and other blood disorders.

[0135] I. INTRODUCTION

[0136] Plasma DNA is an increasingly practiced analyte for molecular diagnostics. Research efforts are ongoing regarding its clinical applications, especially in non-invasive prenatal testing (1-7) and oncology (8-12). Despite the variety of clinical applications, the tissue origin of circulating DNA is not fully understood.

[0137] It has been shown that using sex-mismatched bone marrow transplantation as a model system, circulating DNA is predominantly released from hematopoietic cells (13, 14). Kun et al. recently demonstrated that a significant proportion of plasma DNA has methylation signatures of neutrophils and lymphocytes (15). However, there is currently no information on whether DNA from the erythroid lineage (erythroblasts) can also be detected in plasma.

[0138] Red blood cells (RBCs) are the largest population of hematopoietic cells in the blood. The concentration of red blood cells (RBCs) is about 5 x 1011per liter of blood. Given that the life span of each RBC is about 120 days, the body needs to produce 2 x 1011RBCs / day or 9.7 x 1011RBCs / hour. Mature RBCs in humans have no nucleus. 12 11 9 In the context of anemia, embodiments can be used to study a person suspected of having anemia or who has been diagnosed with anemia: (i) to elucidate the cause of the anemia; (ii) to monitor the progression of the clinical state over time; (iii) to study the effect of a therapy; or (iv) to determine the prognosis. Thus, embodiments have identified erythrocyte DNA as a major component of a hitherto unrecognized circulating DNA pool and as a non-invasive biomarker for differential diagnosis and monitoring of anemia and other blood disorders.

[0139] During the enucleation step, erythroblasts lose their nuclei and mature into reticulocytes in the bone marrow (16). The enucleation process is a complex multi-step process involving tightly regulated cellular signaling and cytoskeletal actions. The nuclear material of erythroblasts is phagocytosed and degraded by bone marrow macrophages in the erythroblastic islands (e.g., in the bone marrow) (17). We hypothesize that some of the degraded DNA material from the erythroid lineage of the bone marrow will be released into the circulation.

[0140] ​​Embodiments can identify methylation markers of DNA from cells of red blood cell origin and use these markers to determine whether red blood cell DNA is detectable in human plasma. High-resolution reference methylomes of different tissues and hematopoietic cell types have become publicly available through collaborative projects including the BLUEPRINT project (18, 19) and the Roadmap Epigenomics Project (20). We and others have previously demonstrated that it is possible to track the origin of plasma DNA through analysis of tissue-associated methylation markers (15, 21, 22). Further details of such analysis to determine the contribution of certain tissues to a cell-free mixture (e.g., plasma) can be found in PCT Patent Application No. WO 2016 / 008451 entitled "Analysis of Methylation Patterns of Tissues in DNA Mixtures," the disclosure of which is incorporated by reference in its entirety for all purposes.

[0141] To validate our hypothesis and demonstrate the presence of red blood cell DNA in plasma, we identified erythroblast-specific differential methylation regions (DMRs) by analyzing the methylation profiles of erythroblasts and other tissue types. Based on these findings, we developed digital polymerase chain reaction (PCR) assays targeting erythroblast-specific DMRs to enable quantitative analysis of red blood cell DNA in biological samples. Specifically, using high-resolution methylation profiles of erythroblasts and other tissue types, three genomic loci were found to be hypomethylated in erythroblasts but highly methylated in other cell types. Digital PCR assays were developed for measuring red blood cell DNA using the differential methylation regions of each locus.

[0142] We applied these digital PCR assays to investigate plasma samples of healthy subjects and patients with different types of anemia. We also explored the potential clinical utility of the assays in anemia assessment. Although the examples use PCR assays, other assays such as sequencing can also be used.

[0143] In subjects with anemia of different etiologies, we showed that quantitative analysis of circulating red blood cell DNA (e.g., using methylation markers) reflects erythropoietic activity in the bone marrow. For patients with reduced erythropoietic activity, such as exemplified by aplastic anemia, the percentage of circulating red blood cell DNA is reduced. For patients with increased but ineffective erythropoiesis, such as exemplified by beta-thalassemia major, the percentage is increased. Furthermore, we found that plasma levels of red blood cell DNA correlate with treatment response in aplastic anemia and iron deficiency anemia. Plasma DNA analysis using digital PCR assays targeting the other two differential methylation regions showed similar findings.

[0144] II. Erythroblast differential methylation regions (DMRs)

[0145] We hypothesized that the process of red blood cell enucleation or other processes involved in RBC maturation would make a significant contribution to the circulating cell-free DNA pool. To determine the contribution of circulating DNA from erythroblasts, we identified differential methylation regions (DMRs) in erythroblast DNA by comparing the DNA methylation profiles of erythroblasts to other tissues and blood cells. We studied the methylation profiles of erythroblasts and other blood cells (neutrophils, B and T lymphocytes) and tissues (liver, lung, colon, small intestine, pancreas, adrenal gland, esophagus, heart, brain, and placenta) from the BLUEPRINT project and Roadmap Epigenomics project and methylation panels generated by our group (18-20, 23).

[0146] In a simple example, one or more DMRs can be used directly to determine the contribution of circulating DNA from red blood cells, for example, by determining the percentage of methylated (for highly methylated DMRs) or unmethylated (for lowly methylated DMRs) DNA fragments. The percentages can be used directly or modified (e.g., multiplied by a scaling factor). Other embodiments can perform more complex procedures, such as solving a system of linear equations. As described in PCT Patent Application No. WO 2016 / 008451, the methylation levels at N genomic loci can be utilized to calculate the contributions of M tissues, where M is less than or equal to N. The methylation level at each locus can be calculated for each tissue. A system of linear equations, Ax = b, can be solved, where b is a vector of measured methylation densities at the N loci, x is a vector of contributions of the M tissues, and A is a matrix of M rows and N columns, where each row provides the methylation densities of the N tissues at the particular loci of that row. If M is less than N, then a least squares optimization can be performed. The matrix A of size N x M can be formed from tissue-specific methylation levels of the reference tissues, as obtained from the sources described above.

[0147] A. Identification of DMRs

[0148] To identify differential methylation regions (DMRs), a particular type of tissue / cell lineage (e.g., erythroblasts) can be isolated and then analyzed, for example, using methylation-aware sequencing, as described herein. The methylation densities of loci across tissue types (e.g., only two types of erythroblasts and others) can be analyzed to determine if there is sufficient difference to identify loci for use in DMRs.

[0149] In some embodiments, one or more of the following criteria can be used to identify methylation markers of erythroblasts. (1) A CpG site is hypo-methylated in erythroblasts if the methylation density of the CpG site is less than 20% in erythroblasts and more than 80% in other blood cells and tissues, and vice versa. (2). To be a DMR, a region can be required to include multiple CpG sites (e.g., 3, 4, 5, or more) that are hypo-methylated. Thus, a stretch of multiple CpG sites within a DMR can be selected for analysis by analysis in order to improve the signal-to-noise ratio and specificity of the DMR. (3) A DMR can be selected to be representative of the size of DNA molecules in the cell-free mixture. In plasma, the majority of DNA fragments are short, with most being shorter than 200 bp (1, 24, 25). For embodiments that determine the presence of erythroblast DNA molecules in plasma, a DMR can be defined within the representative size of plasma DNA molecules (i.e., 166 bp) (1). Variations of these criteria can be used in combination with the three criteria, e.g., different thresholds than 20% and 80% can be used to identify a CpG site as hypo-methylated. As discussed later, some results use selected CpG sites within three erythroblast-specific DMRs that are hypo-methylated in erythroblasts.

[0150] Using the above-defined criteria, we identified three erythroblast-specific DMRs in the whole genome. One DMR is within an intronic region of the ferrochelatase (FECH) gene on chromosome 18. In this region, the difference in methylation density between erythroblasts and other cell types is the largest among the three identified DMRs. The FECH gene encodes ferrochelatase, which is the enzyme responsible for the final step of heme biosynthesis (26). As shown, all four selected CpG sites within the erythroblast-specific DMR are hypo-methylated in erythroblasts but highly methylated in other blood cells and tissues. Figure 1

[0151] Figure 1 Methylation density of CpG sites within the promoter of the ferrochelatase (FECH) gene according to embodiments of the present application is shown. The FECH gene is located on chromosome 18, and the genomic coordinates of the CpG sites are shown on the X-axis. As shown, the methylation density of the CpG sites is within an intronic region of the FECH gene. Four CpG sites located within the region 110 bounded by two vertical dashed lines are all hypo-methylated in erythroblasts but highly methylated in other tissues or cell types. For illustrative purposes, the individual results for lung, heart, small intestine, colon, thymus, stomach, adrenal gland, esophagus, bladder, brain, ovary, and pancreas are not shown. Their average values are represented by "other tissues."

[0152] Because the CpG sites located within this region are hypo-methylated, Figure 1 ​Unmethylated sequences of all four CpG sites within the two dashed lines in the above diagram will be enriched for DNA originating from erythroblasts. Thus, the amount of hypomethylated sequences in a DNA sample will reflect the amount of DNA from erythroblasts.

[0153] An assay is developed to detect methylated or unmethylated DNA at the identified CpG sites. The higher the number of CpG within a plasma DNA molecule, the more specific the assay will be. Most plasma DNA molecules are less than 200 bp, with an average of 166 bp. Thus, the CpG sites can all be within 166 bp of each other, but also within 150, 140, 130, 120, 110, or 100 bp of each other. In other embodiments, only pairs of CpG sites can be within such distances of each other.

[0154] In other embodiments, a CpG site can be defined as hypomethylated in erythroblasts if the methylation density of the CpG site is less than 10% (or other threshold) in erythroblasts and more than 90% (or other threshold) in all other tissues and blood cells. A CpG site can be defined as hypermethylated in erythroblasts if the methylation density of the CpG site is more than 90% (or other threshold) in erythroblasts and less than 10% (or other threshold) in all other tissues and blood cells. In some embodiments, a DMR can have at least two CpG sites within 100 bp, all showing differential methylation in erythroblasts.

[0155] In one embodiment of a diagnostically useful DMR, all CpG sites within 100 bp (or some other length) can be required to show hypomethylation or hypermethylation in erythroblasts compared to all other tissues and blood cells. For example, multiple CpG sites can span 100 bp or less on a reference genome corresponding to a mammal. As another example, each CpG site can be within 100 bp of another CpG site. Thus, the CpG sites can span more than 100 bp.

[0156] In some embodiments, one or more differentially methylated regions can be identified in the following manner. A methylation index (e.g., density) of a plurality of sites can be obtained for each of a plurality of cell lineages, including a particular blood cell lineage and other cell lineages, e.g., as described in Figure 1The methylation index of the plurality of cell lineages can be compared to each other at each of the plurality of sites. Based on the comparison, one or more of the plurality of sites can be identified that each have a methylation index in a particular blood cell lineage that is below / above a first methylation threshold and a methylation index in each other cell lineage that is above / below a second methylation threshold. In this way, hypo- and / or hyper-methylation sites can be identified. An example of the first methylation threshold is 10%, 15%, or 20% for hypo-methylation sites, where an example of the second methylation threshold can be 80%, 85%, or 90%. The differentially methylated regions containing one or more sites can then be identified, for example, using the criteria described above.

[0157] B. Detecting methylated and unmethylated DNA sequences

[0158] To detect methylated and unmethylated DNA sequences at the erythroblast-specific DMR, two digital PCR assays can be developed: one targeting the unmethylated sequence and the other targeting the methylated sequence. In other embodiments, other methods can also be used to detect and / or quantify the methylated and unmethylated sequences of the DMR, such as methylation-aware sequencing (e.g., bisulfite sequencing or sequencing following biochemical or enzymatic processes that will differentially modify DNA based on its methylation status), real-time methylation-specific PCR, methylation-sensitive restriction enzyme assays, and microarray analysis. Thus, other types of assays can be used in addition to PCR assays.

[0159] In one example, the erythroblast DMR can be detected following bisulfite treatment. The methylation status of the CpG site can be determined based on the detection results (e.g., PCR signal). For the FECH gene, the following primers can be used to amplify the erythroblast DMR following bisulfite treatment for sequencing: 5'-TTTAGTTTATAGTTGAAGAGAATTTGATGG-3' and 5'-AAACCCAACCATACAACCTCTTAAT-3'.

[0160] In another example, to improve the specificity of the assay, two forward primers can be used that encompass both the methylated and unmethylated status of a particular CpG. The following lists such primer sets for the two digital PCR assays that specifically target the methylated and unmethylated sequences.

[0161] Primer / Probe Sequence Forward Primer-1 5'-TTGAAGAGAATTTGATGGTATGGGTA-3' Forward Primer-2 5'-TGAAGAGAATTTGATGGTACGGGTA-3' Reverse 5'-CTCAAATCTCTCTAATTTCC A AACAC A ]]> Fluorescent Probe 5'-FAM-TTG T GTGG T GTAGAGAG-MGB-3']]

[0162] Table 1: Assay to specifically detect unmethylated sequences.

[0163] Primer Sequence Forward 5'-TTGAAGAGAATTTGATGGTATGGGTA-3' 5'-TGAAGAGAATTTGATGGTACGGGTA-3' Reverse 5'-CAAATCTCTCTAATTTCC G AACAC G -3']] Fluorescent Probe 5'-VIC-TG C GTGG C GTAGAG-MGB-3']]

[0164] Table 2: Assay to specifically detect methylated sequences

[0165] The underlined nucleotides in the reverse primers and probes are differentially methylated cytosines at CpG sites. Due to the difference at the underlined nucleotides, the reverse primers and probes of the unmethylated and methylated assays bind specifically to the unmethylated and methylated sequences.

[0166] C. Confirmation using universally methylated and unmethylated DNA

[0167] The analysis of universally methylated and unmethylated DNA was performed to confirm the accuracy of the two assays.

[0168] The universally methylated sequence from CpGenome Human Methylated DNA (EMD Millipore) and the universally unmethylated sequence from EpiTect Unmethylated Human Control DNA (Qiagen) were used to confirm the specificity of the two digital PCR assays designed to detect and quantify methylated and non-methylated sequences at the erythroid-specific DMR. CpGenome Human Methylated DNA was purified from HCT116 DKO cells followed by enzymatic methylation of all CpG nucleotides using M. Sssl methyltransferase. The universally methylated and unmethylated DNA sequences were run on the same plate as the positive and negative controls. The cutoff for positive fluorescent signal was determined with reference controls. The number of methylated and unmethylated DNA sequences in each sample was calculated using combinatorial counting from the replicates followed by Poisson correction (4).

[0169] Figure 2A and 2B The analysis of universally methylated and unmethylated DNA using digital PCR assays designed to detect methylated and unmethylated DNA according to an embodiment of the application is shown. The vertical axis corresponds to the intensity of the relative fluorescent signal of the unmethylated sequences. The horizontal axis corresponds to the intensity of the relative fluorescent signal of the methylated sequences. The data was generated using DNA of known methylation or unmethylation. These assays were intended to demonstrate the specificity of the assays for methylated or unmethylated DNA.

[0170] For the analysis of the universally unmethylated DNA, an amplified signal (corresponding to a positive FAM signal in the sub-plot 205 of Fig. 2 Figure 2A of the blue dot 210) was detected using the assay for unmethylated DNA, where no blue dot 210 (sub-plot 255 of Fig. 2 Figure 2B was detected when using the assay for methylated DNA. For the analysis of the universally methylated DNA, an amplified signal (corresponding to a positive VIC signal in the sub-plot 255 of Fig. 2 Figure 2Bgreen dots 220 in sub-panel 250), where no green dots 220 were detected using the assay for unmethylated DNA Figure 2A Black dots in each panel represent droplets with no amplification signal. Thick vertical and horizontal lines within each of the four panels represent the threshold fluorescence signal for a positive result. These results confirm the specificity of both assays for the methylation of the erythroid-specific DMR and for unmethylated DNA.

[0171] To further assess the analytical sensitivity of the assays based on the FECH gene-associated DMR, samples with unmethylated sequences were serially diluted at a specific fraction concentration (i.e., percentage of unmethylated sequences among all (unmethylated and methylated) sequences at the FECH gene-associated DMR). There were a total of 1,000 molecules per reaction. Unmethylated sequences were detected down to 0.1% of the total amount of methylated and unmethylated sequences (see Table 3).

[0172]

[0173] Table 3: Sensitivity assessment of the assay targeting the FECH gene-associated DMR, measured as concentration (percentage of unmethylated sequences) at different input concentrations of unmethylated sequences.

[0174] In addition, to assess potential variations (e.g., from pipetting), we repeated the measurement of the percentage of unmethylated sequences in an artificial mixture of methylated and unmethylated sequences at a specific fraction concentration (unmethylated sequences % = 30%) in 20 separate reactions. We used a total of 500 methylated and unmethylated molecules per reaction. This number is comparable to the number of methylated and unmethylated molecules we observed in the total number of molecules in the digital PCR analysis of plasma DNA samples. We observed a mean of 30.4% and a standard deviation of 1.7% for the percentage of unmethylated sequences in the 20 replicates. The coefficient of variation within the assay was calculated to be 5.7%.

[0175] III. Specificity and sensitivity of the assays for different samples

[0176] To confirm the tissue specificity of the digital PCR analysis of the FECH gene-associated DMR targeting erythrocyte DNA, we tested the digital PCR analysis in various samples with different amounts of erythroblasts, as measured using techniques other than these digital PCR analyses. The amount of unmethylated DNA sequences detected by the digital PCR analysis should reflect the amount of erythrocyte DNA. Similarly, the amount of methylated sequences should reflect DNA from other tissues or cell types. Thus, we defined the percentage of erythrocyte DNA in a biological sample (E%) as the percentage of unmethylated sequences among all detected (unmethylated and methylated) sequences at the erythroblast-specific DMR. Thus, blood samples were analyzed using assays specific for methylated and unmethylated sequences of the DMR region to determine the correlation between the percentage of unmethylated sequences, Unmeth% (also referred to as E%), and the presence of DNA from erythroblasts. Unmeth% (E%) is an example of a methylation level.

[0177] The percentage of erythrocyte DNA (E%) was calculated as:

[0178]

[0179] Since the difference in methylation density between erythrocytes and other cell types is greatest for the DMR within the FECH gene, we first performed E% analysis based on this marker site to demonstrate our hypothesis. Subsequently, we analyzed E% based on two additional erythroblast-specific DMRs in a subset of samples to validate the E% results from the FECH gene-associated DMR. The E% results based on the DMR within the FECH gene will be denoted as E%(FECH). Other percentages or ratios can also be used, such as the percentage of methylated sequences, or just the ratio of methylated sequences to unmethylated sequences, where either value can be in the numerator and denominator of the ratio.

[0180] Specifically, digital PCR was used to determine the number of methylated and unmethylated DNA sequences in each sample at four CpG sites on the FECH gene. Figure 1 Subsequently, the percentage of unmethylated DNA in the sample (Unmeth% / E%) was calculated. In one embodiment, all four CpG sites were unmethylated for a DNA fragment to be considered unmethylated.

[0181] Two cases were used to test the ability of the assay to quantify erythroblasts. One case was cord blood versus adult blood, since these two types of samples differ in the amount of erythroblasts. Also, for the other case, a subject with beta-thalassemia major has a significant number of erythroblasts in his blood.

[0182] A. Erythroid-enriched samples compared to buffy coat of healthy subjects

[0183] The number of erythroblasts in adult blood is extremely low. Umbilical cord blood has a higher number of erythroblasts. Therefore, the E% of the four CpG sites in umbilical cord blood should be much higher than in healthy patients. Thus, to confirm the tissue specificity of the digital PCR analysis of the FECH gene-related DMR targeting erythrocyte DNA, we tested the digital PCR analysis in samples including DNA extracted from 12 different normal tissue types and erythroid-enriched samples. We included 4 samples from different individuals for each tissue type. Erythroid-enriched samples were prepared from umbilical cord blood for analysis.

[0184] Specifically, to confirm the relationship between methylation density and E% at the DMR, venous blood samples were collected from 21 healthy subjects and 30 pregnant women (10 in early pregnancy, 10 in mid-pregnancy, and 10 in late pregnancy). The blood samples were centrifuged at 3,000g for 10 minutes to separate the plasma and blood cells. The buffy coat was collected after centrifugation. The plasma sample was collected and centrifuged again at 30,000g to remove the remaining blood cells.

[0185] As for 12 different normal tissue types, we included 4 samples from different individuals for each tissue type. As shown in Table 4, the median E% (FECH) from all tissue DNA was low (median range: 0.00% to 2.63%).

[0186] Tissue Median E% Liver 0.12％ Lung 1.33％ Esophagus 2.63％ Stomach 2.58％ Small Intestine 2.33％ Colon 1.51％ Pancreas 0.12％ Adrenal Gland 0.00％ Bladder 1.20％ Heart 0.82％ Brain 1.94％ Placenta 0.10％

[0187] Table 4. Table showing the median percentage of erythrocyte DNA (E% (FECH)) in 4 groups of 12 tissue types, where each tissue sample was obtained from different individuals.

[0188] The experimental procedure for enrichment from umbilical cord blood by flow cytometry and cell sorting and subsequent DNA extraction is described below. After delivery, 1-3 mL of umbilical cord blood was collected from each of eight pregnant women. After density gradient centrifugation using the Ficoll-Paque PLUS kit (GE Healthcare), mononuclear cells were isolated from the umbilical cord blood samples. After collecting the mononuclear cells, 1 x 10 8A 1 mL mixture of anti-CD235a (Glycophorin A) conjugated with fluorescein isothiocyanate (FITC) and anti-CD71 antibody (Miltenyi Biotec) conjugated with phycoerythrin (PE) was incubated together at 1 : 10 dilution in phosphate-buffered saline for 30 minutes at 4°C in the dark. Subsequent sorting and analysis of CD235a+CD71+ cells was performed using a BD FACSAria Fusion cell sorter (BD Biosciences). Since CD235a and CD71 are specifically present in erythroblasts, CD235a+CD71+ cells will be enriched for erythroblasts (Bianchi et al. Prenatal Diagnosis 1993; 13: 293-300).

[0189] Due to the low number of cells obtained from each individual, cells from eight individuals were pooled for downstream analysis. Both antibodies are specific for erythroblasts and attach to the surface of erythroblasts. Both antibodies are conjugated with FITC and phycoerythrin, respectively. Both substances bind to magnetic beads and the beads can be sorted using a cell sorter. Thus, Ab-labeled erythroblasts can be captured. Erythroblasts were enriched from 8 cord blood samples using flow cytometry and cell sorting with anti-CD71 (transferrin receptor) and anti-CD235a (Glycophorin A) antibodies (see Supplementary Materials and Methods) and then pooled. DNA was extracted from the pooled sample.

[0190] The E% (FECH) of DNA from the pooled cord blood sample was 67% at the four CpG sites tested in the analysis of CD235a+CD71+ cells (mostly erythroblasts). The E% (FECH) of buffy coat DNA from 20 healthy subjects, who have an undetectable number of erythroblasts in their peripheral blood, was 2.2% (interquartile range: 1.2-3.1%) median E% of buffy coat DNA. The observation of a low proportion of erythroblast-specific unmethylated sequences in the buffy coat of healthy subjects is in line with the fact that mature RBCs do not have a nucleus. Since CD235a and CD71 are cell surface markers specific for erythroblasts (Bianchi et al. Prenatal Diagnosis 1993; 13: 293-300), the high E% (FECH) in cells enriched for CD235a and CD71 shows that the analysis of unmethylated DNA at erythroblast-specific DMRs is able to detect DNA of erythroblast origin. Thus, this high E% of samples enriched for erythroblasts, and the low E% results of DNA from other tissue types and buffy coat DNA of healthy subjects show that the digital PCR analysis of unmethylated FECH sequences is specific for DNA of erythroblast origin.

[0191] B. For patients with β-thalassemia major

[0192] In patients with β-thalassemia major, the bone marrow attempts to make large numbers of red blood cells (RBCs). However, the production of hemoglobin is defective. As a result, many RBCs do not contain enough hemoglobin and contain large excesses of alpha globin chains. These defective RBCs will be removed from the bone marrow and will never become mature RBCs. There are two types of globin chains: alpha and beta. A hemoglobin molecule requires two alpha chains and two beta chains. If beta chains are not produced, excess alpha chains will clump together and cannot form functional hemoglobin.

[0193] In patients with β-thalassemia major, the increased but ineffective erythropoiesis results in a decreased production of mature RBCs (Schrier et al. Current Opinion in Hematology 2002; 9: 123-6). This is accompanied by compensatory extramedullary hematopoiesis and the presence of nucleated red blood cells in the circulation. As described below, patients with β-thalassemia major will have more nucleated red blood cells than healthy patients. The number of nucleated RBCs in the peripheral blood can be counted on a blood smear and expressed as the number of nucleated RBCs per 100 white blood cells (WBCs).

[0194] Since patients with thalassemia major typically have a higher number of erythroblasts in the peripheral blood than healthy individuals due to the failed erythropoiesis (27), such patients also provide a good mechanism to test the specificity and sensitivity of the assay. Thus, we tested the sensitivity of our digital PCR assay in buffy coat DNA from fifteen patients with β-thalassemia major. As measured by an automated hematology analyzer (UniCel DxH 800 Coulter Cell Analysis System, Beckman Coulter) and confirmed by manual counting, all of these patients had a detectable number of erythroblasts in the peripheral blood.

[0195] Figure 3A is a graph showing the correlation between E% (FECH) and the number of nucleated RBCs (erythroblasts) in blood cells according to an embodiment of the application. E% was measured by digital PCR analysis targeting the FECH gene-associated DMR. As indicated by the axes, the graph shows the correlation between the percentage of red blood cell DNA sequences in the buffy coat DNA (E% (FECH)) and the percentage of erythroblasts in all peripheral white blood cells, as measured using an automated hematology analyzer.

[0196] As Figure 3AAs shown, the E% of the buffy coat DNA (FECH) correlates well with the percentage of erythroblasts in the peripheral blood leukocytes measured by a hematology analyzer (r = 0.94, P < 0.0001, Pearson correlation). The good linear relationship between the E% in the buffy coat and the erythroblast count in patients with thalassemia shows that the digital PCR analysis provides a good quantitative measure of the amount of erythrocyte DNA in the sample, as the erythroblasts are unmethylated at the DMR and other blood cells are methylated. Thus, the higher the proportion of erythroblasts in a blood sample, the higher the E% will be. The purpose of this experiment was to demonstrate that the analysis can be used to reflect the amount of DNA in a sample that is derived from erythroblasts. These results further support that the E% of the FECH gene reflects the proportion of DNA derived from erythroblasts.

[0197] This correlation will exist for other patients as well. However, because the number of erythroblasts can be higher for patients with beta-severe thalassemia, their samples provide a good test for identifying this correlation. From Figure 3A As can be seen, the patients have a wide range of E% and a large number of erythroblasts, providing a good mechanism for testing the correlation.

[0198] C. Methods of determining the amount of cell DNA of a particular cell lineage

[0199] In some embodiments, the amount of unmethylated or methylated DNA fragments in a cell-free mixture (e.g., a plasma or serum sample) can be used to determine the number of cells (or other amount of DNA) of a particular cell lineage when the amount is counted at one or more DMRs specific to the particular cell lineage. As Figure 3A As shown in Example 2, the percentage of DNA fragments unmethylated at the FECH DMR correlates with the number of erythroblasts in a blood sample. Absolute concentrations can also be used. For highly methylated DMRs, the amount of methylated DNA fragments (e.g., percentage or absolute concentration) can be used. As described herein, various cell lineages can be used.

[0200] To determine the number of cells, a calibration function can be used. In Figure 3A In the example of a line fitted to the data points, the calibration function can be stored by its functional parameters (e.g., the slope and y-intercept of a line, or more parameters for other functions), or by a set of data points from which the curve fit can be obtained. When determining the number of erythroblasts, the data points (e.g., referred to as calibration data points) can have known values for the amount of DNA (e.g., the number of cells) of the cell lineage, as can be determined by another technique.

[0201] Thus, a method can determine the amount of DNA from a particular cell lineage in a blood sample. As described herein, a number of methylated or unmethylated sequences of one or more DMRs can be determined from the analysis. The level of methylation can be determined and compared to the calibration values of the calibration function. For example, the level of methylation can be compared to the line (or other calibration function) to determine the intersection of the function and the level of methylation, and thus the corresponding amount of DNA (e.g., the value on the horizontal axis of the graph of Figure 3A In other embodiments, the level of methylation can be compared to individual calibration data points, e.g., that have a level of methylation close to the measured level of methylation of the sample.

[0202] Figure 3B A flowchart of a method 300 of determining the amount of cells of a particular cell lineage in a biological sample by analyzing cell-free DNA is illustrated according to embodiments of the present invention. The method 300 can use measurements such as those shown in Figure 3A The portions of the method 300 can be performed manually, while other portions can be performed by a computer system. In one embodiment, a system can perform all of the steps. For example, a system can include robotic elements (e.g., to obtain the sample and perform the analysis), a detection system for detecting signals from the analysis, and a computer system for analyzing the signals. Instructions for controlling this system can be stored in one or more computer-readable media, such as the configuration logic of a field programmable gate array (FPGA), a flash memory, and / or a hard drive. Figure 27 Such a system is shown.

[0203] At block 310, a cell-free mixture of a biological sample is obtained. The biological sample can be a blood sample, but can also be other samples that include cell-free DNA, as described herein. Examples of cell-free mixtures include plasma or serum. The cell-free mixture can include cell-free DNA from multiple cell lineages.

[0204] At block 320, DNA fragments in the cell-free mixture are contacted with an assay corresponding to one or more differentially methylated regions. Each of the one or more differentially methylated regions is specific to a particular cell lineage (e.g., a particular blood cell lineage, such as erythroblasts) by being hypomethylated or hypermethylated relative to other cell lineages.

[0205] In various embodiments, the analysis can involve PCR or sequencing. The contacting with the DNA fragments can involve flow cells, droplets, beads, or other mechanisms to provide for analysis of the interaction with the DNA fragments. Examples of such analysis include whole genome bisulfite sequencing, targeted bisulfite sequencing (by hybrid capture or amplicon sequencing), other methylation-aware sequencing (e.g., Pacific Biosciences' single molecule, real-time (SMRT) DNA sequencing), real-time methylation-specific PCR, and digital PCR. Other examples of analysis that can be used in the method 300 are described herein (e.g., in Section XII). Although examples are described in terms of methylation, the same analysis can be used to detect unmethylation. Figure 3A For erythroblasts, other cell lineages can be used, including other blood cell lineages.

[0206] At block 330, a first number of methylated or unmethylated DNA fragments in the cell-free mixture at the one or more differentially methylated regions is detected based on signals obtained from the analysis. The analysis can provide various signals, such as optical or electrical signals. The signals can provide a specific signal for each DNA fragment, or an aggregate signal indicative of the total number of DNA fragments with a methylation marker (e.g., as in real-time PCR).

[0207] In one embodiment, sequencing can be used to obtain sequence reads for the DNA fragments, and the DNA fragments can be aligned to a reference genome. If a DNA fragment aligns to one of the DMRs, a counter can be incremented. Given that the signal is from a specific methylation, unmethylation analysis, it can be assumed that the DNA fragment has the methylation marker. In another embodiment, reads from PCR (e.g., optical signals from positive wells) can be used to increment such a counter.

[0208] At block 340, the first number is used to determine a first methylation level. The first methylation level can be normalized or be an absolute concentration, such as per volume of the biological sample. Figure 25 Examples of absolute concentrations are provided in Section III.

[0209] For normalized values, the first and total number of DNA fragments in the cell-free mixture at the one or more differentially methylated regions can be used to determine a methylation level. As described above, the methylation level can be a percentage of unmethylated DNA fragments. In other embodiments, the percentage can be a percentage of methylated DNA fragments, which would have an inverse relationship relative to the above example for erythroblasts. In various embodiments, the methylation level can be determined using a percentage of all sites in the DMR, by an average of individual percentages for each site, or a weighted average for each site.

[0210] At block 350, one or more calibration data points are obtained. Each calibration data point can specify (1) a cell amount of a particular blood cell lineage and (2) a calibration methylation level. The one or more calibration data points are determined from a plurality of calibration samples.

[0211] The cell amount can be specified as a particular amount (e.g., a number or concentration) or a range of amounts. The calibration data points can be determined from calibration samples having known cell amounts, which can be measured by various techniques described herein. At least some of the calibration samples will have different amounts of cells, but some of the calibration samples can have the same amount of cells.

[0212] In various embodiments, the one or more calibration points can be defined as a discrete point, a set of discrete points, as a function, as a discrete point and a function, or any other combination of discrete or continuous sets of values. As an example, a calibration data point can be determined from a calibration methylation level of a sample having a particular cell amount of a particular lineage.

[0213] In one embodiment, measurements of the same methylation level from multiple samples of the same cell amount can be combined to determine a calibration data point for a particular cell amount. For example, an average of methylation levels from samples of the same cell amount can be obtained to determine a particular calibration data point (or to provide a range corresponding to the calibration data point). In another embodiment, multiple data points having the same calibration methylation level can be used to determine an average amount of cells.

[0214] In one implementation, methylation levels are measured for a number of calibration samples. A calibration value of the methylation level is determined for each calibration sample, where the methylation level can be plotted against the known cell amount of the sample (e.g., as shown in FIG. 3). A function can then be fit to the data points of the plot, where the function fit defines a calibration data point for determining a cell amount of a new sample. Figure 3A

[0215] At block 360, the first methylation level is compared to the calibration methylation level at the at least one calibration data point. The comparison can be made in various ways. For example, the comparison can be whether the first methylation level is higher or lower than the calibration methylation level. The comparison can involve a comparison to a calibration curve (composed of the calibration data points), and thus the comparison can identify a point on the curve having the first methylation level. For example, a calculated value X of the first methylation level can be used as an input to a function F(X), where F is a calibration function (curve). The output of F(X) is the cell amount. An error range can be provided, which can be different for each X value, providing a range of values as the output of F(X).

[0216] ​At block 370, the amount of cells of a particular cell lineage in the biological sample is estimated based on the comparison. In one embodiment, it can be determined whether the first methylation level is above or below the threshold calibration methylation level, and thus whether the amount of cells of the sample is above or below the amount of cells corresponding to the threshold calibration methylation level. For example, if the calculated biological first methylation level X1 is above the calibration methylation level X C , then it can be determined that the amount of cells N1 of the biological sample is above the amount of cells N C corresponding to X C . This relationship of above and below can depend on how the parameters are defined. In this embodiment, only one calibration data point can be required.

[0217] In another embodiment, the comparison is done by inputting the first methylation level into a calibration function. The calibration function can effectively compare the first methylation level to the calibration methylation level by identifying the point on the curve corresponding to the first methylation level. The estimated amount of cells is then provided as the output value of the calibration function.

[0218] IV. Sources of cell-free DNA of erythroblasts in plasma

[0219] Using the relationship established between Unmeth% and DNA of erythrocyte origin, the Unmeth% of plasma can be used to quantify DNA of erythrocyte origin in plasma. The Unmeth% in plasma was determined using the analysis described above. Differences in Unmeth% were observed between buffy coat and plasma. Analysis showed that cell-free DNA of erythroblasts in plasma originated from erythropoiesis in the bone marrow, rather than from erythroblasts in the bloodstream.

[0220] Having confirmed that the Unmeth% determined by the two digital PCR assays accurately reflects the amount of DNA of erythrocyte origin in the sample, we further compared the proportion of DNA of erythrocyte origin in the buffy coat and plasma of healthy control subjects and pregnant women.

[0221] Figure 4 Unmeth% in the buffy coat and plasma of healthy non-pregnant subjects and pregnant women at different gestational ages according to embodiments of the application is shown. For each group of subjects, the plasma samples had significantly higher Unmeth% than the buffy coat (Wilcoxon sign-rank test for each pair of comparisons between plasma and buffy coat, P < 0.01).

[0222] Figure 4The results show that the amount of erythroblast-derived DNA is lower in blood cells as expected due to the lower number of nucleated RBCs. Surprisingly, the amount of erythroblast-derived DNA is higher in plasma. If the erythroblast-derived DNA in plasma was derived from blood cells, one would expect similar amounts. Thus, this data shows that the source of erythroblast-derived DNA in plasma is from erythropoiesis in the bone marrow.

[0223] Figure 5 Figure 6 is a graph showing the lack of correlation between Unmeth% in buffy coat and plasma. No significant correlation was observed between Unmeth% of buffy coat and plasma DNA (R 2 = 0.002, P = 0.99, Pearson correlation). The lack of correlation can be seen for all subjects, including non-pregnant subjects, early pregnant women, mid-pregnant women, and late pregnant women. As with the results in Figure 5, this is surprising if the source of erythroblast-derived DNA is from blood cells in the bloodstream, one would expect these two to be correlated. Figure 4

[0224] The observation that plasma DNA has a much higher Unmeth% than buffy coat and that there is a lack of correlation between Unmeth% of plasma and buffy coat suggests that circulating cell-free DNA carrying the erythroblast methylation signature can originate in the bone marrow during the process of erythropoiesis, rather than from circulating blood cells. Thus, cell-free plasma DNA with the erythroblast methylation signature is generated in the bone marrow, as opposed to being generated from nucleated RBCs in the bloodstream, because the number of nucleated RBCs in the bloodstream is extremely low in healthy subjects and in pregnant women. And, because the contribution of white blood cells (WBCs) to the erythroblast methylation signature is extremely low, this contribution does not provide a measurable dependency of cell-free plasma DNA with the erythroblast methylation signature.

[0225] V. Methylation levels as a measure of erythropoietic activity

[0226] ​Based on the above observations, we determined that Unmeth% at the Erythroid DMR would reflect the activity of erythropoiesis in the bone marrow. High Unmeth% would indicate high erythropoietic activity. In other words, analysis of erythroid DNA in plasma / serum would be used as a liquid biopsy of the bone marrow. This analysis would be particularly useful in the study of anemia, for example, to determine whether anemia is due to reduced erythropoiesis (e.g., aplastic anemia), defective erythropoiesis (e.g., failure of mature RBC production in thalassemia), or increased RBC consumption (e.g., blood loss and hemolytic anemia). To this end, we recruited 35 healthy subjects and 75 anemic patients with different etiologies. Peripheral blood sample collection and processing, plasma and buffy coat DNA extraction, and bisulfite conversion of DNA were performed. Further details on the methods are described in Section XII.

[0227] A. Measurement of cell-free erythroid DNA in plasma of healthy subjects

[0228] After confirming the specificity of our assays, we used these assays to analyze the plasma of healthy subjects. We analyzed the E% (FECH) in the plasma of 35 healthy subjects, including 20 subjects of the same group that also provided a buffy coat sample. The median E% (FECH) of plasma DNA was 30.1% (interquartile range: 23.8-34.8%). This indicates that erythroid DNA constitutes a significant proportion of the circulating DNA pool in the plasma of healthy individuals. To determine the origin of the plasma erythroid DNA, we compared the corresponding E% (FECH) results in the plasma and buffy coat of 20 healthy subjects.

[0229] Figure 6A and 6B shows the percentage of erythroid DNA (E% (FECH)) in healthy subjects. Figure 6A shows the E% of buffy coat DNA and plasma DNA of healthy subjects, where the values of E% are higher in plasma (cell-free fraction) than in buffy coat (cellular fraction). The median E% of plasma DNA (median: 26.7%, interquartile range: 23.7-30.4%) was significantly higher than the values in the paired buffy coat DNA (median: 2.2%, interquartile range: 1.2-3.1%) (P < 0.0001, Wilcoxon signed-rank test).

[0230] Figure 6B shows the lack of correlation between E% in buffy coat DNA and plasma DNA of the corresponding healthy subjects. There was a lack of correlation between the paired E% (FECH) results in plasma DNA and buffy coat DNA (r = 0.002, P = 0.99, Pearson correlation). In Figure 6A and Figure 6BTwo of the findings in the present study show that circulating erythroid DNA is unlikely to originate mainly from circulating erythroblasts in the peripheral blood.

[0231] Figure 7 The E% (FECH) results of plasma DNA show a lack of correlation with the age of healthy subjects. The graph shows that the E% (FECH) results are not related to the age of the subjects (r = 0.21, p = 0.23, Pearson correlation).

[0232] B. Distinction between patients with β-thalassemia major and aplastic anemia

[0233] After determining that erythroid DNA in plasma does not originate mainly from intact erythroblasts in circulation, we propose that these DNA molecules are more likely to be released from the bone marrow during erythropoiesis. We infer that the quantification of erythroid DNA in plasma can provide information about the erythropoietic activity in the bone marrow.

[0234] To confirm the ability to measure erythropoietic activity in the bone marrow using plasma, patients with β-thalassemia major and aplastic anemia were recruited from the Department of Medicine of the Prince of Wales Hospital in Hong Kong, China. Venous blood samples were collected before transfusion. The Unmeth% of plasma DNA was determined for each patient by digital PCR. These results were correlated with the hemoglobin level. The hemoglobin level can be measured by techniques known to those skilled in the art, for example by photometric techniques performed on an automated blood cell counter. The hemoglobin level can be measured from the RBC fraction, for example after centrifugation.

[0235] The patients of these two groups (β-thalassemia major and aplastic anemia) represent two different spectra of erythropoietic activity. In patients with β-thalassemia major, erythropoiesis has a high activity. However, the production of mature RBCs is reduced due to the defect in the production of functional β-globin chains. In patients with aplastic anemia, erythropoiesis is reduced, leading to a reduction in the production of RBCs.

[0236] Figure 8Figure is a plot of Unmeth% for hemoglobin concentration in patients with aplastic anemia, beta-thalassemia major, and healthy control subjects according to embodiments of the application. In beta-thalassemia patients, hemoglobin concentration is decreased, but Unmeth% is significantly increased compared to healthy control subjects (P < 0.01, Mann- Whitney rank-sum test). In fact, 10 of 11 (89%) beta-thalassemia patients have Unmeth% values higher than all healthy control subjects. This observation is consistent with increased, but defective, erythropoiesis in these patients.

[0237] In contrast, for six patients with aplastic anemia who are on regular transfusions, their Unmeth% is lower than all healthy control subjects. This observation is consistent with decreased erythropoiesis in these patients.

[0238] For three aplastic anemia patients in clinical remission, their hemoglobin levels are normal and they do not require regular transfusions. Their Unmeth% values are not significantly different from those of healthy control subjects (P = 0.53, Mann- Whitney rank-sum test). Thus, quantification of erythroid-specific DNA in plasma will be useful for monitoring patients with bone marrow dysfunction, for example, to determine if aplastic anemia is in remission. Furthermore, quantification of erythroid-specific DNA can be used to guide therapy. For example, patients with aplastic anemia who are not in remission can be treated with regular transfusions.

[0239] Thus, Unmeth% is higher in thalassemia patients and lower in aplastic anemia patients. For thalassemia, the bone marrow is active because the patient is anemic and the bone marrow wants to produce more RBCs for circulation. Thus, the rate of erythropoiesis is higher than in healthy subjects who are not anemic. For patients with aplastic anemia, anemia is due to decreased RBC production. In summary, these results indicate that analysis of the erythroid-specific methylation profile will be useful to reflect the activity of erythropoiesis in the bone marrow.

[0240] Patients can be diagnosed by a combination of hemoglobin measurements and Unmeth%. For example, patients with hemoglobin lower than 11.8 and E% higher than 50 can be classified as having beta-thalassemia. However, patients with hemoglobin lower than 11.8 and E% lower than 25 can be classified as having aplastic anemia.

[0241] C. Iron Deficiency Anemia and Treatment

[0242] Anemia can be due to a deficiency of nutrients (e.g., iron, B12, folate, etc.), blood loss (e.g., due to heavy menstrual periods or bleeding from the gastrointestinal tract), or chronic illness (e.g., cancer, inflammatory bowel disease).

[0243] Figure 9 Figure of plasma Unmeth% in patients with iron deficiency (Fe) anemia and acute blood loss. Three patients with iron deficiency anemia and one patient with acute gastrointestinal blood loss were studied. In two of the iron deficiency patients, anemia was due to heavy menstrual periods. For one patient, a blood sample was collected prior to starting iron supplementation. For the other patient, a blood sample was collected 1 week after starting iron supplementation therapy. The third patient with iron deficiency anemia had inflammatory bowel disease and a blood sample was collected prior to starting iron supplementation.

[0244] Plasma Unmeth% was determined for each patient and compared to the values for healthy control subjects. An increase in plasma Unmeth% was observed in the patient with acute gastrointestinal blood loss. For the two iron deficiency patients for which samples were collected prior to starting iron supplementation therapy, their plasma Unmeth% values did not increase compared to healthy subjects, despite their low hemoglobin levels. For the Fe deficiency patient for which a sample was collected 1 week after starting iron supplementation, an increase in plasma Unmeth% was observed.

[0245] These results show that plasma Unmeth% reflects erythropoietic activity in response to therapy. For example, treatment with iron supplementation shows increased erythropoietic activity. Furthermore, these results show that the response of Unmeth% will be faster than the rise in hemoglobin levels. The use of Unmeth% can be an early indicator of whether the therapy is effective and therefore whether the therapy should continue or stop. Thus, Unmeth% can provide guidance to predict the response to anemia therapy (e.g., iron therapy) before changes in hemoglobin levels are observed.

[0246] In some embodiments, plasma Unmeth% can be used to reflect the response to anemia therapy. For example, in patients with iron deficiency anemia, the response to oral iron supplementation can differ for different subjects due to changes in iron absorption through the gut. In this case, insufficient increase in plasma Unmeth% after starting oral iron supplementation can be used to indicate the need for intravenous iron therapy.

[0247] D. Distinction of various anemia conditions

[0248] We recruited anemic patients with aplastic anemia (AA), chronic renal failure (CRF), iron deficiency anemia due to chronic blood loss, and β-thalassemia major. Different disease entities were recruited to represent the two ends of the spectrum of erythropoietic activity in the bone marrow.

[0249] Figure 10Figure 1 shows the relationship between the percentage of cell-free DNA in plasma (E%(FECH)) and hemoglobin levels in patients with aplastic anemia, chronic renal failure (CRF), beta-thalassemia major, iron deficiency anemia, and healthy subjects according to embodiments of the present application. E%(FECH) of plasma DNA was plotted against hemoglobin levels for anemic patients and 35 healthy controls. The horizontal dotted line represents the median E% in healthy subjects. The vertical line corresponds to the cutoff value of hemoglobin levels measured between subjects with anemia and subjects without anemia (11.5, as shown).

[0250] We analyzed E% of plasma DNA in 13 AA patients who met the diagnostic criteria (28) but failed to respond to immunosuppressive therapy. The median E% of plasma DNA in the AA group was 12.4% (interquartile range: 7.5-13.7%), which was significantly lower than the median E% in healthy controls (P < 0.0001, Mann-Whitney rank sum test; Figure 10 ). Similarly, the median E% result for 18 CRF patients requiring dialysis was 16.8% (interquartile range: 12.2-21.0%), which was also significantly lower than the median E% in healthy controls (P < 0.0001, Mann-Whitney rank sum test; Figure 10 ). These findings are consistent with the pathophysiology of reduced erythropoietic activity in AA (28, 29) and CRF patients (30).

[0251] For patients with beta-thalassemia major, the bone marrow is attempting to compensate for the hypoxic stress with increased but failing erythropoiesis (31). The median E% of plasma DNA in 17 recruited beta-thalassemia major patients was 65.3% (interquartile range: 60.1-78.9%), which was significantly higher than the median E% in healthy controls (P < 0.0001, Mann-Whitney rank sum test; Figure 10 ).

[0252] For subjects with iron deficiency anemia, we recruited 11 patients (transferrin saturation < 16% or serum ferritin level < 30 ng / ml) suffering from iron deficiency anemia due to menorrhagia or peptic ulcer disease. The median E% of their plasma DNA was 37.8% (interquartile range: 31.8-43.0%), which was significantly higher than the median E% in healthy controls (P = 0.002, Mann-Whitney rank sum test; Figure 10 ). This finding can be explained by a compensatory increase in bone marrow erythropoietic activity in response to chronic blood loss (32).

[0253] Thus, a patient can be diagnosed by a combination of hemoglobin measurements and E% values. For example, a patient with a hemoglobin level below 11.5 (or other value) and an E% above 50 can be classified as having anemia with increased erythropoietic activity, such as beta-thalassemia. However, a patient with a hemoglobin level below 11.5 and an E% below 50 and above 28 can be classified as having anemia with moderate erythropoietic activity, such as iron deficiency anemia. However, a patient with a hemoglobin level below 11.5 and an E% below 28 can be classified as having anemia with decreased erythropoietic activity, such as aplastic anemia or chronic renal failure.

[0254] In some embodiments, to determine the classification of the blood disorder, the hemoglobin level of the blood sample can be measured. The hemoglobin level can be compared to a hemoglobin threshold value (e.g., 11.5). Thus, in addition to the methylation level, the classification of the blood disorder can be further based on the comparison of the hemoglobin level to the hemoglobin threshold value.

[0255] An overview of the E% (FECH), red blood cell, and reticulocyte parameters for subjects is shown in Tables 5 and 6, respectively, and Figure 11A and Figure 11B

[0256]

[0257] Table 5. Table summarizes the median percentage of red blood cell DNA in plasma DNA (E% (FECH)) for healthy subjects and anemic patients.

[0258] In Table 6 below, the median and interquartile range (in parentheses) are shown. The following abbreviations are used: hematocrit is Hct, mean corpuscular volume is MCV, mean corpuscular hemoglobin is MCH, mean corpuscular hemoglobin concentration is MCHC, and red blood cell distribution width is RDW.

[0259]

[0260] Table 6. Red blood cell (RBC) parameters for recruited healthy controls and anemic patients.

[0261] Figure 11A and 11B ​The relationship between reticulocyte count / index and hemoglobin level in anemic patients with aplastic anemia, chronic renal failure (CRF), beta-thalassemia major, and iron deficiency anemia is shown. The reticulocyte index is calculated as follows: reticulocyte count x hematocrit / normal hematocrit. As can be seen, the amount of reticulocytes (immature RBCs) in the blood does not provide a reliable distinction between the different conditions. These results show that reticulocyte count and reticulocyte index are not able to distinguish between anemias of different etiology, e.g., between thalassemia and aplastic anemia.

[0262] E. Myelodysplastic syndrome and polycythemia vera

[0263] Figure 12 is a plot of plasma Unmeth% in patients with myelodysplastic syndrome and polycythemia vera. In patients with myelodysplastic syndrome, increased plasma Unmeth% is observed at reduced hemoglobin levels. Increased plasma Unmeth% is also observed in patients with polycythemia vera. These results show that detection and quantification of erythroid DNA methylation markers in plasma are suitable for detecting and monitoring abnormal proliferation or dysplasia of the bone marrow involving myeloblast cells.

[0264] Thus, as can be seen, these two blood conditions also show higher erythroid cell-free DNA, allowing detection of blood conditions. In some embodiments, accurate diagnosis can be based on histological examination of bone marrow biopsies. Thus, a bone marrow biopsy can be performed in response to detection of high Unmeth%. Similarly, a bone marrow biopsy can be performed in response to detection of low Unmeth% in the presence of anemia but absence of nutritional deficiency (e.g., iron deficiency, vitamin B12 deficiency, or folate deficiency). These substrates for bone marrow biopsy can reduce the number of these biopsies while still allowing monitoring of the health of the bone marrow. Thus, Unmeth% will be more suitable for monitoring treatment response.

[0265] F. Other distinctions of anemia

[0266] Distinctions between other conditions are also possible.

[0267] 1. Aplastic anemia (AA) and myelodysplastic syndrome (MDS)

[0268] Both aplastic anemia and MDS are bone marrow failure conditions. Despite their similar clinical features of pancytopenia, these two disease entities have different pathophysiological mechanisms. In AA, there is cytopenic bone marrow without dysplastic features. In MDS, there is usually cytomegalic bone marrow and dysplasia involving one or more lineages (33), but cytopenic MDS is also recognized.

[0269] Figure 13A The percentage of erythrocyte DNA in plasma (E% (FECH)) in patients with aplastic anemia (AA) and myelodysplastic syndrome (MDS) according to embodiments of the application is shown. The median E% of plasma DNA from 8 MDS patients was 50.3% (range: 37.4-60.8%). Two had MDS with single lineage dysplasia, 4 had MDS with multilineage dysplasia, and 2 had MDS with excess blasts (34). All of their previous bone marrow biopsies showed hypercellularity of erythrocyte cells. The median E% in MDS patients was significantly higher than the median E% in 13 recruited AA patients (P < 0.0001, Mann-Whitney rank sum test; Figure 13A ). The median E% result in MDS patients is consistent with the bone marrow biopsy findings and pathophysiology of ineffective erythropoiesis in MDS.

[0270] Thus, MDS can be distinguished from aplastic anemia using E% or other methylation levels. For example, a cutoff of 30 can be used to classify a sample as corresponding to aplastic anemia or MDS.

[0271] 2. Treatment responsive and non-responsive groups of AA

[0272] Figure 13B The percentage of erythrocyte DNA in plasma (E% (FECH)) between the treatment responsive and non-responsive groups in aplastic anemia according to embodiments of the application is shown. We analyzed an additional 8 patients with aplastic anemia who responded to immunosuppressive therapy, raising their hemoglobin levels. The median E% of plasma DNA in the treatment responsive group was 22.5% (interquartile range: 17.2-27.1%), which was higher than the non-responsive group (median: 12.3%; interquartile range: 7.5-13.7%) (P = 0.0003, Mann-Whitney rank sum test; Figure 13B ). There was a small but significant difference between the E% results of the treatment responsive group and healthy controls (P = 0.01, Mann-Whitney rank sum test).

[0273] These results reflect the restoration of erythropoietic activity in the bone marrow. Since the restoration of E% can occur earlier than hemoglobin levels, E% can be used to determine early whether a patient responds to immunosuppressive therapy. When a patient does not respond, other treatments (e.g., more aggressive treatments) can be performed, such as stem cell transplantation or prescription of bone marrow stimulants (e.g., sargramostim, filgrastim, and pegfilgrastim).

[0274] G. Leukemia

[0275] Other blood disorders besides leukemia can also be detected using erythroblast-specific DMRs, such as those detected in FECH.

[0276] Figure 14 Figure showing plasma Unmeth% for hemoglobin concentration in a normal subject and two patients with leukemia according to embodiments of the application. Unmeth% was determined using FECH DMR. The Unmeth% value in the plasma of patients with leukemia or myeloproliferative disorders is higher than the median Unmeth% of the plasma of normal subjects. This observation is consistent with the observation of increased but defective erythropoiesis in patients with leukemia. Thus, a cutoff of about 45 can be used for Unmeth% to distinguish between healthy subjects and subjects with leukemia, determining the level of blood disorder. Hemoglobin levels can also be used, for example, patients with hemoglobin below 8 can be identified as having leukemia rather than beta-thalassemia, which typically has hemoglobin levels between 8 and about 11.8, as shown in Figure 10

[0277] VI. Results for other methylation markers

[0278] We analyzed E% in plasma from a subset of samples based on two other DMRs to validate the above E% results from the DMR associated with the FECH gene. Using the other two erythroblast-specific DMRs as DMRs in the FECH gene, a similar difference in the percentage of red blood cell DNA in plasma between healthy subjects and patients with aplastic anemia and beta-severe thalassemia was observed.

[0279] A. Other two erythroblast-specific DMRs

[0280] The other two DMRs, located on chromosome 12, are also hypomethylated. The genomic regions associated with these two DMRs were not previously identified as within any annotated gene.

[0281] Figure 15A and 15B Figure showing the methylation density of CpG sites within the erythroblast-specific DMR on chromosome 12 according to embodiments of the application. Figure 15A Figure showing region 1510:48227688-48227701 at genomic coordinates on chromosome 12, which includes 3 sites. Figure 15B ​The region 1560:48228144-48228154 at genomic coordinates on chromosome 12 is shown, which also includes 3 loci. The genomic coordinates correspond to human reference genome hgl9. The selected CpG loci located within the shaded region are all hypomethylated in erythroblasts, but highly methylated in other tissue or cell types. The other tissues represent lung, colon, small intestine, pancreas, adrenal gland, esophagus, heart, and brain.

[0282] These two other erythroblast-specific DMRs are labeled Ery-1 and Ery-2. The E% based on the other two DMRs (chromosome 12:48227688-48227701 and chromosome 12:48228144-48228154) will be denoted E%(Ery-1) and E%(Ery-2), respectively.

[0283] Figure 16 The histone modifications (H3K4mel and H3K27Ac) for the two other erythroblast-specific DMRs (Ery-1 and Ery-2) from the ENCODE database are shown. We reviewed the public data for histone modification and CHIP-seq datasets for these two DMRs in the erythroblast cell type from the ENCODE database. The Ery-1 and Ery-2 DMRs are marked by two enhancer-associated histone modifications (H3K4mel and H3K27Ac), which suggest regulatory functions, especially of enhancers. The nearest downstream gene is the HDAC7 gene, approximately 15 kb away.

[0284] B. Erythroblast-enriched samples

[0285] We analyzed the percentage of erythroid DNA based on the other two DMRs in erythroblast-enriched samples from the eight cord blood samples described previously. The E%(Ery-1) and E%(Ery-2) of DNA extracted from the pooled samples were 66.5% and 68.5%. These E% values are similar to the E% based on the DMR associated with the FECH gene, which is 67%. Given the similar findings from all three DMRs, the lower than expected E% values (i.e., lower than expected when enrichment is performed) can be attributed to incomplete selectivity of the enrichment protocol.

[0286] C. Correlation of E% in buffy coat with erythroblasts in beta-thalassemia major patients

[0287] The percentage of erythroid DNA based on the two DMRs was analyzed in the buffy coat DNA of the same group of beta-thalassemia major patients. The E% of the two DMRs in the buffy coat DNA correlated well with the percentage of erythroblasts in a similar manner as Figure 3A the percentage of erythroblasts in a similar manner as

[0288] Figure 17A and 17B Correlation between the percentage of erythroid DNA sequences (E%) in buffy coat DNA of β-thalassemia major patients measured by digital PCR analysis targeting Ery-1 marker (left panel) Figure 17A ) and Ery-2 marker (right panel) and the percentage of erythroblasts in all peripheral white blood cells measured using an automated hematology analyzer. E% (Ery-1) and E% (Ery-2) of buffy coat DNA were well correlated with the percentage of erythroblasts in peripheral white blood cells measured by hematology analyzer (r = 0.938 and r = 0.928, both P < 0.0001, Pearson correlation).

[0289] Figure 18A and 18B Correlation between the E% (FECH) results of buffy coat DNA of β-thalassemia major patients and E% (Ery-1) and E% (Ery-2). E% results derived from these two DMRs were also well correlated with the paired E% results derived from the FECH gene marker site in buffy coat DNA of 15 β-thalassemia major patients.

[0290] C. E% in plasma of healthy subjects and anemic patients

[0291] We analyzed E% (Ery-1) and E% (Ery-2) of plasma DNA in healthy subjects and patients with aplastic anemia and β-thalassemia major. E% results based on three erythroid-specific DMRs were analyzed in the same group of healthy subjects, 7 patients with aplastic anemia and 9 patients with β-thalassemia major.

[0292] Figure 19The percentage of erythrocyte DNA in healthy subjects and patients with aplastic anemia and beta-thalassemia major using digital PCR analysis targeting three erythroblast-specific DMRs according to embodiments of the application is shown. The median E% of plasma DNA of 13 healthy subjects (Ery-1) was 16.7% (interquartile range: 10.9-23.5%) and the median E% of the same group of healthy subjects (Ery-2) was 25.0% (interquartile range: 22.2-27.3%). Based on the Ery-1 marker, the E% of patients with aplastic anemia and beta-thalassemia major (Ery-1) was 13.78% and 61.69%, respectively. Based on the Ery-2 marker, the E% of patients with aplastic anemia and beta-thalassemia major (Ery-2) was 14.13% and 64.95%, respectively. Using two erythroblast-specific DMRs as DMRs in the FECH gene, a similar difference in the percentage of erythrocyte DNA in plasma between healthy subjects and patients with aplastic anemia and beta-thalassemia major was observed.

[0293] VII. Treatment Results

[0294] As described above, E% can be used to monitor the efficacy of treatment of anemia.

[0295] A. Measurement of E% (FECH) of plasma DNA in patients with iron deficiency anemia before and after iron therapy

[0296] We monitored the serial changes in hemoglobin levels, reticulocyte counts, and E% of plasma DNA in 4 patients with iron deficiency anemia who received intravenous iron therapy due to intolerance to oral iron because of gastrointestinal side effects. Instead of patients on oral iron therapy, we chose to observe changes in this group of patients to avoid the possible confounding factor of different treatment responses due to changes in gastrointestinal absorption. We measured these parameters before treatment and two days after treatment.

[0297] Figure 20A and 20B Serial measurements of the percentage of erythrocyte DNA in plasma (E% (FECH)) and the percentage of reticulocyte counts in patients with iron deficiency anemia who received intravenous iron therapy before the state of treatment and two days after treatment according to embodiments of the application are shown. Figure 20A Serial changes in E% of plasma DNA are shown. Figure 20B Serial changes in the percentage of reticulocyte counts are shown.

[0298] Except for Subject 1, E% of plasma DNA and reticulocyte count increased, while hemoglobin levels remained static only initially after treatment initiation. Regarding the final change in hemoglobin levels, Subjects 3 and 4 experienced dramatic changes, at 84.7% and 75.3%, respectively. Subject 2 was followed by default and did not provide additional samples for hemoglobin measurement after treatment. Subject 1, with minimal change in E% of plasma DNA, showed the least increase in hemoglobin levels (12.2%). Therefore, changes in E% of plasma DNA can confirm the dynamic response of bone marrow erythropoietic activity to iron therapy and can serve as an early predictor of patient response to treatment.

[0299] The lack of an increase in reticulocyte count in Subject 1 indicates that RBC production is not adequately responding to iron therapy. A lack of response to iron therapy can also be reflected by a lack of increase in E% (equivalent to bone marrow activity). However, in Subject 1, the pre-iron therapy hemoglobin level was higher than in the other three subjects and closer to the reference range for healthy subjects. The lack of an increase in E% (FECH) in Subject 1 could reflect the absence of a compensatory increase in erythropoietic activity in the bone marrow, due to the smaller deficiency in hemoglobin levels compared to normal. Subject 1's reticulocyte count was initially roughly the same as the other subjects and therefore does not indicate an adequate level of bone marrow activity. Therefore, for anemia with moderate erythropoietic activity, an E% above the upper limit of normal in healthy patients could indicate a positive response to treatment, or at least an indeterminate response, and thus treatment may not be discontinued in such cases.

[0300] To restore hemoglobin levels to normal, increased RBC production is required. Therefore, in iron deficiency anemia, the normal range for E% can be considered inappropriate. An increase in E% in subjects 2–4 indicates an appropriate response to iron therapy, as normal levels are expected for subjects with iron deficiency anemia (see [link to relevant documentation]). Figure 10 This could be an E% threshold that is either exactly higher than the normal range or just above the upper limit of E%. Therefore, the threshold for E% used to determine whether a treatment is effective can depend on the initial value of E%. The threshold for E% can specify a particular change in value relative to the initial value, where the amount of change can depend on the initial value.

[0301] The effects of oral iron therapy were also studied. For example, patients with chronic blood loss due to menorrhagia are prone to iron deficiency anemia. Iron supplements will be used to correct the iron deficiency.

[0302] Figure 21A This invention illustrates a series of changes in plasma E% of erythroblast DMR in patients with iron deficiency anemia due to menorrhagia receiving oral iron therapy according to embodiments of the present invention. The plasma E% of patients with iron deficiency anemia receiving iron therapy was analyzed before and seven days after iron therapy. Figure 21AIn the middle, E% increased after iron treatment. These results show that plasma E% can reflect erythropoietic activity in response to treatment.

[0303] Figure 21B Hemoglobin levels have not significantly increased after iron treatment, while E% (FECH - increased at the same time point after treatment. This is in agreement with Figure 20A Similarly, it shows that E% can be used as an early detection of whether treatment is effective.

[0304] B. Treatment of Chronic Kidney Disease (CKD)

[0305] In CKD patients, one of the main causes of anemia is the decrease in erythropoietin production due to kidney impairment. Erythropoietin is a hormone produced by the kidney in response to low tissue oxygen levels. It stimulates the bone marrow to produce red blood cells. Exogenous erythropoietin can be used to treat CKD anemia.

[0306] Figure 22 Serial changes in plasma Unmeth% at the erythroblast DMR in patients with chronic kidney disease (CKD) receiving treatment with recombinant erythropoietin (EPO) or erythropoiesis stimulating agents (ESAs) are shown. Plasma Unmeth% was analyzed in seven CKD patients receiving EPO treatment before and 7 to 14 days after EPO treatment. Lines of different shapes (colors) correspond to different patients. All patients showed an increase in Unmeth% after EPO treatment. Unmeth% values show different levels of efficacy in different patients. These results show that plasma Unmeth% reflects erythropoietic activity in response to treatment.

[0307] C. ATG treatment of aplastic anemia

[0308] Immunosuppressive therapy of aplastic anemia patients can lead to blood recovery in 60-70% of patients (Young et al. Blood 2006; 108(8): 2509-2519). Plasma Unmeth% values were analyzed in four patients with aplastic anemia receiving immunosuppressive therapy before the start of the immunosuppressive therapy and 2 and 4 months after. All patients did not respond to the treatment and hemoglobin levels did not recover to normal levels during this period; all four patients required regular blood transfusions.

[0309] Figure 23AThe successive changes in plasma Unmeth% at the erythroid DMR in patients with aplastic anemia receiving anti-thymocyte globulin (ATG) treatment or cyclosporine as immunosuppressive therapy according to embodiments of the application are shown. In three patients, there was no change in Unmeth% in plasma. One patient showed a significant increase in Unmeth%. This coincided with the appearance of symptoms of paroxysmal nocturnal hemoglobinuria (PNH) clone, i.e. passing dark urine containing hemoglobin. Even if Unmeth% increased, this symptom can be used to determine that the patient is not responding to the treatment. PNH is known for its appearance in patients with aplastic anemia and has a pathophysiological mechanism of hemolytic anemia. The increase in Unmeth% reflects an increase in erythropoietic activity due to PNH hemolysis.

[0310] Figure 23B The series of changes in hemoglobin in patients with aplastic anemia receiving treatment is shown. The hemoglobin level did not increase significantly. These results show that plasma Unmeth% reflects changes in erythropoietic activity during the course of treatment, which did not change in erythropoietic activity because all patients did not respond to the treatment, as exemplified by the lack of change in hemoglobin, as shown in Figure 23B

[0311] Figure 24A 24B A graph showing Unmeth% of plasma for hemoglobin concentration in four patients with aplastic anemia is shown. Each line corresponds to a patient and tracks the changes in Unmeth% and hemoglobin level before treatment and after 4 months of treatment. Figure 23A It is shown that Unmeth% did not change significantly except for the patient with PNH. Figure 23B It is shown that hemoglobin level did change, but not significantly.

[0312] VIII. Use of the absolute concentration of red blood cell DNA

[0313] For measuring the amount of red blood cell DNA in plasma / serum, some embodiments use the parameter E% (also called Unmeth%) at the hypomethylated marker, but also highly methylated markers specific to the cell lineage can be used, if present. E% corresponds to the amount of erythroid DNA normalized to the total amount of DNA in the sample, which is mostly highly methylated.

[0314] An alternative parameter is to measure the absolute concentration of red blood cell DNA per unit volume of plasma. To calculate E%, embodiments can measure the absolute concentration of unmethylated DNA and the absolute concentration of methylated DNA. In a digital PCR analysis, each dot can represent one DNA molecule (e.g., as Figure 2A and​​Figure 2B The counts of methylated and unmethylated DNA can be directly counted (as shown in FIG. 6). In previous sections, normalized values (e.g., E%) were calculated, but embodiments can also use the absolute concentration of unmethylated molecules for hypomethylated markers or the absolute concentration of methylated molecules for hypermethylated markers.

[0315] Figure 25 Box-whisker plots showing the absolute concentration of red blood cell DNA at the FECH gene-associated DMR (copies per mL of plasma) in healthy subjects and anemic patients according to embodiments of the application are illustrated. The box and internal line represent the interquartile range and median, respectively. The top and bottom whiskers represent the maximum and minimum values.

[0316] As Figure 25 While the absolute concentration of red blood cell DNA can be used to observe separate clusters between different patient groups, as shown in FIG. 6, the normalized values allow for better separation between groups. In theory, the E% parameter of plasma can also be affected by the concentration of circulating DNA of non-red blood cell origin (e.g., bone marrow or lymphoid-derived DNA). For example, in anemic conditions where other hematopoietic lineages are also affected (e.g., aplastic anemia or myelodysplastic syndrome), altered red blood cell DNA release can be masked in some cases.

[0317] IX. Other blood lineages

[0318] This plasma DNA-based approach for hematological assessment can be extended to markers of other blood cell lineages, such as the myeloid, lymphoid, and megakaryocytic lineages. Previous studies using blood lineage-specific DNA methylation markers have focused mainly on whole blood or blood cells (Houseman EA et al. Current Environmental Health Reports 2015; 2: 145-154). Our data presented above clearly show that plasma DNA does contain information not present in blood cells. Therefore, analysis of plasma DNA using epigenetic markers from multiple blood cell lineages can provide valuable diagnostic information about an individual's blood system. It is thus a non-invasive alternative to bone marrow biopsy. Assays can be designed to specifically detect methylation markers of a particular cell lineage in plasma or serum, such that the activity of different cell lineages in the bone marrow can be monitored.

[0319] This approach is applicable to the assessment of many clinical situations, including (but not limited to) the following conditions. Examples of relevant lineages are provided for the conditions.

[0320] 1. Malignant hematological tumors, such as leukemias and lymphomas (lymphocytic lineage)

[0321] 2. Bone marrow disorders, such as aplastic anemia, myelofibrosis (bone marrow cells and lymphocyte lineage)

[0322] 3. Monitoring the immune system and its functions: e.g. immune deficiencies and establishment of immune responses during disease and treatment (lymphocyte lineage)

[0323] 4. Effects of drugs on bone marrow, e.g. azathioprine (bone marrow cell lineage)

[0324] 5. Autoimmune disorders with hematological manifestations, e.g. immune thrombocytopenia (ITP), which is a disorder characterized by low platelet counts but with normal bone marrow. Plasma DNA analysis using blood lineage markers (e.g. megakaryocyte markers) will provide valuable diagnostic information about this condition. (megakaryocyte markers)

[0325] 6. Infections with hematological complications, e.g. infection with parvovirus B19, which can be complicated by reduced erythropoiesis or even more severe aplastic crisis. (erythroid lineage)

[0326] X. Methods

[0327] Figure 26 is a flowchart showing a method 2600 of analyzing a blood sample of a mammal according to an embodiment of the present application. Portions of the method 2600 can be performed manually, while other portions can be performed by a computer system. In one embodiment, a system can perform all of the steps. For example, a system can include robotic elements (e.g., to obtain samples and perform analyses), a detection system for detecting signals from the analyses, and a computer system for analyzing the signals. Instructions for controlling this system can be stored in one or more computer-readable media, such as configuration logic of a field programmable gate array (FPGA), flash memory, and / or a hard drive. Figure 27 Such a system is shown.

[0328] At block 2610, a cell-free mixture of a blood sample is obtained. Examples of cell-free mixtures include plasma or serum. The cell-free mixture can include cell-free DNA from multiple cell lineages.

[0329] In some embodiments, the blood sample is separated to obtain a cell-free mixture. Plasma and serum are different. Both correspond to the fluid portion of blood. To obtain plasma, an anticoagulant is added to the blood sample to prevent it from clotting. To obtain serum, the blood sample is allowed to clot. Thus, during clotting, clotting factors will be consumed. With respect to circulating DNA, some DNA will be released from blood cells to the fluid portion during clotting. Thus, serum has a higher DNA concentration than plasma. The DNA of cells during clotting can dilute DNA specific to plasma. Thus, plasma can be advantageous.

[0330] At block 2620, the DNA fragments in the cell-free mixture are contacted with an assay corresponding to one or more differentially methylated regions. Each of the one or more differentially methylated regions (DMRs) is specific to a particular blood cell lineage by being hypomethylated or hypermethylated relative to other cell lineages. Examples of DMRs for the erythroblast lineage are provided herein.

[0331] In various embodiments, the assay can involve PCR or sequencing, and thus is a PCR assay or a sequencing assay. The contacting with the DNA fragments can involve flow cells, droplets, beads, or other mechanisms to provide for the assay to interact with the DNA fragments. Examples of such assays include whole genome bisulfite sequencing, targeted bisulfite sequencing (by hybrid capture or amplicon sequencing), other methylation sensing (e.g., single molecule real-time (SMRT) DNA sequencing by Pacific Biosciences), real-time methylation specific PCR, and digital PCR. Other examples of assays that can be used in the method 300 are described herein (e.g., in Section XII). Although examples use erythroblasts, other cell lineages, including other blood cell lineages, can also be used.

[0332] At block 2630, a first number of methylated or unmethylated DNA fragments is detected in the cell-free mixture at the one or more differentially methylated regions based on a signal obtained from the assay. The assay can provide various signals, such as optical or electrical signals. The signal can provide a specific signal for each DNA fragment, or an aggregate signal indicative of the total number of DNA fragments with the methylation signature (e.g., as in real-time PCR).

[0333] In one embodiment, sequencing can be used to obtain sequence reads for the DNA fragments, and the DNA fragments can be aligned to a reference genome. If a DNA fragment aligns to one of the DMRs, a counter can be incremented. Given that the signal is from a specific methylation-unmethylation assay, it can be assumed that the DNA fragment has the methylation signature. In another embodiment, reads from PCR (e.g., optical signals from positive wells) can be used to increment such a counter.

[0334] At block 2640, the first number is used to determine a methylation level. The first methylation level can be normalized or an absolute concentration, such as per volume of biological sample. Figure 25 Examples of absolute concentrations are provided. Examples of normalized methylation levels include E% (also referred to as Unmeth%).

[0335] For normalized values, the methylation level can be determined using the first number and total number of DNA fragments in the cell-free mixture at one or more differentially methylated regions. As described above, the methylation level can be a percentage of unmethylated DNA fragments. In other embodiments, the percentage can be a percentage of methylated DNA fragments, which would have an inverse relationship to the above example of erythroblast formation. In various embodiments, the methylation level can be determined using a percentage of all sites in the DMR, by an average of individual percentages for each site, or a weighted average for each site.

[0336] At block 2650, the methylation level is compared to one or more cutoff values as part of determining a classification of a blood disorder in the mammal. The one or more cutoff values can be selected from empirical data, such as shown in Figure 8-10 and Figure 12-14 The cutoff values can be selected to provide optimal sensitivity and specificity to provide accurate classification of a blood disorder, such as based on supervised learning from a dataset of samples known to be normal and to have a disorder.

[0337] As an example for determining cutoff values, a plurality of samples can be obtained. Each sample is known to have a particular classification of a blood disorder, such as by other techniques known to one of skill in the art. The plurality of samples have a classification of at least two blood disorders, such as having the disorder and not having the disorder. Different types of disorders can also be included, such as shown in Figure 10 The methylation level for one or more differentially methylated regions can be determined for each of the plurality of samples as a data point in Figure 8-10 and Figure 12-14

[0338] A first set of samples can be identified as having a first classification of a blood disorder, such as a first classification of being healthy. The first set can be clustered together, such as shown in Figure 8-10 and Figure 12-14 ​The second set of samples can discriminate a second classification of having a blood disorder. The second set can be patients having the disorder or a different type of disorder than the first classification. The two classifications can correspond to different degrees of having the same disorder. When the first set of samples collectively have a statistically higher methylation level than the second set of samples, a cutoff value can be determined that discriminates between the first set of samples and the second set of samples within a particular specificity and sensitivity. Thus, a balance between specificity and sensitivity can be used to select an appropriate cutoff value.

[0339] XI. SUMMARY

[0340] RBCs are the most abundant cell type in peripheral blood, but lack a nucleus. In this disclosure, we determined that cells of the erythroid lineage contribute a significant proportion of the plasma DNA pool. Prior to this study, it was known that hematopoietic cells contribute significantly to the circulating DNA pool (13, 14), but it was assumed that this hematopoietic DNA was only from the leukocytic lineage (15). Recent results using DNA methylation markers have shown that plasma DNA carries DNA methylation markers of neutrophils and lymphocytes (15).

[0341] Using a high-resolution reference methy lome that includes several tissues of erythroblasts (18, 20), we distinguished DNA molecules of erythroid origin from other tissue types in the plasma DNA pool. We were able to quantitatively analyze these DNA molecules in plasma based on digital PCR analysis of erythroblast-specific methylation markers. This approach allowed us to demonstrate the presence of a large amount of erythroid DNA in the plasma DNA pool of healthy subjects.

[0342] Our results are consistent with the results of cells of the erythroid lineage in the bone marrow contributing DNA to plasma. The corollary of this hypothesis is that quantitative analysis of erythroid DNA in plasma reflects erythropoietic activity in the bone marrow and, thus, can aid in the diagnosis of anemia. We have established a reference value for erythroid DNA in plasma of healthy subjects. We further demonstrated that anemic patients can increase or decrease the proportion of circulating erythroid DNA, depending on the exact nature of their pathology and treatment. In particular, we can distinguish between two syndromes of bone marrow failure, aplastic anemia (AA) and myelodysplastic syndrome (MDS), in the patients we recruited by analyzing the percentage of erythroid DNA in plasma.

[0343] Reticulocyte counts can be used to provide information about the bone marrow response in anemic patients. In our study of 11 β-thalassemia patients, four patients had reticulocyte counts in the peripheral blood below 1%, the limit of detection. For the other seven patients, the reticulocyte counts ranged between 1% and 10%. For all nine patients with aplastic anemia, whether or not they received regular transfusions, the reticulocyte counts were below 1%. Thus, reticulocyte counts can not unambiguously define normal and reduced erythropoietic activity in the bone marrow at low concentrations of reticulocytes due to the high inaccuracy of automated methods (35, 36).

[0344] We have shown that analysis of reticulocyte counts or reticulocyte indices cannot distinguish between the causes of anemia with reduced erythropoietic activity in our patient population Figure 11A and Figure 11B ). Our results indicate that plasma Unmeth% (e.g., as shown in Figure 10 ) is more accurate than reticulocyte counts in reflecting erythropoietic activity in the bone marrow. In all patients where both parameters were measured, there was no correlation between plasma Unmeth% and reticulocyte counts or reticulocyte indices (P = 0.3, linear regression), as shown in Figure 11A and Figure 11B .

[0345] Similarly, the presence of an abnormally high number of erythroblasts in the peripheral blood implies an abnormal erythropoietic stress (37). However, the absence of erythroblasts in the peripheral blood does not imply normal or reduced erythropoietic activity. In contrast, the quantitative analysis of plasma erythroblast-derived DNA yields information about the bone marrow erythropoietic activity that is not provided by routine blood parameters of the peripheral blood.

[0346] For β-thalassemia and aplastic anemia, both conditions are usually diagnosed by analyzing blood and the pattern of hemoglobin for iron. However, both β-thalassemia and aplastic anemia exhibit low hemoglobin levels, so this technique cannot distinguish between β-thalassemia and aplastic anemia. With Unmeth%, more specificity can be provided by enabling the distinction between these two pathologies, as shown in Figure 8 and Figure 10 .

[0347] Unmeth% can also be used to monitor treatment. For example, analysis of Unmeth% in patients with iron deficiency anemia can be used to monitor the response of the bone marrow to oral iron therapy, as shown in Figure 9In some patients, oral iron supplements can be effectively absorbed through the gastrointestinal tract. As a result, erythropoiesis can not be increased after treatment is initiated. The absence of an increase in Unmeth% can be used as an indicator of an adverse reaction to oral iron treatment, allowing for initiation of other treatments, such as parenteral iron treatment (e.g., dextran iron, iron gluconate, and sucrose iron). Alternatively, the absence of an increase in Unmeth% can be used to discontinue (stop) treatment, thereby saving the cost of ineffective treatment. In another embodiment, the absence of an increase in Unmeth% can be used to identify when to increase the dosage of treatment, such as the dosage of iron. When Unmeth% increases, treatment can be continued. If the increase in Unmeth% is high enough (e.g., based on a threshold), then treatment can be stopped, as it can be assumed that the erythropoietic activity has reached a level sufficient to eventually restore hemoglobin levels to healthy levels, thereby avoiding an expensive and potentially harmful over-treatment.

[0348] Thus, the blood disorder of the mammal can be treated in response to determining that the classification of the blood disorder indicates that the mammal has the blood disorder. After treatment, the analysis can be repeated to determine updated methylation levels, and a determination can be made as to whether to continue treatment based on the updated methylation levels. In one embodiment, the determination as to whether to continue treatment can include stopping treatment, increasing the dosage of treatment, or performing a different treatment when the updated methylation levels do not change from the methylation levels to within a specified threshold. In another embodiment, the determination as to whether to continue treatment can include continuing treatment when the updated methylation levels change from the methylation levels to within a specified threshold.

[0349] As another example of monitoring treatment, Unmeth% analysis can be used to determine whether ineffective erythropoiesis has been sufficiently inhibited by treatment, such as by transfusion, in a patient with thalassemia. Extramedullary erythropoiesis is the cause of skeletal deformities in patients with thalassemia. Extramedullary erythropoiesis can be inhibited by transfusion and restoration of hemoglobin levels. Unmeth% can show the patient's response to these treatments, and failure to inhibit Unmeth% can be used to indicate that treatment should be intensified.

[0350] Unmeth% can also be used to distinguish patients with iron deficiency alone from patients with iron deficiency and other causes of anemia (e.g., anemia due to chronic disease). Patients with iron deficiency alone are expected to increase Unmeth% after iron treatment, but those with multiple causes of anemia will not show a response of increased Unmeth%.

[0351] Therefore, we have demonstrated that the percentage of circulating erythrocyte DNA in patients with iron deficiency anemia responds to increased iron therapy, thus reflecting increased erythropoietic activity in the bone marrow. The dynamic changes in the proportion of erythrocyte DNA show that quantification of plasma erythrocyte DNA allows for non-invasive monitoring of relevant cellular processes. The rapid kinetics of plasma DNA (e.g., a half-life sequence of tens of minutes (38, 39)) suggest that such monitoring can provide near-real-time results. Similarly, the percentage of circulating cell-free DNA in other cell lineages also allows for non-invasive monitoring of relevant cellular processes in the bone marrow of other cell lineages.

[0352] Current research serves as evidence to confirm the presence of nuclear material from circulating hematopoietic progenitors and precursor cells. Additionally, the presence of circulating DNA released from precursor cells of other hematopoietic lineages can also be used.

[0353] In summary, we have demonstrated that erythrocyte DNA contributes a significant proportion of the plasma DNA pool. This discovery fills a crucial gap in our understanding of the fundamental biology of circulating nucleic acids. Clinically, the measurement of erythrocyte DNA in plasma opens a new avenue for studying and monitoring different types of anemia and heralds the beginning of a new family of cell-free DNA blood tests.

[0354] XII. Materials and Methods

[0355] This section describes the techniques that have been used and can be used to implement embodiments.

[0356] A. Sample collection and preparation

[0357] In some embodiments, normal tissues (liver, lung, esophagus, stomach, small intestine, colon, pancreas, adrenal gland, bladder, heart, brain, and placenta) from four different cases were harvested from anonymized surgical specimens, formalin-fixed and paraffin-embedded (FFPE). The collected tissues were confirmed to be normal on histological examination. With modifications to the manufacturer's fixation protocol, DNA was extracted from the FFPE tissue protocol using the QIAamp DNA Mini Kit (Qiagen). A dewaxing solution (Qiagen) was used instead of xylene to remove paraffin. An additional step of incubation at 90°C for 1 hour was performed after lysis with buffer ATL and proteinase K to reverse the formaldehyde modification of the nucleic acids.

[0358] To prepare erythrocyte-enriched samples for analysis, 1–3 mL of umbilical cord blood was collected from each of eight pregnant women immediately postpartum into tubes containing EDTA. Mononuclear cells were isolated from the umbilical cord blood samples by density gradient centrifugation using Ficoll-Paque PLUS solution (GE Healthcare). 1 × 10⁻⁶ 8One 1 mL mixture of isolated mononuclear cells and two antibodies—FITC-conjugated anti-CD235a (Medrin Biosciences) and phycoerythrin (PE)-conjugated anti-CD71 (Medrin Biosciences)—was diluted 1:10 in phosphate-buffered saline and incubated together in the dark at 4°C for 30 min. CD235a+ and CD71+ cells were subsequently sorted by BD FACSAria fusion flow cytometry (BD Biosciences) for erythrocyte enrichment (1). CD235a+CD71+ cells from eight cases were pooled for downstream analysis. DNA was extracted from the pooled CD235a+CD71+ cells using the QIAamp DNA Blood Mini Kit (Kagem) according to the manufacturer's instructions.

[0359] Peripheral blood samples were collected into tubes containing EDTA and immediately stored at 4°C. 10 mL of peripheral venous blood was collected from each patient. Plasma separation was performed within 6 hours of blood collection. Plasma DNA was extracted from 4 mL of plasma. Plasma and erythrocyte sedimentation rate (ESR) amber DNA were obtained as previously described (2). In short, the blood sample was first centrifuged at 1,600 g for 10 minutes at 4°C, and the plasma fraction was then centrifuged at 16,000 g for 10 minutes at 4°C. After a further centrifugation at 2,500 g for 10 minutes, the cellular fraction was collected to remove any residual plasma. DNA from the plasma and ESR amber fraction was extracted using the QIAamp DSP DNA Mini Kit (Qiagen) and the QIAamp DNA Blood Mini Kit (Qiagen), respectively.

[0360] B. Bisulfite transformation of DNA

[0361] Plasma DNA and genomic DNA extracted from blood cells and FFPE tissue were subjected to two rounds of bisulfite treatment using the Epitect Plus Bisulfite Kit (Kagem) according to the manufacturer’s instructions (3).

[0362] In one embodiment, DNA extracted from a biological sample is first treated with bisulfite. Bisulfite treatment converts unmethylated cytosine to uracil while preserving methylated cytosine. Therefore, after bisulfite conversion, methylated and unmethylated sequences can be distinguished based on sequence differences at the CpG dinucleotide. To analyze the plasma samples presented in selected embodiments of this application, DNA is extracted from 2-4 mL of plasma. To analyze DNA extracted from blood cells, 1 μg of DNA is used for downstream analysis in this example. In other embodiments, different volumes of plasma and amounts of DNA may be used.

[0363] In the examples in this application, each sample underwent two rounds of bisulfite treatment using the EpiTect bisulfite kit according to the manufacturer's instructions. The bisulfite-converted plasma DNA was eluted in 50 μL of water. The bisulfite-converted cellular DNA was eluted in 20 μL of water and then diluted 100-fold for downstream analysis.

[0364] C. Methylation status

[0365] Various methylation analyses can be used to quantify the amount of DNA from a specific cell lineage.

[0366] 1. PCR analysis

[0367] Two digital PCR analyses were performed, targeting each of the three erythroblast-specific DMRs: one targeting the unmethylated sequence of bisulfite conversion and the other targeting the methylated sequence. The primer and probe designs used for the analyses are listed in Supplementary Table 7.

[0368] FECH gene marker loci (chromosome 18: 55250563-55250585)

[0369]

[0370] Ery-1 marker sites (chr 12:48227688-48227701)

[0371]

[0372] Ery-2 marker sites (chromosome 12:48228144-48228154)

[0373]

[0374] Table 7. Oligonucleotide sequences for digital PCR analysis of methylated and unmethylated sequences of erythroblast-specific DMRs. Underlined nucleotides in the reverse primers and probes represent differentially methylated cytosine at CpG sites. VIC and FAM represent two fluorescent reporter molecules.

[0375] As an example, a PCR reaction may include 50 μL of template DNA converted with 3 μL of bisulfite, each primer at a final concentration of 0.3 μM, 0.5 μM MgCl2, and 25 μL of 2×KAPA HiFi HotStart Uracil ReadyMix. The following PCR thermal profile can be used: 95°C for 5 minutes and 35 cycles of 98°C for 20 seconds, 57°C for 15 seconds, and 72°C for 15 seconds, followed by a final extension step at 72°C for 30 seconds. In other embodiments, non-priority whole-genome sequencing may be combined with alignment, but such a procedure may not be cost-effective.

[0376] In some embodiments, for digital PCR analysis of samples, a 20 μL reaction mixture is prepared after bisulfite treatment. In one embodiment, the reaction mixture contains 8 μL template DNA, a final concentration of 450 nM for each of the two forward primers, a 900 nM reverse primer, and a 250 nM probe. In other embodiments, a total volume of 20 μL of each reaction mixture is prepared, containing 8 μL template DNA, a final concentration of 900 nM forward primer, 900 nM reverse primer, and 250 nM probe. The reaction mixture is then used to generate droplets using a BioRad QX200 ddPCR droplet generator. Typically, 20,000 droplets are generated per sample. In some implementations, droplets are transferred to clean 96-well plates and then thermally cycled using the same conditions for methylation- and non-methylation-specific analyses: 95°C for 10 minutes (1 cycle), 40 cycles of 94°C for 15 seconds and 60°C for 1 minute, 98°C for 10 minutes (1 cycle), followed by a 12°C hold step. After PCR, droplets from each sample are analyzed using a BioRad QX200 droplet reader, and the results are interpreted using QuantaSoft (version 1.7) software.

[0377] 2. Examples of other methylation analyses

[0378] Other examples of methyl-sensing sequencing include the use of single-molecule sequencing platforms, which would allow for the direct elucidation of DNA molecules (including N-methyl-sensing molecules). 6Methylated states of 5-methyladenine, 5-methylcytosine, and 5-hydroxymethylcytosine (Mc-adenine, 5-methylcytosine, and 5-hydroxymethylcytosine) without bisulfite conversion (AB Flusberg et al. 2010 Nature Methods; 7:461-465; J Shim et al. 2013 SciRep; 3:1389); or via immunoprecipitation of methylcytosine (e.g., by using an antibody against methylcytosine or by using a methylated DNA-binding protein or peptide (LG Acevedo et al. 2011 Epigenomics; 3:93-101) followed by sequencing; or via the use of a methylation-sensitive restriction enzyme followed by sequencing.

[0379] In some embodiments, the methylation levels of genomic sites in a DNA mixture can be determined using whole-genome bisulfite sequencing. In other embodiments, the methylation levels of genomic sites can be determined using methylation microarray analysis, such as the Illumina HumanMethylation450 system, or by using methylation immunoprecipitation (e.g., using anti-methylcytosine antibodies) or treatment with methylation-binding proteins followed by microarray analysis or DNA sequencing, or by using methylation-sensitive restriction enzyme treatment followed by microarray or DNA sequencing, or by using methylation-sensing sequencing, such as using single-molecule sequencing methods (e.g., nanopore sequencing (Schreiber et al., Proceedings of the National Academy of Sciences 2013; 110: 18910-18915) or by Pacific Biosciences single-molecule real-time analysis (Flusberg et al., Nature Methods 2010; 7: 461-465)). Tissue-specific methylation levels can be measured in the same manner. As another example, targeted bisulfite sequencing, methylation-specific PCR, and non-bisulfite-based methylation-sensing sequencing (e.g., using a single-molecule sequencing platform (Powers et al., Efficient and accurate whole genome assembly and methylome profiling of E. coli, BMC Genomics, 2013; 14:675) can be used to analyze plasma DNA methylation levels for plasma DNA methylation deconvolution analysis. Therefore, methylation-sensing sequencing results can be obtained in a variety of ways.

[0380] D. Statistical Analysis

[0381] Pearson correlation is used to study the correlation between the percentage of erythrocyte DNA (E% (FECH)) and the percentage of erythroblasts in peripheral blood leukocytes measured by a hematologic analyzer in patients with β-thalassemia major. Pearson correlation is also used to study the correlation between plasma DNA and paired E% (FECH) results in erythrocyte sedimentation rate (ESR) amber layer DNA in healthy controls. Wilcoxon's signed-rank test is used to compare the difference in E% between plasma DNA and paired ESR amber layer DNA in healthy subjects. Man Whitney's rank-sum test is used to compare the difference in E% between plasma DNA in healthy subjects and anemia patients in different disease groups.

[0382] We also developed a bioinformatics pipeline to mine erythrocyte-specific DMRs based on the criteria described in our paper. The bioinformatics pipeline can be implemented on various platforms, such as the Perl platform.

[0383] XIII. Instance System

[0384] Figure 27 A system 2700 according to an embodiment of the present invention is described. The system includes a sample 2705, such as cell-free DNA molecules within a sample holder 2710, wherein the sample 2705 is contactable with an analyzer 2708 to provide a signal of a physical characteristic 2715. An example of a sample holder may be a flow cell comprising a tube through which probes and / or primers or droplets of the analyzer move (the droplets comprising the analyzer). The physical characteristic 2715 from the sample (e.g., a fluorescence intensity value) is detected by a detector 2720. The detector may perform measurements at intervals (e.g., periodic intervals) to obtain data points constituting a data signal. In one embodiment, an analog-to-digital converter converts the analog signal from the detector into digital form multiple times. The data signal 2725 is transmitted from the detector 2720 to a logic system 2730. The data signal 2725 may be stored in a local memory 2735, an external memory 2740, or a storage device 2745.

[0385] The logic system 2730 may be or may include: a computer system, an ASIC, a microprocessor, etc. It may also include or be coupled to a display (e.g., a monitor, an LED display, etc.) and a user input device (e.g., a mouse, a keyboard, buttons, etc.). The logic system 2730 and other components may be part of a stand-alone or network-connected computer system, or they may be directly connected to or integrated into a thermal cycler device. The logic system 2730 may also include optimized software executed in the processor 2750. The logic system 1030 may include computer-readable media storing instructions for controlling the system 1000 to perform any of the methods described herein.

[0386] Any computer system mentioned herein can utilize any suitable number of subsystems. Examples of these subsystems in computer system 10 are shown in... Figure 28 In some embodiments, the computer system includes a single computer device, wherein a subsystem may be a component of the computer device. In other embodiments, the computer system may include multiple computer devices having internal components, each of which is a subsystem. The computer system may include desktop and laptop computers, tablet computers, mobile phones, and other mobile devices.

[0387] Figure 28 The subsystems shown are interconnected via system bus 75. Additional subsystems are shown, such as printer 74, keyboard 78, one or more storage devices 79, monitor 76 coupled to graphics card 82, etc. Peripheral devices and input / output (I / O) devices coupled to I / O controller 71 can be connected via various means known in the art (e.g., input / output (I / O) ports 77, e.g., USB). The computer system 10 can be connected to a computer system using I / O port 77 or external interface 81 (e.g., Ethernet, Wi-Fi, etc.). This connection allows the central processing unit 73 to communicate with each subsystem and control the execution of multiple instructions from system memory 72 or storage device 79 (e.g., a fixed disk, such as a hard disk drive or optical disk), as well as the exchange of information between subsystems. System memory 72 and / or one or more storage devices 79 may be embodied as computer-readable media. 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 the user.

[0388] A computer system may include multiple identical components or subsystems connected together, for example, via an external interface 81 or an internal interface. In some embodiments, the computer system, subsystem, or device may communicate via a network. In these cases, one computer may be regarded as a client and another computer as a server, where each may be part of the same computer system. The client and server may each include multiple systems, subsystems, or components.

[0389] Various aspects of the embodiments may be implemented in the form of hardware control logic (e.g., application-specific integrated circuits or field-programmable gate arrays) and / or using computer software with a general-purpose programmable processor in a modular or integrated manner. As used herein, the processor includes a single-core processor, a multi-core processor on the same integrated chip, or multiple processing units on a single circuit board or networked thereon. Based on this disclosure and the teachings provided herein, those skilled in the art will recognize and understand other ways and / or methods of implementing embodiments of the invention using hardware and combinations of hardware and software.

[0390] Any of the 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) employing techniques such as conventional or object-oriented methods. The software code can be stored as a series of instructions or commands on a computer-readable medium for storage and / or transmission. Suitable non-transitory computer-readable media may include random access memory (RAM), read-only memory (ROM), magnetic media such as hard disk drives or floppy disks, or optical media such as optical discs (CDs) or DVDs (Digital Universal Optical Discs), flash memory, etc. The computer-readable medium can be any combination of such storage or transmission means.

[0391] These programs can also be encoded and transmitted using carrier signals suitable for transmission over wired, optical, and / or wireless networks (including the Internet) conforming to various protocols. Therefore, computer-readable media can be generated using data signals encoded with these programs. Computer-readable media encoded with program code can be packaged with compatible devices or provided separately from other devices (e.g., downloaded via the Internet). Any such computer-readable media can reside on or within a single computer product (e.g., a hard disk drive, CD, or an entire computer system) and can exist on or within different computer products within a system or network. The computer system may include a monitor, printer, or other suitable display for providing a user with any of the results mentioned herein.

[0392] One of the methods described herein can be performed, either entirely or partially, using a computer system comprising one or more processors, which can be configured to perform the steps. Therefore, embodiments can relate to a computer system configured to perform the steps of any of the methods described herein, potentially having different components performing the respective steps or a corresponding set of steps. Although presented as numbered steps, steps of the methods herein can be performed simultaneously or in different orders. Furthermore, portions of these steps can be used in conjunction with portions of other steps of other methods. Moreover, all or part of the steps can be optional. Additionally, any step in any method can be performed using modules, circuitry, or other means for performing those steps.

[0393] The specific details of a particular embodiment may be combined in any suitable manner without departing from the spirit and scope of the embodiments of the invention. However, other embodiments of the invention may relate to specific embodiments relating to each individual aspect or a specific combination of these individual aspects.

[0394] For purposes of illustration and description, exemplary embodiments of the invention have been presented above. This is not intended to be exhaustive, nor is it intended to limit the invention to the precise forms described, and many modifications and variations are possible in light of the foregoing teachings.

[0395] 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 "inclusive or", not "exclusive or". A reference to the "first" component does not necessarily require the provision of the second component. Furthermore, unless explicitly stated otherwise, a reference to the "first" or "second" component does not limit the referenced component to a particular location.

[0396] For all purposes, all patents, patent applications, publications, and descriptions mentioned herein are incorporated herein by reference in their entirety. None of them are acknowledged as prior art.

[0397] XIV. References

[0398] The following references are incorporated herein by reference in their entirety for all purposes.

[0399] 1. Lo YM, Chan KC, Sun H, Chen EZ, Jiang P, Lun FM, et al., “Maternal plasma DNA sequencing reveals the genome-wide genetic and mutational profile of the fetus”, SciTransl Med, 2010; 2:61ra91.

[0400] 2. Chiu RW, Chan KC, Gao Y, Lau VY, Zheng W, Leung TY, et al., “Noninvasive prenatal diagnosis of fetal chromosomal aneuploidy by massively parallel genomic sequencing of DNA in maternal plasma”, Proceedings of the National Academy of Sciences, 2008; 105:20458-63.

[0401] 3. New MI, Tong YK, Yuen T, Jiang P, Pina C, Chan KC et al., “Noninvasive prenatal diagnosis of congenital adrenal hyperplasia using cell-free fetal DNA in maternal plasma”, J Clin Endocrinol Metab, 2014; 99:E1022-30.

[0402] 4. Bianchi DW, Parker RL, Wentworth J, Madankumar R, Saffer C, Das AF et al., DNA sequencing versus standard prenatal aneuploidy screening, The New England Journal of Medicine, 2014; 370:799-808.

[0403] 5. Chiu RW, Akolekar R, Zheng YW, Leung TY, Sun H, Chan KC et al., "Non-invasive prenatal assessment of trisomy 21 by multiplexed maternal plasma DNA sequencing: Large scale validity study" BMJ 2011;342:c7401.

[0404] 6. Bayindir B, Dehaspe L, Brison N, Brady P, Ardui S, Kammoun M, et al., “Noninvasive prenatal testing using a novel analysis pipeline to screen for all autosomal fetal aneuploidies improves pregnancy management”, Eur J Hum Genet, 2015; 23:1286-93.

[0405] 7. Scheffer PG, van der Schoot CE, Page-Christiaens GC, de Haas M. Noninvasive fetal blood group genotyping of rhesus D,c,E and of K in alloimmunized pregnant women: Evaluation of a 7-year clinical experience. BJOG 2011; 118:1340-8.

[0406] 8. Lo YM, Chan LY, Lo KW, Leung SF, Zhang J, Chan AT, et al., Quantitative analysis of cell-free Epstein-Barr virus DNA in plasma of patients with nasopharyngeal carcinoma. Cancer Research, 1999; 59:1188-91.

[0407] 9. Leary RJ, Sausen M, Kinde I, Papadopoulos N, Carpten JD, Craig D, et al., Detection of chromosomal alterations in the circulation of cancer patients with whole-genome sequencing, Science Translational Medicine, 2012; 4:162ra54.

[0408] 10. Chan KC, Jiang P, Zheng YW, Liao GJ, Sun H, Wong J, et al., "Cancer genome scanning in plasma: Detection of tumor-associated copy number aberrations, single-nucleotide variants, and tumoral heterogeneity by massively parallel sequencing", Clin Chem, 2013; 59:211-24.

[0409] 11. Bettegowda C, Sausen M, Leary RJ, Kinde I, Wang Y, Agrawal N et al., Detection of circulating tumor DNA in early-and late-stage human malignancies, Science Translational Medicine, 2014; 6:224ra24.

[0410] 12. Izumchenko E, Chang X, Brait M, Fertig E, Kagohara LT, Bedi A, et al., “Targeted sequencing reveals clonal genetic changes in the progression of early lung neoplasms and paired circulating DNA”, Nature Communications, 2015; 6:8258.

[0411] 13. Lui YY, Chik KW, Chiu RW, Ho CY, Lam CW, Lo YM. "Redominant hematopoietic origin of cell-free DNA in plasma and serum after sex-mismatched bone marrow transplantation", Clinical Chemistry, 2002; 48:421-7.

[0412] 14. Zheng YW, Chan KC, Sun H, Jiang P, Su X, Chen EZ, et al., “Nonhematopoietic derived DNA is shorter than hematopoietic derived DNA in plasma: A transplantation model”, Clinical Chemistry, 2012; 58:549-58.

[0413] 15. Sun K, Jiang P, Chan KC, Wong J, Cheng YK, Liang RH, et al., “Plasma DNA tissue mapping by genome-wide methylation sequencing for noninvasive prenatal, cancer, and transplantation assessments”, Proceedings of the National Academy of Sciences, 2015; 112:E5503-12.

[0414] 16. Keerthivasan G, Wickrema A, Crispino JD, "Erythroblast enucleation", Stem Cells International, 2011; 2011:139851.

[0415] 17. Chasis JA, Mohandas N., "Erythroblastic islands: Niches for erythropoiesis", Blood, 2008; 112:470-8.

[0416] 18. Adams D, Altucci L, Antonarakis SE, Ballesteros J, Beck S, Bird A, et al., “BLUEPRINT to decode the epigenetic signature written in blood”, Nature Biotechnol, 2012; 30:224-6.

[0417] 19. Martens JH, Stunnenberg HG., “BLUEPRINT: Mapping human blood cell epigenomes”, Haematologica 2013; 98:1487-9.

[0418] 20. Kundaje A, Meuleman W, Ernst J, Bilenky M, Yen A, Heravi-Moussavi A et al., “Integrative analysis of 111 reference human epigenomes”, Nature 2015; 518:317-30.

[0419] 21. ChimSS, Tong YK, Chiu RW, Lau TK, Leung TN, Chan LY, et al., “Detection of the placental epigenetic signature of the maspin gene in maternal plasma”, Proceedings of the National Academy of Sciences of the United States of America, 2005; 102:14753-8.

[0420] 22. Lehmann-Werman R, Neiman D, Zemmour H, Moss J, Magenheim J, Vaknin-Dembinsky A et al., Identification of tissue-specific cell death using methylation patterns of circulating DNA, Proceedings of the National Academy of Sciences, 2016; 113: E1826-34.

[0421] 23. Lun FM, Chiu RW, Sun K, Leung TY, Jiang P, Chan KC et al., “Noninvasive prenatal methylomic analysis by genomewide bisulfite sequencing of maternal plasma DNA”, Clinical Chemistry, 2013; 59:1583-94.

[0422] 24. Chan KC, Zhang J, Hui AB, Wong N, Lau TK, Leung TN, et al., Size distributions of maternal and fetal DNA in maternal plasma, Clinical Chemistry, 2004; 50:88-92.

[0423] 25. Yu SC, Chan KC, Zheng YW, Jiang P, Liao GJ, Sun H, et al., Size-based molecular diagnostics using plasma DNA for noninvasive prenatal testing, Proceedings of the National Academy of Sciences, 2014; 111:8583-8.

[0424] 26. Ferreira GC, Franco R, Lloyd SG, Moura I, Moura JJ, Huynh BH., “Structure and function of ferrochelatase”, Journal of Bioenerg Biomembr 1995; 27:221-9.

[0425] 27. Buoro S, Vavassori M, Pipitone S, Benegiamo A, Lochis E, FUMAGALLI S, et al., "Evaluation of nucleated red blood cell count by Sysmex XE-2100 in patients with thalassemia or sickle cell anemia and in neonates", Blood Transfusion, 2015; 13:588-94.

[0426] 28. Killick SB, Bown N, Cavenagh J, Dokal I, Foukaneli T, Hill A, et al., Guidelines for the diagnosis and management of adult aplastic anaemia, British Journal of Hematology (Br J Haematol) 2016; 172:187-207.

[0427] 29. Young NS., Acquired aplastic anemia, Ann Intern Med 2002; 136:534-46.

[0428] 30. Eschbach JW., “Erythropoietin 1991--an overview”, American Journal of Kidney Diseases, 1991; 18:3-9.

[0429] 31. Schrier SL., "Pathobiology of thalassemic erythrocytes", Curr Opin Hematol, 1997; 4:75-8.

[0430] 32. Kasper CK, Whisper DY, Wallerstein RO., Clinical aspects of iron deficiency, JAMA 1965;191:359-63.

[0431] 33. Bejar R, Stensma DP., Recent developments in myelodysplastic syndromes, Blood, 2014; 124:2793-803.

[0432] 34. Arber DA, Orazi A, Hasserjian R, Thiele J, Borowitz MJ, Le Beau MM, et al., “The 2016 revision to the World Health Organization classification of myeloid neoplasms and acute leukemia”, Blood, 2016; 127:2391-405.

[0433] 35. Buttarello M., “Laboratory diagnosis of anemia: Are the old and new red cell parameters useful in classification and treatment, how?”, International Journal of Laboratory Hematology, 2016; 38 Supplement 1: 123-32.

[0434] 36. Buttarello M, Bulian P, Farina G, Temporin V, Toffolo L, Trabuio E, Rizzotti P., “Flow cytometric reticulocyte counting”, “Parallel evaluation of five fully automated analyzers: An NCCLS-ICSH approach”, American Journal of Clinical Pathology (Am J Clin Pathol) 2001; 115:100-11.

[0435] 37. Danise P, Maconi M, Barrella F, Di Palma A, Avino D, Roveetti A, et al., Evaluation of nucleated red blood cells in the peripheral blood of hematological diseases, Clin ChemLab Med, 2012; 50:357-60.

[0436] 38. Lo YM, Zhang J, Leung TN, Lau TK, Chang AM, Hjelm NM. Rapid clearance of fetal DNA from maternal plasma. American Journal of Human Genetics (Am J Hum Genet) 1999; 64:218-24.

[0437] 39. To EW, Chan KC, Leung SF, Chan LY, To KF, Chan AT, et al. Rapid clearance of plasma Epstein-Barr virus DNA after surgical treatment of nasopharyngeal carcinoma. Clin Cancer Research, 2003; 9:3254-9.

Claims

1. A computer product comprising a computer readable medium storing a plurality of instructions for controlling a system to analyze a blood sample of a mammal by performing the following operations: detecting, based on signals obtained from the analysis, a first number of methylated or unmethylated DNA fragments in a cell-free mixture of the blood sample at one or more differentially methylated regions, the cell-free mixture comprising cell-free DNA from a plurality of cell lineages, each of the one or more differentially methylated regions having specificity for a particular blood cell lineage by being hypo- or hypermethylated relative to the other cell lineages, wherein the particular blood cell lineage is red blood cells; determining a methylation level using the first number; comparing the methylation level to one or more cutoff values; measuring a hemoglobin level of the blood sample; comparing the hemoglobin level to a hemoglobin threshold value; and determining a classification of a blood tumor in the mammal based on the comparison of the methylation level to the one or more cutoff values and the comparison of the hemoglobin level to the hemoglobin threshold value.

2. The computer product of claim 1, wherein the blood tumor is leukemia.

3. The computer product of claim 1, wherein the blood tumor is lymphoma.

4. The computer product of claim 1, wherein the blood tumor is polycythemia vera.

5. The computer product of claim 1, wherein the blood tumor is myelodysplastic syndrome.

6. The computer product of claim 1, wherein the plurality of instructions further control the system to perform the following operations: determining a total number of DNA fragments in the cell-free mixture at the one or more differentially methylated regions; and determining the methylation level using the first number and the total number.

7. The computer product of claim 1, wherein the plurality of instructions further control the system to perform the following operations: determining a volume of the cell-free mixture, wherein the methylation level is determined using the first number and the volume of the cell-free mixture.

8. The computer product of claim 1, wherein the plurality of instructions further control the system to perform the following operations to identify the one or more differentially methylated regions: obtaining a methylation index for a plurality of sites for each of a plurality of cell lineages, the plurality of cell lineages including the particular blood cell lineage and the other cell lineages; at each site of the plurality of sites, comparing the methylation index for the plurality of cell lineages; identifying one or more sites of the plurality of sites having a methylation index in the particular blood cell lineage that is below / above a first methylation threshold and having a methylation index in each of the other cell lineages that is above / below a second methylation threshold; and identifying a differentially methylated region containing the one or more sites. ​ 9. The computer product of claim 1, wherein the plurality of instructions further control the system to determine the one or more cutoff values by: obtaining a plurality of samples, each known to have a particular classification of the hematological neoplasm, the plurality of samples having at least two classifications of the hematological neoplasm; determining a methylation level of the one or more differentially methylated regions for each of the plurality of samples; identifying a first set of samples having a first classification of the hematological neoplasm; identifying a second set of samples having a second classification of the hematological neoplasm, the first set of samples collectively having a statistically different methylation level than the second set of samples; and determining a cutoff value that distinguishes the first set of samples and the second set of samples within a particular specificity and sensitivity.

10. The computer product of claim 1, wherein the plurality of instructions further control the system to: determine the presence of a hematological neoplasm based on a comparison of the methylation level to one or more cutoff values; and perform a bone marrow biopsy in response to determining the presence of the hematological neoplasm.

11. The computer product of claim 1, wherein the one or more differentially methylated regions comprise CpG sites.

12. The computer product of claim 11, wherein a first region of the one or more differentially methylated regions comprises a plurality of CpG sites within 100 bp of each other, and wherein the plurality of CpG sites are all either hypo-methylated or hyper-methylated.

13. The computer product of claim 1, wherein the one or more differentially methylated regions are hypo-methylated.

14. The computer product of claim 1, wherein one of the one or more differentially methylated regions is in the FECH gene.

15. The computer product of claim 1, wherein one of the one or more differentially methylated regions is in chromosome 12 at genomic coordinates 48227688-48227701 of the human reference genome hg19.

16. The computer product of claim 1, wherein one of the one or more differentially methylated regions is in chromosome 12 at genomic coordinates 48228144-48228154 of the human reference genome hg19.

17. A system comprising: the computer product of any one of claims 1-16; and one or more processors to execute the instructions stored on the computer-readable medium.

18. A computer system configured to analyze a blood sample of a mammal by: detecting, based on signals obtained from the analysis, a first number of methylated or unmethylated DNA fragments in a cell-free mixture of the blood sample at one or more differentially methylated regions, the cell-free mixture comprising cell-free DNA from a plurality of cell lineages, each of the one or more differentially methylated regions having specificity for a particular blood cell lineage by being hypo-methylated or hyper-methylated relative to other cell lineages, wherein the particular blood cell lineage is red blood cells; determining a methylation level using the first number; comparing the methylation level to one or more cutoff values; measuring a hemoglobin level of the blood sample; comparing the hemoglobin level to a hemoglobin threshold value; and determining a classification of a blood tumor in the mammal based on the comparison of the methylation level to one or more cutoff values and the comparison of the hemoglobin level to the hemoglobin threshold value.

19. The computer system of claim 18, wherein the blood tumor is leukemia.

20. The computer system of claim 18, wherein the blood tumor is lymphoma.

21. The computer system of claim 18, wherein the blood tumor is polycythemia vera.

22. The computer system of claim 18, wherein the blood tumor is myelodysplastic syndrome.

23. The computer system of claim 18, wherein the computer system is further configured to implement the following operations: determining a total number of DNA fragments in the cell-free mixture at the one or more differentially methylated regions; and determining the methylation level using the first number and the total number.

24. The computer system of claim 18, wherein the computer system is further configured to implement the following operations: determining a volume of the cell-free mixture, wherein the methylation level is determined using the first number and the volume of the cell-free mixture.

25. The computer system of claim 18, wherein the computer system is further configured to implement the following operations to identify the one or more differentially methylated regions: obtaining a methylation index for a plurality of sites for each of a plurality of cell lineages, the plurality of cell lineages including the particular blood cell lineage and the other cell lineages; at each site of the plurality of sites, comparing the methylation index for the plurality of cell lineages; identifying one or more sites of the plurality of sites that have a methylation index below / above a first methylation threshold in the particular blood cell lineage and a methylation index above / below a second methylation threshold in each of the other cell lineages; and identifying a differentially methylated region containing the one or more sites.

26. The computer system of claim 18, wherein the computer system is further configured to implement the following operations to determine the one or more cutoff values: obtaining a plurality of samples, each sample known to have a particular classification of the blood tumor, the plurality of samples having at least two classifications of the blood tumor; determining a methylation level of the one or more differentially methylated regions for each of the plurality of samples; identifying a first set of samples having a first classification of the blood tumor; identifying a second set of samples having a second classification of the blood tumor, the first set of samples collectively having a statistically different methylation level than the second set of samples; and determining a cutoff value that distinguishes the first set of samples and the second set of samples within a particular specificity and sensitivity.

27. The computer system of claim 18, wherein the computer system is further configured to perform the following operations: determining the presence of a blood tumor based on a comparison of the methylation levels to one or more cutoff values; and performing a bone marrow biopsy in response to determining the presence of the blood tumor.

28. The computer system of claim 18, wherein the one or more differentially methylated regions comprise CpG sites.

29. The computer system of claim 28, wherein a first region of the one or more differentially methylated regions comprises a plurality of CpG sites within 100 bp of each other, and wherein all of the plurality of CpG sites are either hypomethylated or hypermethylated.

30. The computer system of claim 18, wherein the one or more differentially methylated regions are hypomethylated.

31. The computer system of claim 18, wherein one of the one or more differentially methylated regions is in the FECH gene.

32. The computer system of claim 18, wherein one of the one or more differentially methylated regions is in chromosome 12 at genomic coordinates 48227688-48227701 of the human reference genome hg19.

33. The computer system of claim 18, wherein one of the one or more differentially methylated regions is in chromosome 12 at genomic coordinates 48228144-48228154 of the human reference genome hg19.

34. A method of identifying one or more markers of a particular blood cell lineage, wherein the particular blood cell lineage is red blood cells, the method comprising: analyzing a first set of cell-free DNA molecules from the particular blood cell lineage in a blood sample to determine a first methylation status at one or more sites of each of the first set of cell-free DNA molecules; analyzing a second set of cell-free DNA molecules from other blood cell lineages in the blood sample to determine a second methylation status at one or more sites of each of the second set of cell-free DNA molecules; identifying a region having three or more CpG sites, the region being less than 200 bp; for each CpG site of the three or more CpG sites: determining a first percentage of the first set of cell-free DNA molecules that are methylated at the CpG site; determining a second percentage of the second set of cell-free DNA molecules that are methylated at the CpG site; comparing the first percentage to a first threshold value and comparing the second percentage to a second threshold value; and based on the comparisons, determining that the CpG site is differentially methylated; and based on the three or more CpG sites being all hypomethylated or all hypermethylated in the particular blood cell lineage, determining that the region is differentially methylated.

35. The method of claim 34, wherein analyzing the first set of cell-free DNA molecules comprises methylation-aware sequencing.

36. The method of claim 34, wherein all of the three or more CpG sites are hypomethylated.

37. The method of claim 36, wherein the first percentage is less than 20% of the first threshold and the second percentage is greater than 80% of the second threshold.

38. The method of claim 34, wherein all of the three or more CpG sites are hypermethylated.

39. The method of claim 38, wherein the first percentage is greater than 90% of the first threshold and the second percentage is less than 10% of the second threshold.

40. The method of claim 34, wherein the red blood cells comprise erythroblasts.

41. The method of claim 34, wherein the region is less than 166 bp.

42. The method of claim 34, wherein one of the first threshold and the second threshold is 20% or less.

43. The method of claim 34, wherein one of the first threshold and the second threshold is 80% or more.

44. The method of claim 34, wherein one of the first threshold and the second threshold is 10% or less and the other of the first threshold and the second threshold is 90% or more.

45. The method of claim 34, wherein the region has five or more CpG sites.

46. A computer product comprising a computer readable medium storing a plurality of instructions that, when executed, control a computer system to implement the method of any one of claims 34-45.

47. A system comprising: the computer product of claim 46; and one or more processors to execute the instructions stored on the computer readable medium.

48. A computer product comprising a computer readable medium storing a plurality of instructions for controlling a system to analyze a blood sample of a mammal by: detecting, based on signals obtained from the analysis, a first number of methylated or unmethylated DNA fragments in a cell-free mixture of the blood sample at one or more differentially methylated regions, the cell-free mixture comprising cell-free DNA from a plurality of cell lineages, each of the one or more differentially methylated regions having specificity for a particular blood cell lineage by being hypomethylated or hypermethylated relative to other cell lineages, wherein the particular blood cell lineage is red blood cells; determining a methylation level using the first number; and comparing the methylation level to one or more cutoff values as part of a classification to determine whether the mammal has a blood disorder that is treatment responsive, wherein the one or more cutoff values correspond to a range in which the blood disorder is treatment responsive.

49. The computer product of claim 48, wherein the blood disorder is aplastic anemia.

50. The computer product of claim 49, wherein the treatment is immunosuppressive therapy.

51. The computer product of claim 48, wherein the blood disorder is iron deficiency anemia and the treatment is iron therapy.

52. The computer product of claim 48, wherein the blood disorder is chronic kidney disease.

53. The computer product of claim 48, wherein the plurality of instructions further control the system to perform the following operations: measure a hemoglobin level of the blood sample; compare the hemoglobin level to a hemoglobin threshold; and determine a classification of the blood disorder to which the mammal is therapeutically responsive based further on the comparison of the hemoglobin level to the hemoglobin threshold.

54. The computer product of claim 48, wherein one of the one or more differentially methylated regions is in a FECH gene.

55. The computer product of claim 48, wherein one of the one or more differentially methylated regions is in chromosome 12 at genomic coordinates 48227688-48227701 of human reference genome hg19.

56. The computer product of claim 48, wherein one of the one or more differentially methylated regions is in chromosome 12 at genomic coordinates 48228144-48228154 of human reference genome hg19.

57. The computer product of claim 48, wherein the plurality of instructions further control the system to perform the following operations: determine a total number of DNA fragments in the cell-free mixture at the one or more differentially methylated regions; and determine the methylation level using the first number and the total number.

58. The computer product of claim 48, wherein the plurality of instructions further control the system to perform the following operations: determine a volume of the cell-free mixture, wherein the methylation level is determined using the first number and the volume of the cell-free mixture.

59. The computer product of claim 48, wherein the one or more differentially methylated regions comprise CpG sites.

60. The computer product of claim 59, wherein a first region of the one or more differentially methylated regions comprises a plurality of CpG sites within 100 bp of each other, and wherein the plurality of CpG sites are all either hypo-methylated or hyper-methylated.

61. The computer product of claim 48, wherein the one or more differentially methylated regions are hypo-methylated.

Images