Diagnosing brain disease or brain damage

Abstract

Methods of diagnosing a brain disease or damage or pancreatic disease or damage in a subject comprising receiving methylation data comprising CpA and / or CpT methylation data from cfDNA from a subject and analyzing or quantifying the methylation data is provided. Methods of detecting neuronal DNA or pancreatic islet DNA incfDNA comprising ascertaining the methylation status of at least two CpA and / or CpT dinucleotides and at least two CpC dinucleotides on a continuous sequence of a cfDNA molecule are also provided.

Description

NEURONAL DNA DETECTION BY NON-CPG METHYLATION CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of priority of U.S. Provisional Patent Application No. 63 / 756,301 filed February 10, 2025, the contents of which are incorporated herein by reference in their entirety.FIELD OF INVENTION

[0002] The present invention is in the field of neuronal DNA detection.BACKGROUND OF THE INVENTION

[0003] There is an unmet need for blood biomarkers that can indicate brain damage. In recent years some protein biomarkers specific for Alzheimer’s disease were developed, such as phospho-tau and amyloid beta 42 / 40 ratio, but it is not clear how effective they are for early detection of disease, and they are not relevant for other brain pathologies.

[0004] Liquid biopsies based on circulating cell-free DNA (cfDNA) may indicate the presence of cfDNA fragments that originate in the brain. In patients with brain tumors this can potentially be done via detection of somatic mutations in cfDNA; in brain tumors as well as in other pathologies, epigenetic marks on cfDNA may allow one to determine their brain origin. Such epigenetic marks may include brain-specific CpG methylation patterns, fragmentation patterns typical of brain DNA (normal or pathological) or chromatin marks on cell-free chromatin indicative of brain-specific gene expression.

[0005] However, published reports on cfDNA in brain diseases, using either somatic mutations or epigenetic marks, demonstrate a poor sensitivity which is insufficient for clinical utility in most cases, potentially due to activity of the blood-brain barrier. Neuronal cfDNA in the blood as determined by CpG methylation has been reported previously in some cases of head trauma or relapsing multiple sclerosis (Lehmann-Werman et al, Identification of tissue-specific cell death using methylation patterns of circulating DNA. PNAS 113(13):E1826-34, 2026), in some patients with acute psychosis (Lubotzky et al, Elevated brain-derived cell-free DNA among patients with first psychotic episode - a proof-of-concept study, Elife, doi: 10.7554 / eLife.76391, 2022) and in some patients with brain metastases(Lubotzky et al, Liquid biopsy reveals collateral tissue damage in cancer. JCI Insight doi: 10.1172 / jci.insight.153559., 2022) or after brain irradiation (Makranz et al, Plasma based biomarkers detect radiation induced brain injury in cancer patients treated for brain metastasis: A pilot study. PlosOne doi: 10.1371 / joumal. pone.0285646, 2023). However, sensitivity was poor, and assay performance clearly insufficient for clinical utility. The general consensus in the field is that cfDNA signals originating in the brain are extremely weak. For example, Miller et al (“Tracking tumour evolution in glioma through liquid biopsies of cerebrospinal fluid”, Nature, 2019 Jan;565(7741):654-658) claimed that it was impractical to monitor glioblastoma in liquid biopsies, and sufficient material can only be obtained from cerebrospinal fluid. Harvey et al (“Evaluation of a biomarker for amyotrophic lateral sclerosis derived from a hypomethylated DNA signature of human motor neurons”, BMC Med Genomics. 2025 Jan 14; 18: 10) concluded that it is not possible to detect neuronal cfDNA based on methylation analysis in the plasma of patients with ALS. Consequently, cerebrospinal fluid (CSF) samples are seen as a more informative source of information, however this approach is obviously invasive and less applicable than blood draws. A new accurate test for detecting neuronal DNA and brain diseases or damage is greatly needed.

[0006] Given the low amount of cfDNA in plasma (1000 genome equivalents per ml in healthy people), sensitive detection of cfDNA from a given tissue source requires multiple markers that can be assayed in the same sample. A classic example is the early detection of nasopharyngeal cancer resulting from EBV infection; each dying cancer cell releases 500 EBV DNA fragments, which provide sufficient sensitivity and specificity for detection of disease at an early, actionable stage. Unfortunately, the number of genomic loci showing highly brain-specific CpG methylation is not sufficient to allow for detection of brain cfDNA at relevant concentrations, e.g. when brain-derived cell-free DNA contributes <1% of the circulating cell-free DNA.

[0007] Most DNA methylation occurs on cytosines in the context of CpG, and the detection of cell type-specific CpG methylation patterns is the basis for epigenetic liquid biopsies. However, cytosines can be methylated also in other contexts. It has been recognized previously that cytosines followed by adenosine or thymidine (CpA or CpT) can also be methylated, and that this is more common in neurons compared with other tissues. The genomic distribution of such non-CpG methylation (CpH, H denoting A, T or C)) is not well understood. CpH methylation is thought to take place via activity of the de-novo DNA methyltransferase DNMT3a, and to contribute to gene silencing in the brain, similarly to CpG methylation. The present invention makes use of the abundance of non-CpGmethylation in neurons as a biomarker; in addition, it makes use of our discovery that pancreatic islet cells also show non-CpG methylation at levels that far exceed the baseline in other tissues.SUMMARY OF THE INVENTION

[0008] The present invention provides methods of diagnosing a brain disease or damage or pancreatic disease or damage in a subject comprising receiving methylation data comprising CpA and / or CpT methylation data from cfDNA from a subject and analyzing or quantifying the methylation data is provided. Methods of detecting neuronal DNA or pancreatic islet DNA in cfDNA comprising ascertaining the methylation status of at least two CpA and / or CpT dinucleotides and at least two CpC dinucleotides on a continuous sequence of the same cfDNA molecule are also provided.

[0009] According to a first aspect, there is provided a method of diagnosing a brain disease or brain damage in a subject in need thereof, the method comprising:a. receiving methylation data comprising CpA methylation data, CpT methylation data or both from cell-free DNA (cfDNA) from the subject; andb. i. quantifying the total number of cfDNA molecules comprising at least one methylated CpA and / or CpT dinucleotide, wherein a number of molecules above a predetermined cfDNA threshold indicates the subject suffers from the disease or damage; orii. analyzing the data with a trained machine learning model, wherein the machine learning model is trained on the methylation data from cfDNA from subjects suffering from the brain disease or brain damage and samples from subjects without the brain disease or brain damage and wherein the trained machine learning model outputs a disease diagnosis or a disease score, wherein a disease score above a predetermined disease score threshold indicates the subject suffers from the disease or damage; thereby diagnosing a brain disease or damage.

[0010] According to a first aspect, there is provided a method of diagnosing a pancreatic disease or pancreatic damage in a subject in need thereof, the method comprising:a. receiving methylation data comprising CpA methylation data, CpT methylation data or both from cell-free DNA (cfDNA) from the subject; andb. i. quantifying the total number of cfDNA molecules comprising at least one methylated CpA and / or CpT dinucleotide, wherein a number of molecules above a predetermined cfDNA threshold indicates the subject suffers from the disease or damage; orii. analyzing the data with a trained machine learning model, wherein the machine learning model is trained on the methylation data from cfDNA from subjects suffering from the pancreatic disease or pancreatic damage and samples from subjects without the pancreatic disease or pancreatic damage and wherein the trained machine learning model outputs a disease diagnosis or a disease score, wherein a disease score above a predetermined disease score threshold indicates the subject suffers from the disease or damage;thereby diagnosing a pancreatic disease or damage.[Oil] According to some embodiments, the methylation data further comprises CpC methylation data and step (bii) comprises quantifying the total number of cfDNA molecules comprising at least one methylated CpA or CpT and devoid of a methylated CpC.

[0012] According to some embodiments, step (bii) comprises quantifying the total number of cfDNA molecules comprising at least one methylated CpA, containing at least one CpC dinucleotide and devoid of a methylated CpC.

[0013] According to some embodiments, the received methylation data comprises CpA methylation data and the quantifying is quantifying the total number of cfDNA molecule comprising at least one methylated CpA.

[0014] According to some embodiments, at least 1 methylated CpA is a plurality of methylated CpAs.

[0015] According to some embodiments, the plurality of methylated CpAs is at least 6 methylated CpAs.

[0016] According to some embodiments, the methylation data is whole genome methylation data.

[0017] According to another aspect, there is provided a method of detecting neuronal DNA in cell free DNA (cfDNA), the method comprising ascertaining the methylation status of at least two cytosines in a CpA and / or CpT dinucleotide and the methylation status of at least two cytosines being the first cytosine of a CpC dinucleotide on a continuous sequence of the same cfDNA molecule from the sample, and wherein methylation of the at least two cytosines in a CpA and / or CpT dinucleotide and unmethylation of the at least two cytosines of a CpC dinucleotide indicates the DNA is a neuronal DNA, thereby detecting neuronal DNA in a sample.

[0018] According to another aspect, there is provided a method of detecting pancreatic islet DNA in cell free DNA (cfDNA), the method comprising ascertaining the methylation status of at least two cytosines in a CpA and / or CpT dinucleotide and the methylation status of at least two cytosines being the first cytosine of a CpC dinucleotide on a continuous sequence of the same cfDNA molecule from the sample, and wherein methylation of the at least two cytosines in a CpA and / or CpT dinucleotide and unmethylation of the at least two cytosines of a CpC dinucleotide indicates the DNA is a pancreatic islet DNA, thereby detecting pancreatic islet DNA in a sample.

[0019] According to some embodiments, the cfDNA is from a fluid sample, optionally wherein the fluid is selected from the group consisting of blood, plasma, sperm, milk, urine, saliva and cerebral spinal fluid.

[0020] According to some embodiments, the sample is a blood sample.

[0021] According to some embodiments, the method is a method of detecting neuronal DNA, brain disease or brain damage.

[0022] According to some embodiments, brain disease is a disease characterized by death of neuronal cells.

[0023] According to some embodiments, the method is a method of detecting pancreatic islet DNA, pancreatic disease or pancreatic damage.

[0024] According to some embodiments, the pancreatic islet DNA is beta cell DNA.

[0025] According to some embodiments, the continuous sequence comprises no more than 167 nucleotides.

[0026] According to some embodiments, the at least two cytosines in a CpA or CpT dinucleotide is at least two cytosines in a CpA dinucleotide.

[0027] According to some embodiments, the at least two cytosines in a CpA or CpT dinucleotide is at least four cytosines in a CpA or CpT dinucleotide.

[0028] According to some embodiments, at least four cytosines is at least six cytosines.

[0029] According to some embodiments, the at least two cytosines being the first cytosine of a CpC dinucleotide is all cytosines being the first cytosine of a CpC dinucleotide.

[0030] According to some embodiments, the methylation data is produced by or the ascertaining is affected by contacting the cfDNA with an agent that converts unmethylated cytosines to uracils to produce converted cfDNA and sequencing the converted cfDNA.

[0031] According to some embodiments, the agent is bisulfite.

[0032] According to some embodiments, the agent is a deaminase enzyme.

[0033] According to some embodiments, the methylation data is produced by bisulfite methylation-sequencing (BS-Seq) or by enzymatic methylation sequencing (EM-Seq).

[0034] According to some embodiments, the sequencing is deep sequencing, next generation sequencing or massively parallel sequencing.

[0035] According to some embodiments, the sequencing is shallow sequencing.

[0036] According to some embodiments, the method comprises a limit of detection (LOD) of at least 1 neuronal cfDNA molecule in 40,000 cfDNA molecules.

[0037] According to some embodiments, the methylation data is produced or the ascertaining is affected without converting unmethylated cytosines to uracil and sequencing the cfDNA using single-molecule real-time (SMRT) sequencing detecting both the sequence of the DNA and the methylation status of each cytosine.

[0038] According to some embodiments, the sequencing is at a depth of at least 20K reads.

[0039] According to some embodiments, the methylation data is produced by or the ascertaining is affected without converting unmethylated cytosines to uracil and by passing the DNA through a nanopore and detecting both the sequence of the DNA and the methylation status of each cytosine.

[0040] According to some embodiments, the method further comprises quantitating the number of neuronal or pancreatic islet cfDNA molecules in the sample, wherein a number of molecules above a predetermined cfDNA threshold indicates the sample contains neuronal or pancreatic islet DNA.

[0041] According to some embodiments, the predetermined cfDNA threshold is the total number of cfDNA molecules comprising at least one methylated CpA in cfDNA from subjects that do not suffer from the disease or damage.

[0042] According to some embodiments, the predetermined threshold is greater than 12 parts per million (ppm).

[0043] According to another aspect, there is provided a method of diagnosing a brain disease or damage in a subject in need thereof, the method comprising receiving a fluid sample from the subject and detecting neuronal cell free DNA (cfDNA) in the sample by a method of the invention, wherein the presence of neuronal DNA in the sample indicates the subject suffers from a brain disease or damage.

[0044] According to another aspect, there is provided a method of diagnosing a pancreatic disease or damage in a subject in need thereof, the method comprising receiving a fluid sample from the subject and detecting pancreatic islet cell free DNA (cfDNA) in the sample by a method of the invention, wherein the presence of pancreatic islet DNA in the sample indicates the subject suffers from a pancreatic disease or damage.

[0045] According to some embodiments, the brain disease or damage is selected from brain cancer, brain metastases from non -brain cancer, brain injury or trauma, stroke, brain infection, brain toxicity, autoimmune / inflammatory disease and neurodegenerative disease.

[0046] According to some embodiments, the brain disease is a neurodegenerative disease and is selected from amyotrophic lateral sclerosis (ALS), Alzheimer’s disease, Parkinson’s disease, dementia, multiple sclerosis (MS), X-linked adrenoleukodystrophy (X-ALD) and Huntington’s disease.

[0047] According to some embodiments, the brain disease is Alzheimer’s disease.

[0048] According to some embodiments, the brain disease or damage is selected from MS, brain damage by irradiation, brain cancer and systemic lupus erythematosus. According to some embodiments, the brain cancer is glioblastoma.

[0049] According to some embodiments, the pancreatic disease is selected from pancreatic cancer, pancreatitis, diabetes and damage to transplanted islets.

[0050] According to some embodiments, the method further comprises administering a therapeutic agent to a subject determined to suffer from a brain disease or damage.

[0051] According to some embodiments, the method further comprises administering a therapeutic agent to a subject determined to suffer from a pancreatic disease or damage. /

[0052] Further embodiments and the full scope of applicability of the present invention will become apparent from the detailed description given hereinafter. However, it should be understood that the detailed description and specific examples, while indicating preferred embodiments of the invention, are given by way of illustration only, since various changes and modifications within the spirit and scope of the invention will become apparent to those skilled in the art from this detailed description.BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figures 1A-1C: Global CpH methylation levels across human cell types and correction of technical background. (1A) Global methylation percentages for CpA (blue), CpT (orange), and CpC (red) contexts across bulk brain tissue and sorted cell types. Box plots indicate the median (center line), interquartile range (IQR), and I .5 IQR (whiskers). Sample sizes (n) for each group are: bulk brain (n=12), neurons (n=5), megakaryocytes (n=3), erythroblasts (n=9), hepatocytes (n=ll), blood cells (n=28), endothelial cells (n=l), and epithelial cells (n=5). The horizontal dashed line ("Base") indicates the mean CpH methylation level across all non-brain cell types. (IB) Correlation between apparent CpH methylation measured in non-brain human DNA and spiked-in unmethylated phage Lambda DNA. Each point represents a single sample, including a subset of samples from 1A (n=53, colored) and additional bulk tissue, sorted cell types, and cfDNA samples (n=191, grey) from published and in-house datasets. Dashed lines indicate the line of identity (y=x). Pearson’s R2values are shown for each CpH context. Brain-derived samples are excluded here because their high biological CpH methylation obscures the low-level technical background observed in Lambda DNA. (1C) CpH methylation levels after correction for technical background in the subset of samples from a that include Lambda DNA spike-ins. Corrected values were obtained by subtracting the sample-specific Lambda methylation (technical background) from the corresponding human genomic CpH methylation. The higher median CpH methylation observed in neurons in c relative to a is due to the specific subset of samples containing spike-ins. Box plot elements are defined as in 1A.

[0054] Figures 2A-2D: Fragment-level discrimination of neuronal DNA by clustered methylated CpAs. (2A) Schematic illustrating presumed fragment-level DNA methylation patterns in neurons versus blood, and common technical artifacts. Authentic neuronal CpH methylation typically appears in clustered blocks, whereas artifacts arising from stochastic bisulfite conversion failure appear as isolated events. Molecules resultingfrom failed denaturation (incomplete conversion) mimic fully methylated reads. (2B) Representative sequencing reads from neuronal genomic DNA (top) and blood B cells (bottom) at chr5:78, 018, 675-78, 018, 695, a locus exhibiting high CpA methylation in neurons. Top strand reads. (2C) Number of reads per million (RPM) containing >k methylated CpA sites (mCpAs) in neuronal genomic DNA (blue, n=12 samples) and blood cell types (red, n=164 samples). Solid lines show experimental data; dashed lines show expectations from a generative null model assuming independent CpA methylation probabilities. Error bars indicate s.d. (2D) The neuron-to-blood signal-to-noise ratio as a function of the minimum mCpA threshold. The experimental ratio (solid line) peaks at >6 mCpAs. To exclude artifacts derived from failed denaturation, reads were additionally required to contain >2 unmethylated CpCs and no methylated CpCs.

[0055] Figure 3: Fragment-level CpA methylation (mCpA) levels by cell type. Box plots represent the distribution of fragments containing >6 methylated CpAs, >2 CpCs, and no methylated CpCs per million reads. Neurons exhibit high levels of mCpA (nearly 20,000 reads per million), whereas most other cell types show minimal levels (<1 read per million). Notably, pancreatic endocrine cells (Beta, Alpha, and Delta) display intermediate levels ranging from 100-400 mCpA reads per million. Individual data points represent independent biological replicates; horizontal lines within boxes indicate the median.

[0056] Figures 4A-4F: Ultrasensitive detection of neuronal DNA in simulated plasma and in patient samples. (4A-4B) In-silico mixtures. Neuron-derived methylome was computationally spiked into simulated cfDNA plasma methylome samples with coverage of 50x. The simulated samples were composed of DNA from leukocytes, megakaryocytes, endothelial cells, erythroblasts, and hepatocytes, combined in ratios typical of healthy plasma. Analysis was performed (4A) using the criterion of >6 methylated CpAs sites per sequence read, or (4B) using deconvolution based on the neuronal signal in our CpG methylome atlas. (4C-4D) Neuronal cfDNA signal measured in plasma cfDNA from healthy individuals and patients, using (4C) CpA methylation or (4D) CpG methylation. (4E) In-vitro validation. Plasma from a brain-dead donor (measured neuronal fraction: 0.57%) was mixed into healthy control plasma at ratios of 1:10, 1:50, and 1:250. The x-axis represents the expected neuronal fraction calculated based on the 0.57% baseline. The regression (R=l) is calculated on log-transformed data, excluding the 0% mixture. Horizontal dashed live represents the estimated neuronal fraction in healthy plasma (same sample into which braindead plasma was spiked in). (4F) Limit of detection (LOD) analysis (in-siHco). Performance of the CpA method evaluated at varying sequencing coverages (0.00 lx to 50x), and whenneuronal DNA comprises 0.01% or 0.001% of the total. Statistical significance is calculated relative to the 0% dilution (in panels a+b) or relative to healthy controls (in panels c+d). * p<0.05; ** p<0.01; *** p<0.001; **** p<0.0001 (Wilcoxon rank-sum test).

[0057] Figure 5: Levels of reads with >4 methylated CpA in plasma cfDNA from patients with ALS vs. controls. Box and whisker plot of cfDNA molecules that have four or more methylated CpA, no methylated CpC and at least two CpC dinucleotides. The WGBS cell-free DNA samples (Caggiano et al.) were sequenced with 75bp paired-end sequencing, which results in lower coverage of each fragment.

[0058] Figures 6A-6B: Graphs showing the detection of neuron-derived fragments in in-silico simulations using (6A) CpG-based deconvolution as described at Loyfer et al. and (6B) using the 6 methylated CpA method of the invention.DETAILED DESCRIPTION OF THE INVENTION

[0059] The present invention, in some embodiments, provides methods of diagnosing a brain disease or damage or pancreatic disease or damage in a subject. Methods of detecting neuronal DNA or pancreatic islet DNA in cfDNA are also provided.

[0060] The invention is based, at least in part, on the surprising discovery that CpA and CpT methylation is restricted to only neurons and pancreatic islets. In the past it was believed that there are low levels of CpA / CpT methylation in other tissues making it an unreliable biomarker unless specific brain-related loci were examined. It was also thought that CpC methylation occurred in neurons. As demonstrated hereinbelow, both of these beliefs are erroneous. The background CpA / CpT methylation observed in other tissues is in fact artifactual, a result of incomplete conversion of unmethylated cytosines to uracil. In actuality, CpA / CpT methylation is so rare in non-neuronal / non-pancreatic cells that this methylation is a viable biomarker for identifying cfDNA from the brain / pancreatic islets and for diagnosing brain / pancreatic disease and damage. Further, because CpC methylation, even in the brain / pancreas, is almost always artifactual, this artifactual noise can be removed by excluding molecules that are read after sequencing as having “methylated” CpC dinucleotides. This removal increases the accuracy of identifying neuronal / pancreatic DNA as an indication of disease, however the methylation biomarker is so specific that just relying on CpA / CpT methylation is sufficient.

[0061] The instant invention greatly improves the technology of cfDNA detection and DNA methylation analysis in general. The technological field is often hindered by the complexity and variation of CpG methylation in different cells and tissues, however, the instant discovery that CpA and CpT methylation is nearly exclusively a brain phenomenon (with the exception of the pancreatic islets) allows for a greatly improved assay that does not rely on the specific genomic location being investigated, but rather determining that the methylation is genuine and not an artifact. Indeed, the invention can also be viewed as an improvement in bisulfite conversion technology and / or enzymatic conversion technology, as it allows for the filtering out of artifacts of conversion and identifying only genuine, and therefore neuronal, CpA and CpT methylation.

[0062] By a first aspect, there is provided a method of detecting neuronal DNA, the method comprising ascertaining the methylation status of a cytosine in a CpA and / or CpT dinucleotide in a DNA sequence, wherein methylation of said cytosine in a CpA and / or CpT dinucleotide indicates the DNA is a neuronal DNA molecule, thereby detecting neuronal DNA.

[0063] By another aspect, there is provided a method of detecting pancreatic islet DNA, the method comprising ascertaining the methylation status of a cytosine in a CpA and / or CpT dinucleotide in a DNA sequence, wherein methylation of said cytosine in a CpA and / or CpT dinucleotide indicates the DNA is a pancreatic islet DNA molecule, thereby detecting pancreatic islet DNA.

[0064] By another aspect, there is provided a method of detecting neuronal or pancreatic islet DNA, the method comprising ascertaining the methylation status of a cytosine in a CpA and / or CpT dinucleotide in a DNA sequence, wherein methylation of said cytosine in a CpA and / or CpT dinucleotide indicates the DNA is a neuronal or pancreatic islet DNA molecule, thereby detecting neuronal or pancreatic islet DNA.

[0065] By another aspect, there is provided a method of diagnosing a brain disease or brain damage in a subject, the method comprising:a. receiving methylation data from cell free DNA (cfDNA) from the subj ect;andb. analyzing the data with a trained machine learning model, wherein the trained machine learning model outputs a disease diagnosis or a disease score;thereby diagnosing a brain disease or brain damage.

[0066] By another aspect, there is provided a method of diagnosing a pancreatic disease or pancreatic damage in a subject, the method comprising:a. receiving methylation data from cell free DNA (cfDNA) from the subj ect;andb. analyzing the data with a trained machine learning model, wherein the trained machine learning model outputs a disease diagnosis or a disease score;thereby diagnosing a pancreatic disease or pancreatic damage.

[0067] By another aspect, there is provided a method of diagnosing a brain disease or brain damage or a pancreatic disease or pancreatic damage in a subject, the method comprising:a. receiving methylation data from cell free DNA (cfDNA) from the subj ect;andb. analyzing the data with a trained machine learning model, wherein the trained machine learning model outputs a disease diagnosis or a disease score;thereby diagnosing a brain disease or brain damage or a pancreatic disease or pancreatic damage.

[0068] By another aspect, there is provided a method of diagnosing a brain disease or brain damage in a subject, the method comprising:a. receiving methylation data from cell free DNA (cfDNA) from the subj ect;andb. quantifying the number of cfDNA molecules that are neuronal cfDNA molecules wherein a number of molecules above a predetermined threshold indicates the subject suffers from a brain disease or brain damage;thereby diagnosing a brain disease or brain damage.

[0069] By another aspect, there is provided a method of diagnosing a pancreatic disease or pancreatic damage in a subject, the method comprising:a. receiving methylation data from cell free DNA (cfDNA) from the subj ect; andb. quantifying the number of cfDNA molecules that are neuronal cfDNA molecules wherein a number of molecules above a predetermined threshold indicates the subject suffers from a pancreatic disease or pancreatic damage;thereby diagnosing a pancreatic disease or pancreatic damage.

[0070] By another aspect, there is provided a method of diagnosing a brain disease or brain damage or a pancreatic disease or pancreatic damage in a subject, the method comprising:a. receiving methylation data from cell free DNA (cfDNA) from the subj ect;andb. quantifying the number of cfDNA molecules that are neuronal cfDNA molecules wherein a number of molecules above a predetermined threshold indicates the subject suffers from a brain disease or brain damage or a pancreatic disease or pancreatic damage;thereby diagnosing a brain disease or brain damage a pancreatic disease or pancreatic damage.

[0071] By another aspect, there is provided a method of diagnosing a brain disease or brain damage in a subject, the method comprising:a. receiving methylation data from cell free DNA (cfDNA) from the subj ect;andb. quantifying the number of cfDNA molecules comprising at least one methylated CpA and / or CpT dinucleotide, wherein a number of molecules above a predetermined threshold indicates the subject suffers from a brain disease or brain damage;thereby diagnosing a brain disease or brain damage.

[0072] By another aspect, there is provided a method of diagnosing a pancreatic disease or pancreatic damage in a subject, the method comprising:a. receiving methylation data from cell free DNA (cfDNA) from the subj ect;andb. quantifying the number of cfDNA molecules comprising at least one methylated CpA and / or CpT dinucleotide, wherein a number of molecules above a predetermined threshold indicates the subject suffers from a pancreatic disease or pancreatic damage;thereby diagnosing a pancreatic disease or pancreatic damage.

[0073] By another aspect, there is provided a method of diagnosing a brain disease or brain damage or a pancreatic disease or pancreatic damage in a subject, the method comprising:a. receiving methylation data from cell free DNA (cfDNA) from the subj ect;andb. quantifying the number of cfDNA molecules comprising at least one methylated CpA and / or CpT dinucleotide, wherein a number of molecules above a predetermined threshold indicates the subject suffers from a brain disease or brain damage or a pancreatic disease or pancreatic damage;thereby diagnosing a brain disease or brain damage or a pancreatic disease or pancreatic damage.

[0074] By another aspect, there is provided a method of detecting a DNA molecule with genuine CpW methylation, the method comprising:a. receiving methylation data comprising CpH methylation data from a plurality of DNA molecule; andb. identifying at least one DNA molecule of the plurality comprising at least one methylated CpW and devoid of a methylated CpC, wherein the identified DNA molecule is a DNA molecule with genuine CpW methylation,thereby detecting a DNA molecule with genuine CpW methylation.

[0075] In some embodiments, the method is an in vitro method. In some embodiments, the method is an ex vivo method. In some embodiments, the method is a diagnostic method. In some embodiments, the quantifying neuronal DNA molecules is by a method of the invention. In some embodiments, the quantifying pancreatic DNA molecules is by a method of the invention. In some embodiments, the method is a method of detecting neuronal DNA. In some embodiments, the method is a method of detecting pancreatic DNA. In some embodiments, pancreatic DNA is pancreatic islet DNA. In some embodiments, the methodis a method of diagnosing a brain disease orbrain damage. In some embodiments, the method is a method of diagnosing a pancreatic disease or pancreatic damage. In some embodiments, the method is a method of early detection of a brain disease or brain damage. In some embodiments, the method is a method of early detection of a pancreatic disease or pancreatic damage. In some embodiments, the method is a method of monitoring progression of a brain disease or brain damage. In some embodiments, the method is a method of monitoring progression of a pancreatic disease or pancreatic damage. In some embodiments, the method is a method of monitoring response to therapy of a brain disease or brain damage. In some embodiments, the method is a method of monitoring response to therapy of a pancreatic disease or pancreatic damage. In some embodiments, a pancreatic disease is a pancreatic islet disease. In some embodiments, pancreatic damage is pancreatic islet damage. In some embodiments, pancreatic islet is beta cells. In some embodiments, a pancreatic disease is a beta cell disease. In some embodiments, pancreatic damage is beta cell damage. In some embodiments, the method is a method of detecting genuine CpW methylation. In some embodiments, genuine methylation is not an artifact. In some embodiments, an artifact is an artifact of bisulfite conversion. In some embodiments, an artifact is an artifact of enzymatic conversion. In some embodiments, an artificial is an artifact of conversion of unmethylated cytosines to uracils. In some embodiments, the method is a method of improving bisulfite conversion. In some embodiments, the method is a method of improving bisulfite conversion analysis. In some embodiments, the method is a method of improving enzymatic conversion. In some embodiments, the method is a method of improving enzymatic conversion analysis. In some embodiments, the method is a method of improving cfDNA detection methods. In some embodiments, the method is a method of improving detection of CpA, CpT or both methylation. In some embodiments, the method is a method of improving methylationsequencing (methyl-seq). In some embodiments, the method is a method of improving bisulfite methyl-seq. In some embodiments, the method is a method of improving enzymatic methyl-seq (EM-seq).

[0076] In some embodiments, the subject is a subject in need of a method of the invention. In some embodiments, the subject is a mammal. In some embodiments, the mammal is a human. In some embodiments, the subject is suspected of suffering from a brain disease or damage. In some embodiments, the subject is suspected of suffering from a pancreatic disease or damage. In some embodiments, the subject is not suspected of suffering from both a brain disease or damage and a pancreatic disease or damage. In some embodiments, the subject is at risk of developing a brain disease or damage. In some embodiments, the subjectis at risk of developing a pancreatic disease or damage. In some embodiments, the subject is not at risk of developing both a brain disease or damage and a pancreatic disease or damage. In some embodiments, the subject suffered a head injury. In some embodiments, the subject is elderly. In some embodiments, an elderly subject is a subject over 65, 70, 75, 80, 85, or 90 years of age. Each possibility represents a separate embodiment of the invention, some embodiments, an elderly subject is a subject over 65 years of age. In some embodiments, the subject is obese. In some embodiments, the subject has a body mass index (BMI) of 25 or over. In some embodiments, the subject has a body mass index (BMI) of 30 or over. In some embodiments, the elderly subject is at risk of the brain disease. In some embodiments, the obese subject is at risk of the pancreatic disease. In some embodiments, the method further comprises selecting the subject.

[0077] In some embodiments, a neuronal cell is a neuron. In some embodiments, a neuronal cell is a brain cell. In some embodiments, a pancreatic cell is a pancreatic islet cell. In some embodiments, a pancreatic cell is a beta cell. In some embodiments, the method is a method of detecting neuron death. In some embodiments, the method is a method of detecting beta cell death. In some embodiments, the method is a method of detecting neuronal DNA in cfDNA. In some embodiments, the method is a method of detecting pancreatic DNA in cfDNA. In some embodiments, pancreatic DNA is pancreatic islet DNA.

[0078] In some embodiments, the detecting is in the subject. In some embodiments, the detecting is in a sample. In some embodiments, the ascertaining is in a sample. In some embodiments, the sample is from the subject. In some embodiments, the method comprises receiving a sample. In some embodiments, the method comprises taking a sample. In some embodiments, the sample comprises DNA. In some embodiments, the DNA is cell free-DNA (cfDNA). In some embodiments, the DNA molecule is in the sample. In some embodiments, the sample is a fluid sample. In some embodiments, the fluid is a bodily fluid. In some embodiments, the bodily fluid is selected from: blood, serum, plasma, gastric fluid, intestinal fluid, saliva, bile, tumor fluid, breast milk, urine, interstitial fluid, cerebral spinal fluid and stool. In some embodiments, the bodily fluid is selected from: blood, plasma, sperm, milk, urine, saliva and cerebral spinal fluid. In some embodiments, the bodily fluid is selected from: blood, plasma, urine, saliva and cerebral spinal fluid. In some embodiments, the sample is a blood sample. In some embodiments, blood is whole blood. In some embodiments, the sample is a plasma sample. In some embodiments, the DNA is circulating DNA. In some embodiments, the sample is purified or isolated DNA. In some embodiments, the DNA is purified or isolated from the fluid.

[0079] In some embodiments, the blood sample comprises at least 0.001, 0.01, 0.1, 0.5, 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 ml of blood. Each possibility represents a separate embodiment of the invention. In some embodiments, the blood sample comprises at least 5 ml of blood. In some embodiments, the blood sample comprises at most 0.001, 0.01, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 or 20 ml of blood. Each possibility represents a separate embodiment of the invention. In some embodiments, the blood sample comprises 5-10 ml of blood. In some embodiments, the sample comprises at least 3 picogram, 6 picogram, 50 picogram, 500 picogram, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 or 20 nanogram of DNA. Each possibility represents a separate embodiment of the invention. In some embodiments, the sample comprises at most 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 25, 30, 35, 40, 45, 50 ng of DNA. Each possibility represents a separate embodiment of the invention.

[0080] In some embodiments, the method comprises ascertaining the methylation status of a CpA. As used herein, the notation CpX refers to a dinucleotide in which the first nucleotide is a cytosine. The four possible dinucleotides are CpA, CpT, CpC and CpG. Thus, the terms CpX and CpX dinucleotide can be used interchangeably and have the same meaning. It will be understood that the methylation status of the cytosine is what is being ascertained, measured or analyzed. Further, in the case of CpC dinucleotides it will be understood that the methylation status of the first cytosine is being ascertained, measured or analyzed. In some embodiments, the method comprises ascertaining the methylation status of a CpH. According to the single letter nucleotide code, “H” refers to A, T or C. In some embodiments, the method comprises ascertaining the methylation status of a CpW. According to the single letter nucleotide code, “W” refers to A or T. In some embodiments, the method comprises ascertaining the methylation status of a CpT. In some embodiments, the method comprises ascertaining the methylation status of a CpA. In some embodiments, the method comprises ascertaining the methylation status of a CpC. In some embodiments, the method comprises ascertaining the methylation status of a CpA and / or a CpT. In some embodiments, the method comprises ascertaining the methylation status of a CpA, CpT and / or CpC. In some embodiments, the method comprises ascertaining the methylation status of a CpA and a CpC on the same molecule of DNA. In some embodiments, the same molecule of DNA is a single molecule of DNA. In some embodiments, the same molecule of DNA is the same physical molecule of DNA. In some embodiments, the method comprises ascertaining the methylation status of a CpA and / or CpT and a CpC on the same molecule of DNA. In some embodiments, a CpA, CpT and / or CpC is at least one CpA, CpT and / or CpC. In someembodiments, at least one is a plurality. In some embodiments, at least one is at least 2, 3, 4, 5, 6, 7 or 8. Each possibility represents a separate embodiment of the invention. In some embodiments, the ascertaining is ascertaining the methylation status of at least two CpA and / or CpT dinucleotides on the same molecule of DNA. It will be understood that this can refer to two CpAs, two CpTs or 1 CpA and 1 CpT. In some embodiments, the ascertaining is ascertaining the methylation status of at least 3 CpA and / or CpT dinucleotides on the same molecule of DNA. In some embodiments, the ascertaining is ascertaining the methylation status of at least 4 CpA and / or CpT dinucleotides on the same molecule of DNA. In some embodiments, the ascertaining is ascertaining the methylation status of at least 6 CpA and / or CpT dinucleotides on the same molecule of DNA. In some embodiments, CpA and / or CpT dinucleotides is CpA dinucleotides. In some embodiments, the ascertaining is ascertaining the methylation status of at least two CpCs on the same molecule of DNA. In some embodiments, for detecting pancreatic islet DNA, or diagnosing a pancreatic disease or damage CpA methylation is used. In some embodiments, the CpA is followed by a G base (trinucleotide CAG). In some embodiments, the CpA is followed by an H base (A, T or C, trinucleotide CAH). In some embodiments, the CpT is followed by a G base (trinucleotide CTG). In some embodiments, the CpT is followed by an H base (A, T or C, trinucleotide CTH).

[0081] In some embodiments, methylation status is methylated or unmethylated. In some embodiments, the methylation status is ascertained of a dinucleotide in a DNA sequence. In some embodiments, the methylation status is ascertained of a dinucleotide in a DNA molecule. In some embodiments, the sequence is in a molecule. In some embodiments, the methylation status is ascertained in a continuous sequence of a DNA molecule. In some embodiments, the methylation status is ascertained in a continuous sequence of the same physical DNA molecule. In some embodiments, the methylation status is ascertained in a continuous sequence of a single DNA molecule. In some embodiments, the status of more than one dinucleotide is ascertained on the same DNA molecule. In some embodiments, the status of more than one dinucleotide is ascertained in a continuous sequence of the same DNA molecule. Thus, the multiple methylated CpA / CpTs must all be on the same molecule and then that molecule is considered as being neuronal or pancreatic.

[0082] In some embodiments, the continuous sequence comprises no more than 167 nucleotides. The continuous sequence is preferably no longer than 300 nucleotides, 295 nucleotides, 290 nucleotides, 285 nucleotides, 280 nucleotides, 275 nucleotides, 270 nucleotides, 265 nucleotides, 260 nucleotides, 255 nucleotides, 250 nucleotides, 245nucleotides, 240 nucleotides, 235 nucleotides, 230 nucleotides, 225 nucleotides, 220 nucleotides, 215 nucleotides, 210 nucleotides, 205 nucleotides, 200 nucleotides, 195 nucleotides, 190 nucleotides, 185 nucleotides, 180 nucleotides, 175 nucleotides, 170 nucleotides, 167 nucleotides, 165 nucleotides, 160 nucleotides, 155 nucleotides, 150 nucleotides, 145 nucleotides, 140 nucleotides, 135 nucleotides, 130 nucleotides, 125 nucleotides, 120 nucleotides, 115 nucleotides, 110 nucleotides, 105 nucleotides, 100 nucleotides, 95 nucleotides, 90 nucleotides, 85 nucleotides, 80 nucleotides, 75 nucleotides, 70 nucleotides, 65 nucleotides, 60 nucleotides, 55 nucleotides, 50 nucleotides, 45 nucleotides, 40 nucleotides, 35 nucleotides, 30 nucleotides, 25 nucleotides, or 20 nucleotides. In some embodiments, the continuous sequence is between 50-300 nucleotide, 50-250 nucleotides, 50-200 nucleotides, 100-300 nucleotides, 100-250 nucleotides, 100-200 nucleotides, 125-300 nucleotides, 125 to 150 nucleotides or 125-200 nucleotides.

[0083] In some embodiments, the methylation data comprises whole genome methylation. In some embodiments, the methylation data comprises CpA methylation. In some embodiments, the methylation data consists of CpA methylation. In some embodiments, the methylation data comprises CpT methylation. In some embodiments, the methylation data consists of CpT methylation. In some embodiments, the methylation data comprises CpW methylation. In some embodiments, the methylation data comprises CpH methylation. In some embodiments, the methylation data comprises CpA and CpT methylation. In some embodiments, the methylation data consists of CpA and CpT methylation. In some embodiments, the methylation data comprises CpA and CpC methylation. In some embodiments, the methylation data consists of CpA and CpC methylation. In some embodiments, the methylation data comprises CpT and CpC methylation. In some embodiments, the methylation data consists of CpT and CpC methylation. In some embodiments, the methylation data comprises CpG methylation. In some embodiments, the methylation data comprises CpA, CpT, CpC and CpG methylation data. In some embodiments, the methylation data consists of CpA, CpT, CpC and CpG methylation data.

[0084] In some embodiments, methylation of at least one CpA in a DNA molecule indicates the molecule is a neuronal DNA. In some embodiments, methylation of at least one CpT in a DNA molecule indicates the molecule is a neuronal DNA. In some embodiments, methylation of at least one CpA or CpT in a DNA molecule indicates the molecule is a neuronal DNA. In some embodiments, methylation of at least one CpA in a DNA molecule indicates the molecule is a pancreatic islet DNA. In some embodiments, methylation of at least one CpT in a DNA molecule indicates the molecule is a pancreatic islet DNA. In someembodiments, methylation of at least one CpA or CpT in a DNA molecule indicates the molecule is a pancreatic islet DNA. In some embodiments, methylation of at least one CpA in a DNA molecule indicates the molecule is a neuronal or pancreatic islet DNA. In some embodiments, methylation of at least one CpT in a DNA molecule indicates the molecule is a neuronal or pancreatic islet DNA. In some embodiments, methylation of at least one CpA or CpT in a DNA molecule indicates the molecule is a neuronal or pancreatic islet DNA.

[0085] In some embodiments, methylation of at least two CpAs in the same DNA molecule indicates the molecule is a neuronal DNA. In some embodiments, methylation of at least two CpTs in the same DNA molecule indicates the molecule is a neuronal DNA. In some embodiments, methylation of at least two CpA and / or CpT in the same DNA molecule indicates the molecule is a neuronal DNA. In some embodiments, methylation of at least two CpAs in the same DNA molecule indicates the molecule is a pancreatic islet DNA. In some embodiments, methylation of at least two CpTs in the same DNA molecule indicates the molecule is a pancreatic islet DNA. In some embodiments, methylation of at least two CpA and / or CpT in the same DNA molecule indicates the molecule is a pancreatic islet DNA. In some embodiments, methylation of at least two CpAs in the same DNA molecule indicates the molecule is a neuronal or pancreatic islet DNA. In some embodiments, methylation of at least two CpTs in the same DNA molecule indicates the molecule is a neuronal or pancreatic islet DNA. In some embodiments, methylation of at least two CpA and / or CpT in the same DNA molecule indicates the molecule is a neuronal or pancreatic islet DNA.

[0086] In some embodiments, methylation of at least four CpAs in the same DNA molecule indicates the molecule is a neuronal DNA. In some embodiments, methylation of at least four CpTs in the same DNA molecule indicates the molecule is a neuronal DNA. In some embodiments, methylation of at least four CpA and / or CpT in the same DNA molecule indicates the molecule is a neuronal DNA. In some embodiments, methylation of at least four CpAs in the same DNA molecule indicates the molecule is a pancreatic islet DNA. In some embodiments, methylation of at least four CpTs in the same DNA molecule indicates the molecule is a pancreatic islet DNA. In some embodiments, methylation of at least four CpA and / or CpT in the same DNA molecule indicates the molecule is a pancreatic islet DNA. In some embodiments, methylation of at least four CpAs in the same DNA molecule indicates the molecule is a neuronal or pancreatic islet DNA. In some embodiments, methylation of at least four CpTs in the same DNA molecule indicates the molecule is a neuronal or pancreatic islet DNA. In some embodiments, methylation of at least four CpAand / or CpT in the same DNA molecule indicates the molecule is a neuronal or pancreatic islet DNA.

[0087] In some embodiments, methylation of at least six CpAs in the same DNA molecule indicates the molecule is a neuronal DNA. In some embodiments, methylation of at least six CpTs in the same DNA molecule indicates the molecule is a neuronal DNA. In some embodiments, methylation of at least six CpA and / or CpT in the same DNA molecule indicates the molecule is a neuronal DNA. In some embodiments, methylation of at least six CpAs in the same DNA molecule indicates the molecule is a pancreatic islet DNA. In some embodiments, methylation of at least six CpTs in the same DNA molecule indicates the molecule is a pancreatic islet DNA. In some embodiments, methylation of at least six CpA and / or CpT in the same DNA molecule indicates the molecule is a pancreatic islet DNA. In some embodiments, methylation of at least six CpAs in the same DNA molecule indicates the molecule is a neuronal or pancreatic islet DNA. In some embodiments, methylation of at least six CpTs in the same DNA molecule indicates the molecule is a neuronal or pancreatic islet DNA. In some embodiments, methylation of at least six CpA and / or CpT in the same DNA molecule indicates the molecule is a neuronal or pancreatic islet DNA.

[0088] Scoring of molecules is possible only if they contain the dinucleotide that is being investigated, and indeed the requisite number of that dinucleotide if molecules with more than one are being scored. Thus, a prefiltering step is possible in which molecules without the requisite number of CpA, CpT or CpC dinucleotides are removed. In some embodiments, the DNA molecule comprises at least one CpA. In some embodiments, the DNA molecule comprises at least one CpT. In some embodiments, the DNA molecule comprises at least one CpA or CpT. In some embodiments, the DNA molecule comprises at least two CpAs. In some embodiments, the DNA molecule comprises at least two CpTs. In some embodiments, the DNA molecule comprises at least two CpA and / or CpT. In some embodiments, the DNA molecule comprises at least four CpAs. In some embodiments, the DNA molecule comprises at least four CpTs. In some embodiments, the DNA molecule comprises at least four CpA and / or CpT. In some embodiments, the DNA molecule comprises at least sxi CpAs. In some embodiments, the DNA molecule comprises at least six CpTs. In some embodiments, the DNA molecule comprises at least six CpA and / or CpT. In some embodiments, DNA molecules without the required number of CpAs / CpTs are excluded from the analysis.

[0089] In some embodiments, a molecule with methylated CpA / CpT must also have at least one unmethylated CpC. In some embodiments, at least one unmethylated CpC is at least twounmethylated CpCs. In some embodiments, the molecule comprises at least 1 CpC. In some embodiments, the molecule comprises at least 2 CpCs. In some embodiments, the molecule is devoid of methylated CpCs. In some embodiments, all CpCs of the molecule are unmethylated. In some embodiments, the molecule comprises at least 2 CpCs and all CpCs are unmethylated. In some embodiments, molecules without at least 1 CpC are excluded from the analysis. In some embodiments, molecules without at least 2 CpCs are excluded from the analysis. In some embodiments, a molecule comprising a methylated CpA and / or CpT and devoid of methylated CpCs is a neuronal DNA. In some embodiments, a molecule comprising a methylated CpA and / or CpT and devoid of methylated CpCs is a pancreatic DNA. In some embodiments, a molecule comprising a methylated CpA and / or CpT and devoid of methylated CpCs is a neuronal or pancreatic DNA. In some embodiments, the molecule is both devoid of unmethylated CpCs and comprises at least two unmethylated CpC. In some embodiments, the molecule is both devoid of unmethylated CpCs and comprises at least one unmethylated CpC.

[0090] In some embodiments, a DNA molecule with genuine CpW methylation comprises at least one methylated CpW and is devoid of unmethylated CpC. In some embodiments, a DNA molecule with genuine CpW methylation comprises at least two methylated CpW and is devoid of unmethylated CpC. In some embodiments, a DNA molecule with genuine CpW methylation comprises at least three methylated CpW and is devoid of unmethylated CpC. In some embodiments, a DNA molecule with genuine CpW methylation comprises at least four methylated CpW and is devoid of unmethylated CpC. In some embodiments, a DNA molecule with genuine CpW methylation comprises at least five methylated CpW and is devoid of unmethylated CpC. In some embodiments, a DNA molecule with genuine CpW methylation comprises at least six methylated CpW and is devoid of unmethylated CpC. In some embodiments, a DNA molecule with genuine CpW methylation comprises at least one methylated CpW and at least one unmethylated CpC. In some embodiments, a DNA molecule with genuine CpW methylation comprises at least two methylated CpW and at least one unmethylated CpC. In some embodiments, a DNA molecule with genuine CpW methylation comprises at least three methylated CpW and at least one unmethylated CpC. In some embodiments, a DNA molecule with genuine CpW methylation comprises at least four methylated CpW and at least one unmethylated CpC. In some embodiments, a DNA molecule with genuine CpW methylation comprises at least five methylated CpW and at least one unmethylated CpC. In some embodiments, a DNA molecule with genuine CpW methylation comprises at least six methylated CpW and at least one unmethylated CpC. Insome embodiments, a DNA molecule with genuine CpW methylation comprises at least one methylated CpW and at least two unmethylated CpC. In some embodiments, a DNA molecule with genuine CpW methylation comprises at least two methylated CpW and at least two unmethylated CpC. In some embodiments, a DNA molecule with genuine CpW methylation comprises at least three methylated CpW and at least two unmethylated CpC. In some embodiments, a DNA molecule with genuine CpW methylation comprises at least four methylated CpW and at least two unmethylated CpC. In some embodiments, a DNA molecule with genuine CpW methylation comprises at least five methylated CpW and at least two unmethylated CpC. In some embodiments, a DNA molecule with genuine CpW methylation comprises at least six methylated CpW and at least two unmethylated CpC. In some embodiments, a DNA molecule with genuine CpW methylation is devoid of unmethylated CpCs and comprises at least one unmethylated CpC. In some embodiments, a DNA molecule with genuine CpW methylation is devoid of unmethylated CpCs and comprises at least two unmethylated CpC.

[0091] While whole genome methylation is exemplified methods using only PCR of specific regions are also envisioned. Multi-copy sites or repetitive regions (i.e., Alu repeats or LINE elements) also can contain CpW methylation in neurons and pancreatic endocrine cells and so amplification of these regions followed by their sequencing is sufficient for CpW methylation detection (and also CpC). In such cases genome level data is not necessary. Further, a methylation chip or array that can detect CpH methylation can be used to produce the methylation data without sequencing.

[0092] In some embodiments, the quantifying is quantifying the total number of molecules that are neuronal molecules. In some embodiments, neuronal molecules are determined by a method of a the invention. In some embodiments, the quantifying is quantifying the total number of molecules that are pancreatic molecules. In some embodiments, pancreatic molecules are determined by a method of a the invention. In some embodiments, the quantifying is quantifying the total number of cfDNA molecules comprising at least one methylated CpA. In some embodiments, the quantifying is quantifying the total number of cfDNA molecules comprising at least one methylated CpT. In some embodiments, the quantifying is quantifying the total number of cfDNA molecules comprising at least one methylated CpA and / or CpT. In some embodiments, the quantifying is quantifying the total number of cfDNA molecules comprising at least one methylated CpA and at least one unmethylated CpC in the same molecule. In some embodiments, the quantifying is quantifying the total number of cfDNA molecules comprising at least one methylated CpTand at least one unmethylated CpC in the same molecule. In some embodiments, the quantifying is quantifying the total number of cfDNA molecules comprising at least one methylated CpA and / or CpT and at least one unmethylated CpC in the same molecule. In some embodiments, the quantifying is quantifying the total number of cfDNA molecules comprising at least one methylated CpA that are devoid of a methylated CpC. In some embodiments, the quantifying is quantifying the total number of cfDNA molecules comprising at least one methylated CpT that are devoid of a methylated CpC. In some embodiments, the quantifying is quantifying the total number of cfDNA molecules comprising at least one methylated CpA and / or CpT that are devoid of a methylated CpC. In some embodiments, at least one is at least 2. In some embodiments, at least one is at least 4. In some embodiments, at least one is at least 6. In some embodiments, the quantified molecules comprise at least one CpC dinucleotide. In some embodiments, the quantified molecules comprise at least two CpC dinucleotides.

[0093] In some embodiments, the method comprises quantitating the number of neuronal DNA molecules in the sample. In some embodiments, the method comprises quantitating the number of pancreatic islet DNA molecules in the sample. In some embodiments, the method comprises quantitating the number of neuronal and / or pancreatic islet DNA molecules in the sample. In some embodiments, a number of cfDNA molecules above a predetermined threshold indicates the subject suffers from a brain disease or damage. In some embodiments, a number of cfDNA molecules above a predetermined threshold indicates the subject suffers from a pancreatic disease or damage. In some embodiments, a number of cfDNA molecules above a predetermined threshold indicates the subject suffers from a brain disease or damage or pancreatic disease or damage. In some embodiments, above is significantly above. In some embodiments, above is statistically significantly above. In some embodiments, above is at least 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 120, 125, 150, 175, 180, or 200% above. Each possibility represents a separate embodiment of the invention.

[0094] In some embodiments, the predetermined threshold is the number of neuronal cfDNA molecules in cfDNA from a control subject. In some embodiments, the predetermined threshold is the number of pancreatic islet cfDNA molecules in cfDNA from a control subject. In some embodiments, the predetermined threshold is the number of neuronal and / or pancreatic islet cfDNA molecules in cfDNA from a control subject. In some embodiments, in cfDNA from a control subject is in a sample from a control subject. The cfDNA molecules that are neuronal / pancreatic islet are as determined hereinabove (e.g., at least one methylatedCpA / CpT, devoid of methylated CpC, etc.). In some embodiments, a control subject is a healthy subject. In some embodiments, a control subject is a subject that does not suffer from a neuronal disease or damage. In some embodiments, a control subject is a subject that does not suffer from a pancreatic disease or damage. In some embodiments, a control subject is a subject that does not suffer from a neuronal disease or damage or pancreatic disease or damage.

[0095] In some embodiments, the predetermined threshold is at least 12 parts per million (ppm). It will be understood that when the total number of cfDNA molecules are qualified and the total number of neuronal / pancreatic islet cfDNA molecules are quantified then they can be expressed as ppm. In some embodiments, parts per million is molecules per million (mpm). In some embodiments, the million is 1 million included molecules in the analysis and does not comprise excluded molecules (e.g., those without the required number of CpA / CpT / CpC dinucleotides). In In some embodiments, the predetermined threshold is great than 12 ppm. In some embodiments, the predetermined threshold is at or above 12, 13, 14, 15, 16, 17, 18, 19, 20, 25, 30, 35, 40, 45, 50, 55, 60 or 65 ppm. Each possibility represents a separate embodiment of the invention. In some embodiments, the predetermined threshold is between 12 and 69 ppm. In some embodiments, the predetermined threshold is between 12 and 65 ppm. In some embodiments, the predetermined threshold is between 12 and 69, 12 and 65, 1 and 60, 12 and 55, 12 and 50, 15 and 69, 15 and 65, 15 and 60, 15 and 55, 15 and 50, 20 and 69, 20 and 65, 20 and 60, 20 and 55, 20 and 50, 25 and 69, 25 and 65, 25 and 60, 25 and 55 or 25 and 50. Each possibility represents a separate embodiment of the invention.

[0096] In some embodiments, the brain disease or damage is a brain disease. In some embodiments, the brain disease or damage is characterized by neuronal cell death. In some embodiments, the brain disease or damage is characterized by neuronal cfDNA in the subject. In some embodiments, in the subject is in circulation. In some embodiments, the brain disease or damage is selected from brain cancer, brain metastases from non-brain cancer, brain injury or trauma, stroke, brain infection, psychiatric disease, autoimmune disease and neurodegenerative disease. In some embodiments, the brain disease or damage is selected from brain cancer, brain metastases from non-brain cancer, brain injury or trauma, stroke, brain infection, brain toxicity (e.g., from drugs, irradiation, etc.) autoimmune / inflammatory disease and neurodegenerative disease. In some embodiments, the brain disease or damage is selected from brain cancer, brain injury, stroke, brain infection and neurodegenerative disease. In some embodiments, the brain disease or damage isselected from multiple sclerosis (MS), brain damage by irradiation, brain cancer and systemic lupus erythematosus. In some embodiments, brain injury or trauma is a blast injury. In some embodiments, brain injury or damage is brain damage by irradiation. In some embodiments, the brain disease or damage is a neurodegenerative disease. In some embodiments, the neurodegenerative disease is selected from amyotrophic lateral sclerosis (ALS), Alzheimer’s disease, Parkinson’s disease, dementia, multiple sclerosis (MS), X-linked adrenoleukodystrophy (X-ALD) and Huntington’s disease. In some embodiments, the neurodegenerative disease is selected from ALS and Alzheimer’s disease. In some embodiments, the neurodegenerative disease is ALS. In some embodiments, the neurodegenerative disease is Alzheimer’s disease. In some embodiments, the brain disease or damage is not a psychiatric disease. In some embodiments, the brain disease or damage is not schizophrenia. In some embodiments, the autoimmune / inflammatory brain disease is multiple sclerosis (MS). In some embodiments, the neurodegenerative disease is MS. In some embodiments, the brain disease or damage is systemic lupus erythematosus. In some embodiments, the autoimmune / inflammatory disease is systemic lupus erythematosus. In some embodiments, the brain disease or damage is brain toxicity. In some embodiments, the brain toxicity is following immunotherapy. In some embodiments, the immunotherapy is for cancer. In some embodiments, the cancer is no brain cancer. In some embodiments, the cancer is brain cancer. In some embodiments, the brain disease or damage is brain cancer. In some embodiments, the brain cancer is glioblastoma.

[0097] In some embodiments, the pancreatic disease or damage is a pancreatic disease. In some embodiments, the pancreatic disease or damage is characterized by pancreatic cell death. In some embodiments, the pancreatic disease or damage is characterized by beta cell death. In some embodiments, the pancreatic disease or damage is characterized by pancreatic cfDNA in the subject. In some embodiments, the pancreatic disease or damage is characterized by beta cfDNA in the subject. In some embodiments, in the subject is in circulation. In some embodiments, the pancreatic disease or damage is selected from pancreatic cancer, pancreatitis, diabetes, congenital hyperinsulinism and damage to transplanted islets. In some embodiments, the pancreatic disease or damage is selected from pancreatic cancer, pancreatitis, and diabetes. In some embodiments, the pancreatic disease or damage is diabetes. In some embodiments, diabetes is diabetes type 1 and / or diabetes type 2. In some embodiments, diabetes is diabetes type 1. In some embodiments, diabetes is diabetes type 2. In some embodiments, transplanted islets comprise transplanted stem cell-derived beta cells. In some embodiments, transplanted islets comprise transplanted stem cell-derived islets. In some embodiments, the pancreatic disease or damage is pancreatitis. In some embodiments, the pancreatic disease or damage is pancreatic cancer. In some embodiments, pancreatic cancer is selected from ductal adenocarcinoma and islet cell tumors.

[0098] In some embodiments, the methylation data is produced by converting unmethylated cytosines in the DNA to uracils. In some embodiments, the ascertaining is affected by converting unmethylated cytosines in the DNA to uracils. In some embodiments, the methylation data is produced by contacting the DNA with an agent that converts unmethylated cytosines to uracils to produce converted DNA. In some embodiments, the ascertaining is affected by contacting the DNA with an agent that converts unmethylated cytosines to uracils to produce converted DNA. In some embodiments, the agent is bisulfite. In some embodiments, the agent is an enzyme. In some embodiments, the enzyme is a deaminating enzyme. In some embodiments, the enzyme is an Apolipoprotein B mRNA editing enzyme, catalytic polypeptide (APOBEC). In some embodiments, the APOBEC is APOBEC3A. In some embodiments, methylated cytosines are protected from deamination. In some embodiments, the protection is by contact with an oxidizing enzyme. In some embodiments, the oxidizing enzyme is Ten-Eleven-Translocation-2 (TET2). In some embodiments, the protection is by contact with a glucosyltransferase enzyme. In some embodiments, the glucosyltransferase enzyme is T4-P-Glucosyltransferase (T4-BGT). In some embodiments, the agent is a plurality of enzymes, wherein the plurality comprises a deaminase, an oxidase and glycosyltransferase. In some embodiments, the plurality of enzymes comprises APOBEC3 A, TET2 and T4-BGT. In some embodiments, the converting is by deamination. Methods of acquiring methylation data include for example Methyl-Seq, Bisulfite-Seq (BS-Seq), methyC-capture sequencing (MCC-Seq), reduced-representation bisulfite sequencing (RRBS-Seq), and enzymatic methylation-seq (EM-Seq).

[0099] In some embodiments, the production further comprises sequencing the converted DNA. In some embodiments, the sequencing is deep sequencing, next generation sequencing or massively parallel sequencing. In some embodiments, the sequencing is whole genome sequencing (WGS). In some embodiments, the sequencing is shallow sequencing. As used herein the term “deep sequencing” refers to sequencing with at least a depth of 50x, but which generally has a depth of greater than lOOx. “Shallow sequencing” is also known as low-pass whole genome sequencing, as used herein the term refers to sequencing with a depth of at most 5x. In some embodiments, shallow sequencing comprises a read depth of O.lx to 5x. In some embodiments, shallow sequencing comprises a read depth of O.Olx to5x. In some embodiments, shallow sequencing comprises a read depth of 0.05x to 5x. In some embodiments, shallow sequencing comprises a read depth of O.lx to lx. In some embodiments, shallow sequencing comprises a read depth of O.Olx to lx. In some embodiments, shallow sequencing comprises a read depth of 0.05x to lx.

[0100] In some embodiments, the sequencing is at a read depth or comprises a read depth of at least 0.001, 0.01, 0.1, 1, 2, 3, 4, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 100, 500 thousand or 1 million reads. Each possibility represents a separate embodiment of the invention. In some embodiments, the sequencing is at a read depth or comprises a read depth of at least 20K reads. In some embodiments, the sequencing is at a read depth or comprises a read depth of at least 1 million reads. In some embodiments, the sequencing is at a read depth or comprises a read depth of at most 1, 2, 3, 4, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 100, 500 thousand, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 30, 40 or 50 million reads. Each possibility represents a separate embodiment of the invention. In some embodiments, the sequencing is at a read depth or comprises a read depth of at most 20K reads. In some embodiments, the sequencing is at a read depth or comprises a read depth of at most 1 million reads. In some embodiments, the sequencing is at a read depth or comprises a read depth of at most 10 million reads.

[0101] In some embodiments, the sequencing is at a genome coverage or comprises a genome coverage of at least 0.0001, 0.001, 0.01, 0.1, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, or 50x coverage (50x coverage meaning each locus in the genome is sequenced 50 times on average). Each possibility represents a separate embodiment of the invention. In some embodiments, the sequencing is at a genome coverage or comprises a genome coverage of at least 0.05x. In some embodiments, the sequencing is at a genome coverage or comprises a genome coverage of at least O.Olx. In some embodiments, the sequencing is at a genome coverage or comprises a genome coverage of at most 0.01, 0.1, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50, 100, 150, or 200x coverage. Each possibility represents a separate embodiment of the invention. In some embodiments, the sequencing is at a genome coverage or comprises a genome coverage of at most 0.05x. In some embodiments, the sequencing is at a genome coverage or comprises a genome coverage of at most O.Olx. In some embodiments, the sequencing is at a genome coverage or comprises a genome coverage of at most lx. In some embodiments, the sequencing is at a genome coverage or comprises a genome coverage of at most 5x.

[0102] In some embodiments, the method has a limit of detection (LOD) of at least 1 :43,000, 1:40,000, 1:30,000, 1:20,000, 1:15,000, 1:10,000 or 1:5,000. Each possibility represents a separate embodiment of the invention. In some embodiments, the method has a limit ofdetection of at least 1:43,000. As used herein, the term “limit of detection” refers to the number of neuronal reads within all reads that can be detected. One of the unexpected advantages of the instant method is that it is ultra-sensitive with a much lower limit of detection as compared to other methods of neuronal cfDNA detection. Indeed, it is exemplified hereinbelow that 1 neuronal cfDNA molecule within 43,478 total cfDNA molecules can be detected.

[0103] In some embodiments, the methylation data is produced without converting unmethylated cytosines to uracil. In some embodiments, the ascertaining is affected without converting unmethylated cytosines to uracil. In some embodiments, the methylation data is produced by sequencing the DNA using single-molecule real-time (SMRT) sequencing. In some embodiments, the ascertaining is affected sequencing the DNA using SMRT sequencing. In some embodiments, SMRT sequencing detects both the sequence of the DNA and the methylation status of each cytosine. In some embodiments, the methylation data is produced by passing the DNA through a nanopore. In some embodiments, the nanopore is a nanopore sequencer. In some embodiments, the nanopore detects both the sequence of the DNA and the methylation status of each cytosine. Examples of sequencing that may be used include, but are not limited to Oxford Nanopore sequencing, PacBio sequencing,

[0104] In some embodiments, the methylation data is analyzed with a trained machine learning model. In some embodiments, analyzing comprises applying a machine learning algorithm. In some embodiments, analyzing comprises applying a machine learning model. In some embodiments, the machine learning model is a machine learning algorithm. In some embodiments, the machine learning model implements a machine learning algorithm. In some embodiments, the algorithm is a classifier. In some embodiments, the algorithm is a regression model. In some embodiments, the algorithm is supervised. In some embodiments, the algorithm is unsupervised. In some embodiments, the model is a neural network. In some embodiments, diagnosing is diagnosing by the machine learning algorithm. In some embodiments, the machine learning algorithm is the trained machine learning algorithm. In some embodiments, the machine learning algorithm is the machine learning algorithm during training.

[0105] In some embodiments, at an inference stage the trained machine learning algorithm is applied. In some embodiments, the trained machine learning algorithm is applied to a subjects methylation data. In some embodiments, at the inference stage an input is received. In some embodiments, the input comprises the subject’s methylation data. In some embodiments, at the inference stage the trained machine learning algorithm is applied. Insome embodiments, applied is applied to the input. In some embodiments, the input is the received input. In some embodiments, the inference stage is to diagnose disease or damage.

[0106] A neural network, e.g. a neural network implementing machine learning, may refer to an information processing paradigm that may include nodes, referred to as neurons, organized into layers, with links between the neurons. The links may transfer signals between neurons and may be associated with weights. A NN may be configured or trained for a specific task, e.g., pattern recognition or classification. Training a NN for the specific task may involve adjusting these weights based on examples. Each neuron of an intermediate or last layer may receive an input signal, e.g., a weighted sum of output signals from other neurons, and may process the input signal using a linear or nonlinear function (e.g., an activation function). The results of the input and intermediate layers may be transferred to other neurons and the results of the output layer may be provided as the output of the NN. Typically, the neurons and links within a NN are represented by mathematical constructs, such as activation functions and matrices of data elements and weights. A processor, e.g., CPUs or graphics processing units (GPUs), or a dedicated hardware device may perform the relevant calculations.

[0107] In some embodiments, the machine learning model is trained on methylation data from DNA from subjects suffering from the brain disease or damage. In some embodiments, the machine learning model is trained on methylation data from cfDNA from samples from subjects suffering from the brain disease or damage. In some embodiments, the machine learning model is trained on methylation data from DNA from subjects suffering from the pancreatic disease or damage. In some embodiments, the machine learning model is trained on methylation data from cfDNA from samples from subjects suffering from the pancreatic disease or damage. In some embodiments, the machine learning model is trained on methylation data from DNA from subjects suffering from the brain disease or damage and / or pancreatic disease or damage. In some embodiments, the machine learning model is trained on methylation data from cfDNA from samples from subjects suffering from the brain disease or damage and / or pancreatic disease or damage.

[0108] In some embodiments, the machine learning model is trained on methylation data from DNA from subjects not suffering from the brain disease or damage. In some embodiments, the machine learning model is trained on methylation data from cfDNA from samples from subjects not suffering from the brain disease or damage. In some embodiments, the machine learning model is trained on methylation data from DNA from subjects not suffering from the pancreatic disease or damage. In some embodiments, the machinelearning model is trained on methylation data from cfDNA from samples from subjects not suffering from the pancreatic disease or damage. In some embodiments, the machine learning model is trained on methylation data from DNA from subjects not suffering from the brain disease or damage and / or pancreatic disease or damage. In some embodiments, the machine learning model is trained on methylation data from cfDNA from samples from subjects not suffering from the brain disease or damage and / or pancreatic disease or damage. In some embodiments, subjects not suffering from the disease or damage are control subjects.

[0109] In some embodiments, the machine learning model is trained on a training set. In some embodiments, the training set comprises methylation data from subject with the disease / damage and subjects without the disease / damage and labels indicating from which subject the methylation data originated. In some embodiments, the training set comprises methylation data and labels indicating the subject from which the data was derived. In some embodiments, the training set comprises methylation data from subjects suffering from a neuronal disease or damage. In some embodiments, the training set comprises methylation data from subjects suffering from a pancreatic disease or damage. In some embodiments, the training set comprises methylation data from subjects suffering from a neuronal disease or damage and / or a pancreatic disease or damage.

[0110] In some embodiments, the trained machine learning model outputs a diagnosis. In some embodiments, the diagnosis is a disease diagnosis. In some embodiments, the diagnosis is the disease or damage is present or the disease or damage is absent. In some embodiments, the trained machine learning model outputs a disease score. In some embodiments, a disease score is a disease or damage score. In some embodiments, a disease score above a predetermined threshold indicates the subject suffers from the disease or damage. In some embodiments, the predetermined threshold is a the predetermined disease score threshold. In some embodiments, the threshold is a the score produced for control subject. In some embodiments, the score produced is a the average score produced. In some embodiments, control subjects are subjects with the disease or damage.

[0111] In some embodiments, the score is a linearized score. In some embodiments, the score is from 0 to 1. In some embodiments, the threshold is selected from 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, and 0.9. Each possibility represents a separate embodiment of the invention. In some embodiments, the score is from 0 to 10. In some embodiments, the threshold is selected from 1, 2, 0.3, 4, 5, 6, 7, 8, and 9. Each possibility represents a separate embodiment of the invention.

[0112] In some embodiments, the method further comprises administering to a subject determined or diagnosed with a brain disease or damage a brain therapeutic agent. In some embodiments, a brain therapeutic agent is a treatment. In some embodiments, the method further comprises administering to a subject determined or diagnosed with a pancreatic disease or damage a pancreatic therapeutic agent. In some embodiments, a pancreatic therapeutic agent is a treatment. In some embodiments, the treatment is insulin. In some embodiments, the disease is diabetes and the treatment is insulin. In some embodiments, the treatment is an immunosuppressant. In some embodiments, the disease or damage is damage to transplanted islets / beta cells and the treatment is an immunosuppressant. Immunosuppressants to deal with transplant rejection / damage are well known in the art, these include but are not limited to anti-CD3 antibodies (e.g., teplizumab), steroids (e.g., prednisone), tacrolimus, cyclosporine, mycophenolate mofetil, azathioprine and Rapamycin (e.g., sirolimus) to name but a few.

[0113] As used herein, the term "about" when combined with a value refers to plus and minus 10% of the reference value. For example, a length of about 1000 nanometers (nm) refers to a length of 1000 nm+- 100 nm.

[0114] It is noted that as used herein and in the appended claims, the singular forms "a," "an," and "the" include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to "a polynucleotide" includes a plurality of such polynucleotides and reference to "the polypeptide" includes reference to one or more polypeptides and equivalents thereof known to those skilled in the art, and so forth. It is further noted that the claims may be drafted to exclude any optional element. As such, this statement is intended to serve as antecedent basis for use of such exclusive terminology as "solely," "only" and the like in connection with the recitation of claim elements, or use of a "negative" limitation.

[0115] In those instances where a convention analogous to "at least one of A, B, and C, etc." is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., "a system having at least one of A, B, and C" would include but not be limited to systems that have A alone, B alone, C alone, A and B together, A and C together, B and C together, and / or A, B, and C together, etc.). It will be further understood by those within the art that virtually any disjunctive word and / or phrase presenting two or more alternative terms, whether in the description, claims, or drawings, should be understood to contemplate the possibilities of including one of the terms, either of the terms, or both terms. For example, the phrase "A or B" will be understood to include the possibilities of "A" or "B" or "A and B."

[0116] It is appreciated that certain features of the invention, which are, for clarity, described in the context of separate embodiments, may also be provided in combination in a single embodiment. Conversely, various features of the invention, which are, for brevity, described in the context of a single embodiment, may also be provided separately or in any suitable sub-combination. All combinations of the embodiments pertaining to the invention are specifically embraced by the present invention and are disclosed herein just as if each and every combination was individually and explicitly disclosed. In addition, all subcombinations of the various embodiments and elements thereof are also specifically embraced by the present invention and are disclosed herein just as if each and every such sub-combination was individually and explicitly disclosed herein.

[0117] Additional objects, advantages, and novel features of the present invention will become apparent to one ordinarily skilled in the art upon examination of the following examples, which are not intended to be limiting. Additionally, each of the various embodiments and aspects of the present invention as delineated hereinabove and as claimed in the claims section below finds experimental support in the following examples.

[0118] Various embodiments and aspects of the present invention as delineated hereinabove and as claimed in the claims section below find experimental support in the following examples.EXAMPLES

[0119] Generally, the nomenclature used herein and the laboratory procedures utilized in the present invention include molecular, biochemical, microbiological and recombinant DNA techniques. Such techniques are thoroughly explained in the literature. See, for example, "Molecular Cloning: A laboratory Manual" Sambrook et al., (1989); "Current Protocols in Molecular Biology" Volumes I-III Ausubel, R. M., ed. (1994); Ausubel et al., "Current Protocols in Molecular Biology", John Wiley and Sons, Baltimore, Maryland (1989); Perbal, "A Practical Guide to Molecular Cloning", John Wiley & Sons, New York (1988); Watson et al., "Recombinant DNA", Scientific American Books, New York; Birren et al. (eds) "Genome Analysis: A Laboratory Manual Series", Vols. 1-4, Cold Spring Harbor Laboratory Press, New York (1998); methodologies as set forth in U.S. Pat. Nos. 4,666,828; 4,683,202; 4,801,531; 5,192,659 and 5,272,057; "Cell Biology: A Laboratory Handbook", Volumes I-III Cellis, J. E., ed. (1994); "Culture of Animal Cells - A Manual of Basic Technique" by Freshney, Wiley-Liss, N. Y. (1994), Third Edition; "Current Protocols in Immunology"Volumes I-III Coligan J. E., ed. (1994); Stites et al. (eds), "Basic and Clinical Immunology" (8th Edition), Appleton & Lange, Norwalk, CT (1994); Mishell and Shiigi (eds), "Strategies for Protein Purification and Characterization - A Laboratory Course Manual" CSHL Press (1996); all of which are incorporated by reference. Other general references are provided throughout this document.Example 1: Global abundance of non-CpG methylation in humans

[0120] The inventors first asked whether the global abundance of CpH methylation in neurons is sufficient to support ultrasensitive detection of neuronal DNA. To this end, genome-wide non-CpG methylation levels were calculated in multiple cell types, with a focus on neurons and on the major contributors to cfDNA (hematopoietic cells, hepatocytes, and vascular endothelial cells), using a combination of public and in-house methylome datasets. In agreement with previous analyses, it was found that bulk brain tissue exhibits substantial non-CpG methylation, with 3.4% of CpA sites methylated, as well as 1.3% of CpT sites and 0.75% of CpC (Fig. 1A). Breaking down to constituent brain cell types, it was found that in sorted cortical neurons, 11% of CpA sites were methylated, as well as 3.5% of CpT sites and 0.96% of CpC sites. In all other tissues, 0.5-1% of non-CpG sites were methylated, in agreement with previous reports.

[0121] The observation that 11% of CpA sites are methylated in neurons suggests an unusually dense epigenetic signal and potentially an extremely sensitive biomarker. In fact, most of the ~18 million nucleosome-size cfDNA fragments released from each dying neuron (166bp-long, containing a median of 11 CpA sites; see Methods) are expected to carry an identifiable covalent mark. However, the presence of 1% CpA methylation in non -neurons would provide a signal -to-noise ratio too low for utility as a biomarker for presence of rare neuronal DNA in complex mixtures.Example 2: Apparent non-neuronal CpH methylation largely reflects technical artifacts

[0122] What are the origins of the apparent background CpH methylation in non-neuronal tissues? Most methods for detection of cytosine methylation, either chemical (bisulfite) or enzymatic (protective oxidation of methylated cytosines by TET followed by Apobec-mediated deamination) rely on selective deamination of unmethylated cytosines to uracils, while methylated cytosines are protected. Following PCR amplification and sequencing, unmethylated cytosines are read as thymidines and methylated cytosines are read as cytosines. The process of deamination is not perfect in either method, and a certainproportion of unmethylated cytosines - typically close to 1% - fails to convert. It is therefore hypothesized that much of the CpH methylation observed in non-neuronal tissues reflects failed deamination rather than bona fide non-CpG methylation. To test this hypothesis, unmethylated phage Lambda DNA, which is often used as an internal spike-in control, was analyzed. Consistent with a technical origin, the extent of CpH sites read as methylated in Lambda DNA correlated strongly with the apparent CpH methylation levels measured in non-neuronal human genomic DNA (Fig. IB). This correlation was observed across diverse tissues, cell types, and cfDNA samples, and across all non-CpG contexts (CpA, CpT and CpC), indicating that the background signal scales with sample-specific conversion efficiency. Brain-derived samples were excluded from this analysis, as their high endogenous CpH methylation obscures the low-level technical background.

[0123] This relationship enabled direct estimation of true CpH methylation in each sample, by subtracting the sample-specific artifactual Lambda-derived CpH signal from the corresponding genomic CpH measurements. Following this correction, CpH methylation in non-neuronal tissues was reduced to -0.1% (Fig. 1C). In neurons, however, CpH methylation remained high, with corrected levels of 11% for CpA, 3.5% for CpT, and 0.96% for CpC. Thus, accounting for technical under-conversion reveals an effective signal-to-noise ratio of 100:1 between CpA methylation in neurons and non-neuronal tissues.Example 3: Multiple methylated CpAs distinguish neuronal DNA fragments

[0124] Having established that true CpH methylation is abundant in neurons and negligible in other tissues, the inventors next sought to develop a molecular-level classifier that could reliably distinguish neuron-derived DNA fragments from the vast background of blood cfDNA. To this end, the distinct topologies of biological methylation and technical noise that arises from the chemistry of experimental deamination were considered. CpG methylation typically exhibits a regional organization, with adjacent sites showing concordant patterns due to processive enzymatic activity. It was hypothesized that authentic CpH methylation in neurons would exhibit similar local clustering, whereas technical artifacts such as the failed bisulfite conversion of unmethylated cytosines would occur as stochastic, independent events (Fig. 2A). Selected genomic loci harboring methylated CpA sites (mCpA) in neurons showed a pattern of methylation in neurons and blood that was consistent with this idea (Fig. 2B).

[0125] To systematically test the clustering hypothesis, experimental methylomes from neurons and blood were compared to a generative null model that simulates CpA methylationas a random, independent process following a binomial distribution, parameterized by the observed global methylation rates (11% for neurons and 0.53% for blood). DNA fragments were scored by the number of mCpA sites they carried.

[0126] As expected under the null model, fragments with many mCpAs were rare in both neurons and blood. However, the experimental neuronal data deviated strongly from the expected, consistent with biological clustering. As the stringency threshold for methylated CpA increased (that is, requiring more methylated CpAs per fragment to classify a molecule as neuron-derived), the blood background decayed rapidly, closely matching the null prediction. In contrast, the neuronal signal decayed much more slowly, indicating an excess of fragments carrying multiple mCpAs beyond what would be expected by chance (Fig.2C).

[0127] At a threshold of methylated CpA>6 per read, the null model predicted only 2,944 reads per million (RPM) in neuronal DNA, whereas the experimental data yielded 17,943 RPM, representing a six-fold enrichment driven by biological clustering of CpA methylation (Fig. 2C). This divergence between model and experimental data underlies the discriminatory power of the classifier.

[0128] While the simulated signal-to-noise ratio (assuming no clustering of methylated CpA) peaked early and declined (maxing at -23,000:1), the experimental neuron-to-blood ratio continued to increase with stringency, reaching a 16,991:1 ratio at mCpA>6 (neuronal signal: 17,943 reads per million; Blood background: 1.1 RPM; Fig.2C-2D). Stricter criteria reduced the neuronal read count without further increasing discrimination. The inventors thus selected the presence of at least 6 methylated CpA sites as an operational criterion for neuronal DNA.

[0129] Though six methylated CpA sites per molecule was found to be the optimal threshold and seven methylated CpA sites was usable as a strict threshold, it is notable that 5 methylated CpA and even 4 methylated CpA were capable of producing a statistically significant signal and thus can still be used to identify neuronal DNA.

[0130] Incomplete denaturation of DNA represents an additional, qualitatively distinct source of technical noise. Because bisulfite deamination requires single-stranded DNA, failed denaturation protects all cytosines from conversion, producing fragments that appear fully methylated (Fig. 2A). To exclude such artifacts, the inventors exploited the scarcity of authentic CpC methylation in neurons. Fragments were classified as neuron-derived only if they contained at least two unmethylated CpC sites and no methylated CpCs, providing an internal control for successful denaturation and bisulfite activity. Applying this filter inconjunction with the >6 mCpAs criterion yielded 17,943 RPM in neuronal DNA and ~1 RPM in blood, corresponding to a final signal-to-noise ratio of 16,991:1 (Fig. 2D). When pseudo count regularization was omitted, the corresponding ratio exceeded 33,000:1.

[0131] Using a definition of at least 6 methylated CpA per read and at least two CC dinucleotides that are unmethylated (all CpCs must be unmethylated but at least two such dinucleotides must be present), CpA methylation was examined in various cell types. (Fig.3). Blood cells, megakaryocytes, endothelial cells, erythroblasts, hepatocytes, and epithelial cells all showed essentially no reads (<1 RPM) with 6 CpA methylation with two unmethylated CpCs, i.e., they showed no genuine CpA methylation. In contrast, neurons showed very high (17,943 RPM) genuine CpA methylated molecules. Surprisingly, pancreatic islet cells (i.e., endocrine pancreas), in particular beta cells (-280 RPM), but also alpha and delta cells, showed low level genuine CpA methylation. This CpA methylation was not observed in exocrine pancreas cells (Panc-exo: acinar and ductal cells) which also had less than 1 RPM. This potentially reflects the known molecular resemblance of islet cells to neurons. As this methylation in the endocrine pancreas was so much lower than neuronal methylation the two cell types can still be distinguished.

[0132] Together, these analyses establish a robust molecule-level classifier to distinguish neuronal from non-neuronal DNA molecules with high sensitivity and accuracy. It also enabled differentiation of endocrine pancreatic DNA molecules from non-endocrine pancreatic molecules.Example 4: Ultrasensitive detection of neuronal DNA in plasma

[0133] Having established a robust fragment-level classifier for neuronal DNA, the inventors next evaluated its sensitivity in plasma-like mixtures. n silico serial dilution experiments were performed in which neuronal methylomes were computationally spiked into simulated cfDNA plasma methylomes composed of blood, hepatocytes and vascular endothelial DNA in proportions typical of healthy plasma. Neuronal signal was estimated using one billion reads (corresponding to whole-methylome sequencing at 50x depth). Each simulation was repeated 20 times, by randomly subsampling the respective methylomes.

[0134] The classifier readily detected the presence of neuronal DNA in dilutions down to 1 : 100,000 and failed to detect neuronal DNA when comprising 1 : 1,000,000 of the mixture, relative to simulated healthy plasma (Fig. 4A). This analytical sensitivity is in line with the theoretical performance of the classifier (Fig. 2A-2D). For comparison, the same simulatedmixtures were analyzed using neuron-specific CpG methylation markers (see Materials and methods and Loyfer et al., 2023, “A DNA methylation atlas of normal human cell types”, Nature, 613, 355-364, the contents of which are hereby incorporated by reference in their entirety). This CpG-based analysis picked up neuronal DNA only when comprising at least 1 : 1,000 of the mixture, suggesting that CpH-based analysis is two orders of magnitude more sensitive than current state-of-the-art CpG-based analyses (Fig. 4B).

[0135] The inventors then applied the assay to actual plasma samples sequenced at depths ranging from lOx to 50x, converting the proportion of mCpA reads to estimates of neuronal DNA contribution based on the mCpA profile in isolated neurons. Among 53 samples from healthy adults, the neuronal cfDNA signal comprised 0.0023% of the total (1:14,821, or 23 parts per million, Fig. 4C), reflecting a combination of potential true neuronal DNA and a baseline noise. A similar low-level signal was observed in 86 patients with pathologies not involving the brain. The highest outlier in this group (0.06% neuronal signal in plasma) corresponded to be a patient with multiple myeloma complicated by spinal cord compression necessitating irradiation, consistent with collateral neuronal damage.

[0136] In contrast, plasma from patients with conditions involving potential brain damage showed significantly higher levels of neuronal cfDNA. These included relapsing multiple sclerosis (n=2, 0.018%), brain irradiation (n=4, 0.044%), glioblastoma (n=5, 0.02%), and systemic lupus erythematosus (n=4, 0.066%), suggesting that neuronal cell death in these pathologies does release detectable cfDNA to the bloodstream. Most strikingly, plasma samples from brain-dead donors (n=16) exhibited markedly elevated concentrations of neuronal cfDNA (-0.78%), consistent with massive ongoing brain cell death (Fig. 4C).

[0137] For comparison, CpG-based analysis detected neuronal signal in only a small subset of these samples and showed a substantially broader signal distribution among patients without brain pathologies (Fig. 4D). This reduced specificity likely reflects confounding contributions of globally hypomethylated erythroblasts to cfDNA, which limit the detection of tissue-specific contributions based on demethylated CpGs.

[0138] The inventors took advantage of the high proportion of neuronal DNA detected in the plasma of brain-dead donors and performed a more rigorous plasma to plasma spike-in wet experiment. Genome-wide analysis of mCpA revealed a distinct neuronal signal even when brain-dead plasma (from a donor with an estimated neuronal cfDNA contribution of 0.57%) was diluted 1:250 into normal plasma. Thus, the fraction of neuronal cfDNA that could be detected was 0.0023%, or 23 parts per million (Fig. 4E). This suggests a real-lifelimit of detection (LOD) of 1:43,478 for detecting neuronal cfDNA in plasma, in line with theoretical considerations.Example 5: Neuronal DNA detection using shallow sequencing

[0139] A hallmark feature of CpA methylation is its broad distribution across the genome. While this may limit the utility of targeted analysis based on deep sequencing of a few selected loci, shallow whole-methylome sequencing may suffice to detect a signal, with important implications for deployment of a practical blood test. To test this idea, the inventors assessed how sensitivity of detection was affected by sequencing depth. The simulated neuronal spike-in data shown in Figure 4A was reanalyzed by randomly selecting a lower number of sequenced reads that corresponded to lower coverage. It was asked how reduced sequencing depth affected the ability to detect a neuronal signal. Strikingly, the analysis indicated that sequencing depth as low as 0.05x (i.e. one million reads) was sufficient to pick up a neuronal signal comprising 0.01% of the total (Fig. 4F). The performance of shallow sequencing in this experiment indicated that one can detect neuronal cfDNA using minute amounts of blood, and near-real time sequencing methods in clinical settings that require a short turnaround time can be done.Example 6: Diagnostic identification of neuronal DNA in blood

[0140] A published cfDNA dataset from patients with ALS (Caggiano et al., “Comprehensive cell type decomposition of circulating cell-free DNA with CelFiE”, Nat. Comm., 2021 May 11; 12(1):2717, the contents of which are hereby incorporated by reference in their entirety), which failed to detect neuronal DNA using CpG deconvolution was examined. Notably, only 75bp fragments were sequenced, greatly reducing the number of informative fragments. ALS patients showed an elevation in neuronal cfDNA (Fig. 5).

[0141] It is not currently known if ALS patients have significantly elevated blood levels of neuronal cfDNA, thus an in-silico experiment was designed to test sensitivity of the assay. Reads from seven neuron samples and 10 normal cfDNA blood samples were mixed to produce 10 simulated samples of cfDNA with low concentrations (10%, 1%, 0.1%, 0.01% and 0%) of neuronal DNA fragments, with various depths of sequencing (200M, 20M, 2M, 200K, and 20K reads). Each mixture was deconvoluted based on the number of methylated CpA reads (cutoff of six methylated CpAs per molecule) or CpG reads. CpG analysis was performed as described in Loyfer et al., “A DNA methylation atlas of normal human celltypes”, Nature, 613, 355-364, the contents of which are hereby incorporated by reference in its entirety.

[0142] The CpG method was able to identify the neuronal spike-in at higher neuronal fractions. Additionally deeper sequencing depth was required (Fig.6A). In contrast, the CpA method was able to detect neuronal DNA even when it was only 0.1% of the total. Further, even at the lowest sequencing depth (20K reads, equivalent to O.Olx coverage) the CpA method could detect the neuronal cfDNA (Fig. 6B).Example 7: Use of a machine learning (ML) model to diagnose disease

[0143] The overall CpA methylation in cfDNA has thus been shown to be greatly elevated as compared to controls. A cutoff can thus easily be assigned that distinguishes between the control and the patients with disease (e.g., above 12 ppm, above 20 ppm, etc.). It is also tested if a machine learning algorithm can be trained to distinguish between samples from a disease subject and those from a healthy subject. A classifier algorithm is trained on methylation data from subjects with and without neurological disease (e.g., ALS, Alzheimer’s, etc.). The data contains the sequences of the cfDNA as well as the methylation status of various CpX dinucleotide. A training set corresponding to -75% of the samples is used for determining the ideal training set. Various different methylation input data (in addition to the sequence itself) is tested, including total methylation (i.e., all CpX dinucleotides), methylation of just CpA, methylation of CpA and CpC, methylation of CpA and CpT, methylation of CpA, CpC, and CpT. The trained models produced are then tested on the test cohort (the remaining -25% of samples). The model that best differentiates disease samples from healthy samples is selected. Leave one out cross-validation is also performed to ensure the best model is selected.

[0144] Although the invention has been described in conjunction with specific embodiments thereof, it is evident that many alternatives, modifications and variations will be apparent to those skilled in the art. Accordingly, it is intended to embrace all such alternatives, modifications and variations that fall within the spirit and broad scope of the appended claims.

Claims

CLAIMS:

1. A method of diagnosing a brain disease or brain damage in a subject in need thereof, the method comprising:a. receiving methylation data comprising CpA methylation data, CpT methylation data or both from cell-free DNA (cfDNA) from said subject; and b. i. quantifying the total number of cfDNA molecules comprising at least one methylated CpA and / or CpT dinucleotide, wherein a number of molecules above a predetermined cfDNA threshold indicates the subject suffers from the disease or damage; orii. analyzing said data with a trained machine learning model, wherein said machine learning model is trained on said methylation data from cfDNA from subjects suffering from said brain disease or brain damage and samples from subjects without said brain disease or brain damage and wherein said trained machine learning model outputs a disease diagnosis or a disease score, wherein a disease score above a predetermined disease score threshold indicates the subject suffers from the disease or damage;thereby diagnosing a brain disease or damage.

2. The method of claim 1, wherein said methylation data further comprises CpC methylation data and step (bii) comprises quantifying the total number of cfDNA molecules comprising at least one methylated CpA or CpT and devoid of a methylated CpC.

3. The method of claim 2, wherein step (bii) comprises quantifying the total number of cfDNA molecules comprising at least one methylated CpA, containing at least one CpC dinucleotide and devoid of a methylated CpC.

4. The method of any one of claims 1 to 3, wherein said received methylation data comprises CpA methylation data and said quantifying is quantifying the total number of cfDNA molecules comprising at least one methylated CpA.

5. The method of any one of claims 1 to 4, wherein at least 1 methylated CpA is a plurality of methylated CpAs.

6. The method of claim 5, wherein said plurality of methylated CpAs is at least 6 methylated CpAs.

7. The method of any one of claims 1 to 6, wherein said methylation data is whole genome methylation data.

8. A method of detecting neuronal DNA in cell free DNA (cfDNA), the method comprising ascertaining the methylation status of at least two cytosines in a CpA and / or CpT dinucleotide and the methylation status of at least two cytosines being the first cytosine of a CpC dinucleotide on a continuous sequence of the same cfDNA molecule from said sample, and wherein methylation of said at least two cytosines in a CpA and / or CpT dinucleotide and unmethylation of said at least two cytosines of a CpC dinucleotide indicates the DNA is a neuronal DNA, thereby detecting neuronal DNA in a sample.

9. The method of any one of claims 1 to 8, wherein said cfDNA is from a fluid sample, optionally wherein said fluid is selected from the group consisting of blood, plasma, sperm, milk, urine, saliva and cerebral spinal fluid.

10. The method of claim 9, wherein said sample is a blood sample.

11. The method of any one of claims 1 to 10, wherein said method is a method of detecting neuronal DNA, brain disease or brain damage.

12. The method of claim 11, wherein brain disease is a disease characterized by death of neuronal cells.

13. The method of any one of claims 8 to 12, where said continuous sequence comprises no more than 167 nucleotides.

14. The method of any one of claims 8 to 13, wherein said at least two cytosines in a CpA or CpT dinucleotide is at least two cytosines in a CpA dinucleotide.

15. The method of any one of claims 8 to 14, wherein said at least two cytosines in a CpA or CpT dinucleotide is at least four cytosines in a CpA or CpT dinucleotide.

16. The method of claim 15, wherein at least four cytosines is at least six cytosines.

17. The method of any one of claims 8 to 16, wherein said at least two cytosines being the first cytosine of a CpC dinucleotide is all cytosines being the first cytosine of a CpC dinucleotide.

18. The method of any one of claims 1 to 17, wherein said methylation data is produced by or said ascertaining is affected by contacting the cfDNA with an agent that converts unmethylated cytosines to uracils to produce converted cfDNA and sequencing the converted cfDNA.

19. The method of claim 18, wherein said agent is bisulfite or a deaminase enzyme.

20. The method of claim 18 or 19, wherein said methylation data is produced by bisulfite methylation-sequencing (BS-Seq) or by enzymatic methylation sequencing (EM-Seq).

21. The method of any one of claims 18 to 20, wherein said sequencing is deep sequencing, next generation sequencing or massively parallel sequencing.

22. The method of any one of claims 18 to 20, wherein said sequencing is shallow sequencing.

23. The method of any one of claims 1 to 22, comprising a limit of detection of at least 1 neuronal cfDNA molecule in 40,000 cfDNA molecules.

24. The method of any one of claims 1 to 23, wherein said methylation data is produced or said ascertaining is affected without converting unmethylated cytosines to uracil and sequencing said cfDNA using single-molecule real-time (SMRT) sequencing detecting both the sequence of said DNA and the methylation status of each cytosine.

25. The method of any one of claims 18 to 24, wherein said sequencing is at a depth of at least 20K reads.

26. The method of any one of claims 1 to 17, wherein said methylation data is produced by or said ascertaining is affected without converting unmethylated cytosines to uracil and by passing the DNA through a nanopore and detecting both the sequence of said DNA and the methylation status of each cytosine.

27. The method of any one of claims 8 to 26, further comprising quantitating the number of neuronal or pancreatic islet cfDNA molecules in said sample, wherein a number of molecules above a predetermined cfDNA threshold indicates the sample contains neuronal or pancreatic islet DNA.

28. The method of any one of claims 1 to 7 and 27, wherein said predetermined cfDNA threshold is the total number of cfDNA molecules comprising at least one methylated CpA in cfDNA from subjects that do not suffer from the disease or damage.

29. The method of any one of claims 1 to 7 and 27 to 28, wherein said predetermined threshold is greater than 12 parts per million (ppm).

30. A method of diagnosing a brain disease or damage in a subject in need thereof, the method comprising receiving a fluid sample from said subject and detecting neuronal cell free DNA (cfDNA) in said sample by a method of any one of claims 8 to 29, wherein the presence of neuronal DNA in said sample indicates the subject suffers from a brain disease.

31. The method of any one of claims 1 to 7 and 30, wherein said brain disease or damage is selected from brain cancer, brain metastases from non-brain cancer, brain injury or trauma, stroke, brain infection, brain toxicity, autoimmune / inflammatory disease and neurodegenerative disease.

32. The method of claim 31, wherein said brain disease is a neurodegenerative disease and is selected from amyotrophic lateral sclerosis (ALS), Alzheimer’s disease, Parkinson’sdisease, dementia, multiple sclerosis (MS), X-linked adrenoleukodystrophy (X-ALD) and Huntington’s disease.

33. The method of claim 32, wherein said brain disease is Alzheimer’s disease.

34. The method of claim 31, wherein said brain disease or damage is selected from MS, brain damage by irradiation, brain cancer and systemic lupus erythematosus; optionally wherein said brain cancer is glioblastoma.

35. The method of any one of claims 1 to 7 and 29 to 34, further comprising administering a therapeutic agent to a subject determined to suffer from a brain disease or damage.

Images