Methods and devices for parental origin disease allele detection for the diagnosis and management of genetic diseases - Patents.com

Abstract

A method for assigning uniparental origin to haplotypes, subhaplotypes, or alleles associated with haplotypes, wherein genomic chromosome-length haplotypes are generated and a methylation variable status of at least one imprinted methylation variable region (iDMR) associated with each one of the autosomes is determined, and the methylation variable status of the at least one iDMR is used to assign a uniparental origin for each one of the chromosome-length haplotypes.

Description

[Technical field]

[0001] (Reference to Related Application) This application claims priority to and for U.S. purposes the benefit of U.S. Provisional Patent Application No. 63 / 340,712, filed May 11, 2022, which is incorporated herein by reference in its entirety for all purposes.

[0002] FIELD OF THEINVENTION Some embodiments relate to methods or devices for determining the uniparental origin of a haplotype, subhaplotype, or allele in a mammalian subject.Some embodiments relate to methods or devices for determining the uniparental origin of a haplotype, subhaplotype, or allele in a mammalian subject without the use of parental sequence information. [Background technology]

[0003] Phasing is traditionally defined as the task of distinguishing alleles from their maternal and paternal homologs, but in practice, most current phasing methods completely ignore parental information. Instead, chromosomes are described as a series of sub-chromosomal phase blocks, each of which is either composed of alleles grouped into two haplotypes (in the diploid case), or as a chromosome-long phase block that is not assigned to a single parent of origin (PofO). In this sense, true phase information is largely unreachable by current genomic methods that do not incorporate sequence data from at least one parent next to the offspring. 1~3 .

[0004] A notable exception to this paradigm is parental information provided by consistent differences in DNA methylation between maternally and paternally inherited alleles at imprinted differentially methylated regions (iDMRs). Such iDMRs reliably suppress the expression of either the maternal or paternal allele and, crucially, can be detected by nanopore sequencing (Oxford Nanopore Technologies), for example, using the unique ion current signature of 5-methyl-cytosine. 4~7 Long nanopore sequence reads can be used to call both sequence variation and DNA methylation to detect genome-wide allele-specific methylation. 6、7 Despite the fact that phasing using nanopore reads can achieve megabase-scale phase blocks, complete chromosome haplotypes cannot be obtained, and each chromosome is represented by several phase blocks, with likely switches between paternal and maternal origins of blocks along the chromosome. 6 .

[0005] Conversely, some phasing techniques lack parental information but produce phase blocks spanning centromeres, repeat regions, and homozygous runs. 8、9 For example, single-cell Strand-seq is a method that allows sequencing of the parental DNA template strand in a single daughter cell cultured for one round of DNA replication in the presence of BrdU. 10 Reads from the Watson template strand are mapped to the reference genome in a negative orientation, and reads from the Crick template strand are mapped in a positive orientation, meaning that alleles covered by reads with different orientations, which occur with a frequency of 50% for a given chromosome, belong to different homologs. By sequencing multiple cells, this approach allows the construction of global chromosome-length haplotypes. 8Because strand-seq phasing blocks are generally sparse (i.e., they do not phase all single nucleotide variants, SNVs), strand-seq often serves as a scaffold on which reads or sub-chromosomal phasing blocks from other sequencing technologies are combined, effectively phasing them against each other along the entire length of a given chromosome. 11、12 .

[0006] Determining the PofO for germline variants will be useful in clinical genetics. Genetic testing can inform patients' genetic risk for disease. However, predicting which side of the family an autosomal variant originates from is an important limitation of current technology. The inability to determine the parental origin (PofO) of a variant (i.e., if it is maternally or paternally inherited) may result in uncertain patient management and ineffective cascade genetic testing (CGT).

[0007] The impact of missing PofO information in clinical genetics practice can be explained within the framework of considering the patient, the family, and the health care system. In considering the impact of missing PofO information on the patient, genes with PofO effects can be considered. There are genes that require parental segregation of the pathogenic / likely pathogenic variant (PV) to interpret disease risk. This is the most direct use for determining the PofO. Germline PV in SDHD shows a PofO effect and, if inherited through the father, predisposes to a high lifetime risk for paraganglioma and pheochromocytoma 68、69Accurate prediction of maternal PV inheritance allows patients to avoid lifelong annual disease screenings that require high-resolution magnetic resonance imaging, biochemical tests and specialist evaluations, which generate direct and indirect costs for both patients and the health care system. Without knowledge of the PofO, patients may be advised to continue screening or given unclear instructions that lead to non-adherence to screening. Unnecessary and costly procedures or missed tumor diagnoses may occur. Segregation of the variant with parental CGT would not be possible if the parents are deceased, unavailable or decline genetic testing. In this respect, PofO assignment allows immediate improved cancer risk assessment for patients.

[0008] Given the impact of missing PofO information on screening recommendations, management recommendations dependent on phenotype segregation can be considered. Pancreatic ductal adenocarcinoma (PDAC) screening is recommended for BRCA1, BRCA2, MLH1, MSH2, ATM, and PALB2 PV carriers with PDAC in relatives from the same side of the family as the PV. 70 However, segregation is typically unclear. When a parent dies, PV can be separated by tissue examination, but in the case of PDAC, archival tissue is usually limited or absent. Pancreatic cancer screening is invasive (e.g., alternating annual upper endoscopy and magnetic resonance imaging (MRI)), stressful and burdensome for the patient, and costly for the system. PofO assignment predictive of non-segregation could potentially free PV carriers from PDAC screening or make it available if found to predict segregation with PV.

[0009] Given the impact of missing PofO information on variant curation, variant curation guidelines that consider phenotypic segregation within families may be considered. Upgrading or downgrading a variant from "variant of uncertain significance" (VUS) may impact cancer screening / surveillance and risk reduction interventions (e.g., surgical or medical interventions). VUS are typically found in 40% of index cases tested by multigene panels or genomic approaches. Currently, ACMG / AMP guidelines use phenotypic information and variant segregation to guide variant classification. 15 Determining the phase or whether a variant is in cis or trans with other PVs can also be used to classify variants. 15 Suspected variants of uncertain significance, where classification as pathogenic or benign cannot be achieved, may benefit from alternative evidence such as haplotyping and segregation prediction to support variant classification.

[0010] Considering the impact of missing PofO information on family members, the risk of recurrence of the pathogenic variant occurring in another pregnancy can be considered. In the setting of a de novo variant, PofO assignment can inform the risk of recurrence of the pathogenic variant occurring in another pregnancy, depending on the PofO assignment and whether the predicted uniparental origin of the variant in question is a parent of the current pregnancy / embryo / fetus or child in question. For example, if a de novo pathogenic variant is predicted to have originated from the father, there may be a risk of gonadal mosaicism for that pathogenic variant in subsequent pregnancies. However, if the subsequent pregnancy is with the same mother but a different father, there may be no significant risk for the pathogenic variant and there may be no need for amniocentesis or other investigations to evaluate the pathogenic variant. In addition, the identification of a parent who may be potentially mosaic for a pathogenic variant may also indicate an increased risk for cancer, and these individuals may benefit from increased cancer screening.

[0011] Considering the impact of missing PofO information on clarifying PV risk estimates, one can consider the a priori risks, typically around 50% each, assigned to parents carrying an autosomal dominant PV before segregation testing. Accurate prediction of PV PofO alters the a priori risk estimates for family members carrying PV. As the probability of one parent carrying PV moves from 50% to around 100%, the other parent and their extended relatives move from 50% to around 0%, typically returning them to the population risk of cancer. Second-degree relatives from the side of the family segregating PV (e.g., aunts) move from a 25% chance of carrying PV to around 50%. In certain jurisdictions, individuals with a 50% chance of carrying PV in a highly penetrant breast cancer susceptibility gene (e.g., BRCA1) have access to annual breast MRI screening until their genetic status is revealed. Given the lack of PV segregation in most second-degree relatives, the potential to alter the medical management of second-degree relatives by use of PofO prediction in the proband is highly significant.

[0012] Considering the impact of missing PofO information on cascade genetic testing (CGT), one can consider the a priori risk assigned to parents carrying autosomal dominant PVs before segregation testing, which is typically around 50% each, and current cascade genetic testing rates are low. Using PofO predictions to target CGT for entire families could have the greatest impact on the health care system. For example, estimating that of the 300,000 Canadians predicted to have hereditary cancer (HC) susceptibility, less than 10% have been identified in the past 25 years, there is a high unmet public service demand to improve CGT rates. Hereditary breast and ovarian cancer (HBOC) due to germline PVs in BRCA1 or BRCA2 is associated with a substantially elevated lifetime risk of breast, ovarian, prostate, and pancreatic cancer. Lynch syndrome (LS), resulting from germline PVs in MLH1, MSH2, MSH6, or PMS2, predisposes to colon and endometrial cancer. These are two of the most common hereditary cancer syndromes. The Centers for Disease Control and Prevention (CDC), Office of Genomics and Precision Public Health (OGPPH) recognizes HBOC and LS as having Tier 1 evidence that identification of these hereditary cancer syndromes has "significant potential for positive public health impact based on available evidence-based guidelines and recommendations" with both therapeutic and targeted cancer risk reduction utility in patients and family members. Genetic testing for these and other hereditary cancer syndromes has been available since the mid-90s. Despite the potential to prevent or catch cancer early, CGT uptake remains poor, with no relative representation of CGT in half of identified families. 20、71 Lower rates are also consistently observed among ethnic minority groups. 20This may be caused by limited intrafamilial communication, inability to meet relatives, or broken family structures. A challenge to CGT is the lack of universal screening for Lynch syndrome. 19、71 , Germline multigene panel testing for hereditary cancers 72 and tumor testing for targeted therapeutic indications 73 This includes cases where cancer susceptibility PVs are identified outside the context of a known family history of cancer, as seen in . This unmet potential necessitates the development of new strategies and techniques to increase the identification of at-risk family members. The group is developing guidelines to allow for patient-sanctioned direct contact of relatives (PSDCoR), which has been shown to be effective in increasing rates of CGT. 74 PSDCoR dramatically increases the number of patients undergoing CGT across all ethnicities and requires the commitment of additional resources to actively seek out at-risk family members. Offering CGT to parents is the first step in identifying at-risk extended relatives, but parental samples are often not available. Adoption of accurate PofO prediction can significantly reduce CGT costs, shorten the time to diagnosis, and improve risk assessment by predetermining which side of the family is at risk. The focus in terms of costs is often on the actual molecular testing until segregation is confirmed, but significant resources may be spent on pre-test genetic counseling for multiple family members on both sides of the family. Direct and indirect financial and emotional burdens are also associated with CGT of patient family members. Accurate PofO prediction could halve the number of family members requiring CGT, double the mutation detection rate, and achieve these results at a lower cost per family. For an average family size of n=20, 32Accurate PofO prediction is critical to achieve scale of PSDCoR-generated testing in an efficient and cost-effective manner. Although low-cost genetic testing of all relatives can be considered a competitor to parent-of-origin-aware genomic analysis (POAga), relying on testing of both sides of the family when segregation is not possible may actually create an undue burden on the side not at risk, increase the cost of clinical infrastructure, occupy limited genetic counseling resources, dilute the perception of risk across both sides of the family, and delay CGT in the at-risk side by decreasing the sense of urgency. Accurate PofO prediction can save costs and improve risk assessment and time to diagnosis.

[0013] Considering the impact of the lack of PofO information on the health care system, better identification of PV carriers and early detection at a more preventable or treatable stage of cancer has the potential to alleviate significant economic pressure on an overburdened health care system. Population genetic testing (PGT) may be considered as an alternative to CGT, but the need for CGT depends on how quickly and uniformly PGT can be implemented. Given the poor uptake of CGT in HC families with increased risk of cancer, PGT in average-risk individuals may prove more challenging. It is unclear how quickly adoption and saturation of PGT will occur. Of note, the two approaches are not mutually exclusive. Even with PGT, clinics still ensure that they have tested relatives within HC families. Family members may be excluded from PGT or have limited access to PGT due to perceived low risk (e.g., more than 20% of Canadians are recent immigrants). Furthermore, ethnic minorities have the lowest CGT rates. 20,Therefore, instead of relying on those at risk to present or take PGT, a PSDCoR approach in HCs and other families with genetic disease risk may be necessary. Accurate prediction of PofO via POAga would be of significant value to a healthcare system where the timeline for PGT adoption is several years away. PSDCoR facilitated by POAga of individuals with PVs identified from peridiagnostic cancer genetic testing, prior to PGT adoption, could provide real benefits in improving CGT rates and identifying all HC PVs in the population within 10 years if 70% of first-, second-, and third-degree relatives are tested. 32 Finally, once PGT is established, POAga remains important to ensure that those at highest risk are identified and to best inform medical management in redefining risk for family members who decline both PGT and CGT. Use of POAga will be necessary to realize the full potential of precision healthcare, now and in the future.

[0014] In summary, accurate assignment of variant PofO can aid clinical genetics in many ways, including curating variants, determining recurrence risk in future pregnancies for apparent de novo variants identified in the zygote, fetus, or child, efficiently screening relatives for genetic disease, predicting segregation of cancer or other disease susceptibility risk into straightforward management recommendations, and assessing disease risk when a pathogenic variant has a PofO effect, i.e., when a patient's risk of disease depends on which patient it was inherited from (e.g., hereditary paraganglioma-pheochromocytoma syndrome caused by pathogenic variants in SDHD, SDHAF2, and MAX). 13~17 Cascade genetic testing is used to identify pathogenic variants associated with disease, such as inherited cancer or other treatable Mendelian disease risk, with the goal of preventing or catching cancer early, or intervening in disease or disease risk, in family members. 18In the absence of PofO information, either because parents are unavailable, deceased, or decline genetic testing, cascade genetic testing must be offered to both family members until separation is confirmed. This can be costly and burdensome for patients and families, exacerbating the already low uptake of cascade genetic testing. 19、20 Eliminating the need to test one-sided family members is a clear benefit and there is major clinical utility in defining the PofO for pathogenic variants, and more broadly, establishing chromosome-length haplotypes with precise parental segregation of genomic variations has widespread applications.

[0015] There is a general need for improved methods of determining the parental origin for alleles associated with various conditions or diseases, particularly methods that do not require sequence data from family members.

[0016] The foregoing examples of the related art and limitations associated therewith are intended to be illustrative and not exhaustive. Other limitations of the related art will become apparent to those of ordinary skill in the art upon reading the specification and studying the drawings. Summary of the Invention

[0017] The following embodiments and aspects thereof are described and illustrated in conjunction with systems, tools, and methods that are meant to be exemplary and illustrative, not limiting in scope. In various embodiments, one or more of the problems discussed above are reduced or eliminated, while other embodiments are directed to other improvements.

[0018] One embodiment provides a method for assigning uniparental origin to haplotypes, subhaplotypes, or alleles associated with haplotypes. Chromosome-length haplotypes of a genome are generated. A methylation variable status of at least one imprinted methylation variable region (iDMR) associated with each autosome of a subject is determined. The determined methylation variable status of the at least one iDMR is correlated with each one of the chromosome-length haplotypes to assign a uniparental origin for each one of the chromosome-length haplotypes, including the subhaplotypes and alleles associated with the haplotypes. In some embodiments, the step of generating chromosome-length haplotypes is performed by using at least one sequencing method. In some embodiments, partial haplotypes are generated rather than assembling chromosome-length haplotypes.

[0019] In one embodiment, the chromosome length haplotypes are determined by performing a first sequencing method that allows for the determination of long-range phase information, and a second sequencing method that allows for the accurate determination of at least short reads of the sequences and the assignment of the short reads of the sequences to haplotypes. The results from the first and second sequencing methods are used to generate the chromosome length haplotypes. The methylation status of at least one iDMR associated with each autosome of the subject is optionally determined using a second sequencing method, and the determined methylation status of the at least one iDMR is used to assign a uniparental origin for each one of the chromosome length haplotypes. In some embodiments, the results of the first and second sequencing methods are used iteratively to generate the chromosome length haplotypes.

[0020] In some embodiments, the methods are used to assign parental origin to a haplotype, subhaplotype, or allele associated with a haplotype by assessing the methylation status of iDMRs associated with each one of a subject's haplotype pairs using only a sample obtained from the subject, i.e., without reference to parental or familial sequence data.

[0021] In addition to the exemplary aspects and embodiments described above, further aspects and embodiments will become apparent by reference to the drawings and by study of the following detailed descriptions. [Brief description of the drawings]

[0022] Exemplary embodiments are illustrated in the referenced figures in the drawings. It is intended that the embodiments and drawings disclosed herein are to be considered illustrative and not restrictive. [Figure 1] 1 illustrates an example embodiment of a method for determining the uniparental origin of a haplotype, subhaplotype, or allele associated with a haplotype. [Diagram 2] 1 shows a second example embodiment of a method for determining the uniparental origin of a haplotype, subhaplotype, or allele associated with a haplotype. [Figure 3A] 13 illustrates a third example embodiment for determining the uniparental origin of a haplotype, subhaplotype, or allele associated with a haplotype. [Figure 3B] 13 illustrates a fourth example embodiment for determining the uniparental origin of a haplotype, subhaplotype, or allele associated with a haplotype. [Figure 4] Figure 1 shows an overview of the phasing method used in an example embodiment to determine uniparental origin. The top panel shows the workflow including the inputs. The middle panel shows the combination of Strand-seq and nanopore reads to build chromosome-length haplotypes without assignment of each homolog to its uniparental origin. The bottom part of the figure shows how iDMR can be used to distinguish between maternal and paternal homologs to assign a uniparental origin for each one of the chromosome-length haplotypes. [Diagram 5]Comparison of nanopore-only and Strand-seq phasing. The left panel (a) shows sub-chromosomal nanopore phase blocks on chromosome 1. The right panel (b) shows chromosome-scale haplotypes obtained by phasing nanopore-detected variants using Strand-seq for chromosome 1. [Figure 6] In one example, CpG methylation in the paternal and maternal iDMRs (extending upwards) and the paternal iDMR (extending downwards) is shown. [Figure 7] Chromosome-wise results for parental origin assignment of het-SNVs in each of the five genomes tested in one example are shown. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0023] Throughout the following description, specific details are set forth to provide a more thorough understanding to those skilled in the art. However, well-known elements may not have been shown or described in detail to avoid unnecessarily obscuring the present disclosure. Thus, the description and drawings should be regarded in an illustrative sense, rather than a restrictive sense.

[0024] Hundreds of loci in the human genome have alleles that are differentially methylated according to their uniparental origin. DNA methylation at these imprinted loci, termed imprinted methylation variable regions or iDMRs, generally shows little variation across tissues, individuals, and populations. We have shown that such loci can be used to distinguish maternal and paternal homologs for all autosomes without the need for parental DNA. We have integrated the determination of the methylation status of iDMRs using a methylation-detecting sequencing technology (nanopore sequencing in the exemplary embodiment) with long-range phase information obtained using another sequencing technology (Strand-seq in the exemplary embodiment) to determine the uniparental origin of chromosome-length haplotypes for both DNA sequence and DNA methylation. We show that the overlap between SNVs in iDMRs and SNVs defined by sequencing strategies (e.g., Strand-seq) that allow the determination of long-range phase information allows for the first time complete uniparental origin aware haplotyping or phasing without analysis of parental genomes. We validated the determination of uniparental origin in five trios from various global populations. Our method correctly estimates uniparental origin for all autosomes with low mismatch error rates. Since the method can determine whether an inherited disease allele can be maternally or paternally derived, the method can be used for the diagnosis, treatment, and management of genetic diseases. Since the method can also determine whether any autosomal variant can be maternally or paternally derived, the method can also be used for the discovery, investigation, and exploration of genetic ancestry, genotypes, pharmacogenomic variants, and forensic research.

[0025] Some embodiments provide methods and devices for assigning uniparental origin of autosomes using parental imprints and chromosome length haplotypes. In one aspect, the invention provides methods and devices for assigning greater than 99% of autosomal alleles to uniparental origin.

[0026] Some embodiments provide methods and apparatus related to distinguishing alleles from maternal and paternal homologs to resolve haplotypes of parental chromosomes in a diploid genome. In one aspect, the invention provides methods and apparatus for resolving haplotypes of the human genome. In alternative aspects, the invention provides methods and apparatus for resolving haplotypes of other mammalian and diploid genomes.

[0027] The present invention provides a method for determining whether a genetic disease allele may be derived from a subject's mother or father for the diagnosis and management of genetic diseases. Knowledge of parental segregation provides implications for variant curation, understanding disease penetrance, patient and parent management implications, and facilitating focused cascade genetic testing. Such diseases include genetic syndromes such as hereditary cancer syndromes exhibiting parent-of-origin effects, including SDHD, SDHAF2, and MAX-associated hereditary paraganglioma-pheochromocytoma syndrome, as well as other hereditary cancer syndromes such as hereditary breast and ovarian cancer, neighbor-neighbor syndrome, etc. Diseases also include, but are not limited to, treatable cancers and non-cancer related diseases, and include, but are not limited to, those identified by the American College of Medical Genetics and Genomics (ACMG) Recommendations for Reporting of Incidental Findings in Clinical Exome and Genome Sequencing. Some embodiments of the present invention can be applied to any Mendelian disease or trait. These can be found in Online Mendelian Inheritance in Man (OMIM®), a continuously updated catalog of human genes and genetic disorders and traits. Some embodiments of the invention can be applied to pharmacogenetic variants and markers of ancestry. Some embodiments of the invention can reveal somatic mutational signatures inherited from either parent to a child that may affect the child's risk for cancer or disease, and can infer ongoing mutational processes in the parents that may be related to the parents' own cancer risk or disease status.

[0028] In one aspect, the invention provides a method for identifying the parental origin of an allele where there is limited or no access to parental and / or familial DNA samples. In another aspect, the invention provides a method for identifying the parental origin of an allele that may be mosaic in a parent, which may inform a parent of recurrence risk or disease risk who may otherwise test negative for the allele or pathogenic or likely pathogenic variant.

[0029] In one aspect, the invention provides methods for the identification, validation and screening of complex genetic traits for mammalian and diploid genomes. In one aspect, the invention provides methods for livestock genotyping for animal breeding.

[0030] In another aspect, the invention provides a method for delineating ancestral background from either parent.

[0031] In another aspect, the present invention allows for improved haplotyping for use in pharmacogenomic evaluation, HLA haplotyping, cancer vaccines, forensics, and treatments using allele-specific suppression. For example, parent-of-origin effects have been observed in association with HLA loci. 58、59 Similarly, HLA haplotypes may predict the efficacy of cancer vaccines and / or cancer therapies and thus contribute to the selection of cancer vaccines and / or cancer therapies. 60 .

[0032] In another aspect, the present invention provides insight into the complete set of variants inherited from either parent, from which one can infer the mutational processes that occurred in either parent.

[0033] In another aspect, the invention facilitates the discovery of novel variants and genes through the segregation of alleles / variants / haplotypes with parental or familial disease phenotypes.

[0034] In another aspect, with precise elucidation of all autosomal variations, polygenic risks, and genetic and epigenetic modifiers inherited from each parent, the method can be used to inform comprehensive risk assessments of Mendelian traits.In another aspect, with precise elucidation of all autosomal variations, polygenic risks, and genetic and epigenetic modifiers inherited from each parent, the method can be used to inform comprehensive risk assessments of complex traits.

[0035] Some embodiments of the present invention provide methods and apparatus that utilize imprinted methylation variable regions (iDMRs) based on their uniparental origin to integrate megabase-scale phase blocks from nanopore sequencing with long-range phase information in DNA from sequencing techniques that can only generate sparse phase blocks (e.g., Strand-seq data) to provide maternal and paternal homologs for all autosomes. In one aspect of the present invention, nanopore sequencing is performed using a nanopore PromethION instrument. One skilled in the art may select alternative sequencing techniques.

[0036] Referring to Figure 1, a general method 100 for determining the uniparental origin of haplotypes, subhaplotypes or alleles in a diploid genome is shown. In step 102, haplotypes are obtained in any suitable manner, for example, by a suitable long-range sequencing technique capable of elucidating haplotypes (e.g., Strand-seq sequencing, Hi-C sequencing, 3C-Seq sequencing, 10X sequencing, continuous long-read (CLR) sequencing, high-fidelity (HiFi) sequencing, Pore-C sequencing, or other suitable techniques now known or later developed, or combinations thereof), or by sequencing haploid cells. In some embodiments, the long-range sequencing technique used is capable of elucidating haplotypes along the entire length of the chromosomes.

[0037] In step 104, information regarding the methylation status of imprinted methylated variable regions (iDMRs) is determined for each haplotype, e.g., for each autosome. The methylation status of iDMRs can be determined in any suitable manner, e.g., using nanopore sequencing, single-molecule real-time (SMRT) sequencing, bisulfite sequencing, or other technologies (whether now known or later developed) that allow for the determination of the methylation status of iDMRs, or a combination thereof.

[0038] In some embodiments, two different sequencing methods are used in steps 102 and 104 to determine both the long-range phase information for each haplotype and the methylation status of the iDMRs associated with each one of the haplotypes. In other embodiments, if future developments in sequencing technology result in a single sequencing technology that can determine both the long-range phase information and the methylation status of the iDMRs for each haplotype, steps 102 and 104 can be performed using that single sequencing technology.

[0039] In some embodiments, one sequencing technique is used both to help obtain the haplotypes in step 102 and to determine the methylation status of the iDMRs associated with each haplotype in step 104. For example, in one embodiment, nanopore sequencing is used both to help obtain the two haplotypes for a diploid genome in step 102 and to determine the methylation status of the iDMRs associated with each one of the haplotypes in step 104. However, because nanopore sequencing cannot provide long-range phase information, nanopore sequencing is combined with another sequencing technique in step 102 to obtain the complete haplotypes, e.g., Strand-seq, Hi-C, 3C-Seq, 10X, continuous long reads (CLR), high fidelity (HiFi), Pore-C, or other sequencing techniques that allow for the determination of long-range phase information.

[0040] In step 106, a parental origin for each haplotype is assigned based on the methylation status of the iDMR for each one of the haplotypes, for example, for each one of the autosomes.

[0041] In some embodiments, rather than performing long-range sequencing in step 102, haplotypes constructed from haploid cells (e.g., sperm or eggs), hypodiploid cell populations, or single chromosomes are used. 57 can be sequenced to generate accurate haplotypes against which the sequence data from step 104 can be phased. In some embodiments where haplotypes are constructed from haploid cells such as sperm or eggs, the parental somatic tissue cells (e.g., blood) from which such haploid cells are derived can be sequenced in step 104 to infer uniparental origin from such parental autosomal alleles, since haploid cells (e.g., sperm or eggs) are themselves imprinted during their formation.

[0042] Referring to FIG. 2, a second embodiment 200 of a general method for determining the uniparental origin of haplotypes, subhaplotypes or alleles in a diploid genome is shown. In step 202, a first sequencing method is performed on a sample from a subject that allows the determination of long-range phase information for each one of the haplotypes of the diploid genome. Any suitable sequencing technology (whether currently known or later developed) that allows the determination of long-range phase information can be used in step 202, for example, Strand-seq, Hi-C, 3C-Seq, 10X, continuous long read (CLR), high fidelity (HiFi), Pore-C, or combinations thereof.

[0043] In step 204, information regarding the methylation status of imprinted methylated variable regions (iDMRs) is determined for each haplotype, e.g., for each autosome. The methylation status of iDMRs can be determined in any suitable manner, e.g., using nanopore sequencing, single molecule real-time (SMRT) sequencing, bisulfite sequencing, or other technology (whether now known or later developed) that allows for the determination of the methylation status of iDMRs, or a combination of such technologies.

[0044] In some embodiments, two different sequencing methods are used in steps 202 and 204 to determine both the long-range phase information for each haplotype and the methylation status of the iDMRs associated with each one of the haplotypes. In other embodiments, if future developments in sequencing technology result in a single sequencing technology that can determine both the long-range phase information and the methylation status of the iDMRs for each haplotype, steps 202 and 204 can be performed using that single sequencing technology.

[0045] In step 206, chromosome-length haplotypes are generated using the sequence information obtained in steps 202 and 204. In some embodiments, the sequence information obtained in steps 202 and 204 is used iteratively to phase heterozygous variants, such as single nucleotide variants (SNVs) and / or indels, to obtain chromosome-length haplotypes.

[0046] In step 208, a parental origin is assigned to each one of the chromosome length haplotypes (e.g., autosomes) using the results of the methylation status of the iDMR associated with each one of the chromosome length haplotypes determined in step 206.

[0047] Referring to FIG. 3A, a third example embodiment 300 of one particular embodiment of a method for determining the uniparental origin of a haplotype, subhaplotype, or allele is shown. In step 302, a first sequencing method is performed on the sample that allows for the determination of long-range phase information for each one of the haplotypes of the diploid genome. For example, the first sequencing method may generate sparse chromosome-scale haplotypes. Any suitable sequencing technology (whether currently known or later developed) that allows for the determination of long-range phase information can be used in step 302, for example, Strand-seq, Hi-C, 3C-Seq, 10X, continuous long read (CLR), high fidelity (HiFi), Pore-C, or combinations thereof.

[0048] In step 304, a second sequencing method that allows for the determination of the methylation status of iDMRs associated with a particular DNA sequence is performed on the sample. The methylation status of the iDMRs can be determined in any suitable manner, for example, using nanopore sequencing, single molecule real-time (SMRT) sequencing, bisulfite sequencing, or other techniques (whether now known or later developed) that allow for the determination of the methylation status of iDMRs, or a combination of such techniques. The second sequencing method yields unphased variants (e.g., SNVs and / or indels) in 306.

[0049] In step 308, long-range phase information for the variants determined from the first sequencing method is used to phase the variants obtained using the second sequencing method into one of two haplotypes, namely HP1 or HP2. In some embodiments, the variants phased in step 308 are SNVs. In some embodiments, the variants phased in step 308 are indels, and SNPs flanking the indels are used to assign the indels to haplotypes. In some embodiments, the variants phased in step 308 are both SNVs and indels.

[0050] In some embodiments, in step 310, the phased variants obtained in step 308 are used to re-phase all variants (e.g., both SNVs and indels) identified by the first sequencing method into their respective haplotypes, i.e., HP1 or HP2.

[0051] In step 312, the sequence reads obtained using the second sequencing method are phased a second time using all of the phased variants (e.g., both SNVs and indels) obtained in step 310 to obtain a high density of phased variants in 314.

[0052] In step 316, the per-read methylation information from each read obtained by the second sequencing method is integrated with its phase information from 314. Thus, the methylation variable information obtained at each iDMR is phased to each read in either one of the two haplotypes (i.e., HP1 or HP2) to calculate the methylation frequency at each iDMR site for each haplotype. The methylation frequencies are then used to assign each haplotype to its uniparental origin in 318 based on the methylation status of the known iDMRs.

[0053] Referring to Figure 3B, a fourth example embodiment 350 of a method for determining the uniparental origin of a haplotype, subhaplotype or allele is shown. At 352, a first sequencing method is performed that allows for the determination of long-range phase information. At 354, a second sequencing method is performed that allows for the accurate determination of at least short reads of sequences and the assignment of those short reads to haplotypes at 358. In some embodiments, the second sequencing method also provides information about the methylation status of the short sequence reads. At 360, variants (e.g., SNPs and / or indels) can optionally be iteratively phased to improve the phase information obtained at 362.

[0054] At 366, for example, using information obtained from the second sequencing method, the methylation variable status of at least one iDMR for each autosome of the subject is determined. At 368, a uniparental origin for each one of the chromosome length haplotypes is assigned based on the methylation variable status of the at least one iDMR for each autosome of the subject.

[0055] In some embodiments, the methods are used to assign parental origin to a haplotype, subhaplotype, or allele by assessing the methylation status of iDMRs associated with each one of the subject's haplotype pairs using only samples obtained from the subject.

[0056] In some embodiments, the methods are used to trigger cascade genetic testing by selecting an undesirable subhaplotype or allele, or a subhaplotype or allele of interest, determining the parent associated with the undesirable subhaplotype or allele, or the subhaplotype or allele of interest based on the determination of uniparental origin for the undesirable subhaplotype or allele, or the subhaplotype or allele of interest, and performing cascade genetic testing on family members of the parent associated with the undesirable subhaplotype or allele, or the subhaplotype or allele of interest. In some such embodiments, the undesired alleles are SDHD, SDHAF2, MAX, BRCA1, BRCA2, MLH1, MSH2, MSH6, PMS2, EPCAM, ATM, PALB2, TP53, APC, ACTA2, ACTC1, ACVRL1, AIP, ALK, ANKRD26, APOB, ARMC5, ATP7B, ATR, AXIN2, BAG3, BAP1, BARD1, BLM, BMPR1A, BRIP1, BTD, BUB1B, CACNA1S, CASQ2, CASR, CDC73, CDH1, CDK4, CDKN1B, CDKN1C, CDKN2A, CEBPA, CFTR, CHEK1, CHEK2, COL3A1, CPA1, CTC1, CTNNA1, CTRC, CYLD, DDB2, DDX41, DES, DICER1, DIS3L2, D KC1, DLST, DROSHA, DSC2, DSG2, DSP, EGFR, EGLN1, ENG, ERCC1, ERCC2, ERCC3, ERCC4, ERCC5, ETV6, EXT1, EXT 2, EZH2, FAM175A, FAN1, FANCA, FANCB, FANCC, FANCD2, FANCE, FANCF, FANCG, FANCI, FANCL, FANCM, FBN1, FH , FLCN, FLNC, FOCAD, GAA, GALNT12, GATA2, GLA, GPC3, GREM1, HFE, HNF1A, HOXB13, HRAS, KCNH2, KCNQ1, KIF1 B, KIT, LDLR, LMNA, LZTR1, MC1R, MEN1, MET, MITF, MLH3, MRE11, MRE11A, MSH3, MUTYH, MYBPC3, MYH11, MYH7,MYL2, MYL3, NBN, NF1, NF2, NHP2, NOP10, NTHL1, OTC, PAX5, PALLD, PCSK9, PDGFRA, PHOX2B, PIK3CA, PKP2, POLD1, POLE, POLH, POT1, PRKAG2, PRKAR1A, PRSS1, PTCH1, PTCH2, P TEN, RAB43, RABL3, RAD1, RAD50, RAD51C, RAD51D, RB1, RBM20, RECQL, RECQL4, RECQL5, REST, RET, RINT1, RPE65, RPS20, RUNX1, RYR1, RYR2, SAMD9, SAMD9L, SCN5A, SDHA, SDH B, SDHC, SLC45A2, SLX4, SMAD3, SMAD4, SMARCA4, SMARCB1, SMARCE1, SPINK1, SRP72, STK11, SUFU, TERC, TERT, TGFBR1, TGFBR2, TINF2, TMEM127, TMEM43, TNNC1, TNNI3, TNNT2, TP53I3, TPM1, TRDN, TRIP13, TSC1, TSC2, TTN, TTR, TYR, VHL, WRAP53, WRN, WT1, XPA, XPC, or XRCC2 alleles, or any other gene currently known or later determined to be associated with a genetic disorder. A list of currently known addressable genetic disorders is available, for example, from the American College of Medical Genetics and Genomics. A list of currently known pathogenic variants in such genes can be found, for example, from ClinVar, e.g., ACTC1, ACVRL1, APC, APOB, ATP7B, BAG3, BMPR1A, BRCA1, BRCA2, BTD, CACNA1S, CASQ2, COL3A1, DES, DSC2, DSG2, DSP, ENG, FBN1, FLNC, GAA, GLA, HFE, HNF1A, KCNH2, KCNQ1, LDLR, LMNA, MAX, MEN1, MLH1, MSH2, MSH6, MUTYH, MYBPC3, MYH11, MYH7, MYL2, MYL3, NF2, OTC, PALB2, PCSK9, PKP2, PMS2, PRKAG2, PTEN, RB1, RBM20, RET, RPE65, RYR1, RYR2, SCN5A, SDHAF2, SDHB,Available in SDHC, SDHD, SMAD3, SMAD4, STK11, TGFBR1, TGFBR2, TMEM127, TMEM43, TNNC1, TNNI3, TNNT2, TP53, TPM1, TRDN, TSC1, TSC2, TTN, TTR, VHL, and WT1.

[0057] In some embodiments, the undesirable allele is a founder mutation commonly associated with a genetic disorder, such as BRCA1 187delAG, BRCA1 5385insC (also written as BRCA1 5382insC), BRCA2 6174delT.

[0058] In some embodiments, if the allele is a mutant form of succinate dehydrogenase complex subunit D (SDHD) and the parental origin is paternal, the method further comprises periodically evaluating the subject for paraganglioma or pheochromocytoma and / or performing cascade genetic testing. In some embodiments, if the allele is a mutant form of SDHD and the parental origin is maternal, the method further comprises performing cascade genetic testing.

[0059] In some embodiments, if the allele is a mutant form of succinate dehydrogenase complex assembly factor 2 (SDHAF2) and the parental origin is paternal, the method further comprises periodically evaluating the subject for paraganglioma and / or performing cascade genetic testing.

[0060] In some embodiments, if the allele is a mutant form of the MYC-associated factor X gene (MAX) and the parental origin is paternal, the method further comprises periodically evaluating the subject for pheochromocytoma and / or performing cascade genetic testing.

[0061] In some embodiments, if the allele is a mutant form of ATM, BRCA1, BRCA2, MLH1, MSH2, MSH6, EPCAM, PMS2, PALB2, TP53, CHEK2, BRIP1, RAD51C, RAD51D, or MUTYH and the parental origin is determined to be a parent with a family history of pancreatic cancer, the method further comprises initiating pancreatic cancer screening in the subject beginning at age 50 years or 10 years earlier than the earliest pancreatic cancer diagnosis in the family, whichever is earlier.

[0062] In some embodiments, the method further comprises determining whether the two or more variants are cis-phased or trans-phased based solely on samples obtained from the subject (i.e., without reference to samples obtained from either parent of the subject). In some aspects, understanding whether a variant is cis-phased or trans-phased can inform risk assessment. 15 .

[0063] In some embodiments, if the allele is a mutant form of the succinate dehydrogenase complex containing subunits A, B, C, or D and the pathogenic or likely pathogenic variant in SDHD is determined to be paternally inherited (i.e., of paternal uniparental origin), the subject is evaluated periodically for paraganglioma or pheochromocytoma and / or cascade genetic testing is initiated. If the allele is a mutant form of the succinate dehydrogenase complex containing subunits A, B, C, or D and the pathogenic or likely pathogenic variant in SDHD is determined to be maternally inherited (i.e., of maternal uniparental origin), tumor screening may not be necessary in asymptomatic carriers and only cascade genetic testing is initiated.

[0064] In some embodiments, the methods can be used to detect uniparental origin of aberrant methylation patterns in one or both of the determined haplotypes. For example, constitutive methylation of genes has been implicated in certain diseases, e.g., MLH1, MSH2, BRCA2, and many other genes have been implicated in cancers or disorders with aberrant methylation. 61~64 .

[0065] In some embodiments, the methods further comprise predicting a similar pharmacogenomic outcome for a parent carrying the favorable or unfavorable haplotype, subhaplotype, or allele detected in the subject, where parentage has been determined by assessing the uniparental origin of the favorable or unfavorable haplotype, subhaplotype, or allele detected in the subject.

[0066] In some embodiments, the method further comprises determining whether the allele is a disease-risk associated HLA haplotype, subhaplotype, or allele that has a known parental-of-origin effect, and if the haplotype, subhaplotype, or allele is determined to be a disease-risk associated HLA haplotype, subhaplotype, or allele that has a known parental-of-origin effect, periodically monitoring the subject for risk of disease based on the known parental-of-origin effect.

[0067] In some embodiments, the allele is an HLA haplotype, subhaplotype, or allele, and the method further comprises selecting a cancer vaccine or cancer therapy for the parent based on a determination that the parent of the subject is of uniparental origin of the HLA haplotype, subhaplotype, or allele in the subject.

[0068] In some embodiments, the method further comprises comparing the DNA sequence obtained from the crime scene sample with the haplotype, subhaplotype or allele information and parental origin information obtained from the subject, and concluding that the subject's identified parental origin is associated with the crime scene if the DNA sequence obtained from the crime scene sample matches the DNA sequence of one of the haplotypes, subhaplotypes or alleles from the subject.

[0069] In some embodiments, the haplotype, subhaplotype, or allele comprises an apparent de novo germline variant, and a biological parent of the subject tests negative for the germline variant, and the method further comprises determining a parental origin for the de novo germline variant to identify a biological parent at potential risk for recurrence of the germline variant, or determining the risk to the parent of having postzygotic somatic mosaicism indicating the parent's risk of developing the disease.

[0070] In some embodiments, the method further comprises identifying founder mutations by comparing haplotypes from two individuals with corresponding mutations to identify founder haplotypes.

[0071] In some embodiments, the method further comprises administering a therapy to the parent determined to be of uniparental origin of the undesired haplotype, subhaplotype or allele, the therapy optionally comprising an antisense oligonucleotide, an allele-specific targeting construct, or an allele-specific gene editing targeted to the undesired allele. In some embodiments, the undesired allele comprises a mutation involved in a disorder with a dominant-negative effect, optionally a mutation involved in a disorder with a dominant-negative effect as associated with an SNV, copy number variant (CNV), indel, methylation, or triple repeat.

[0072] In some embodiments, the methods can be used to provide improved structural variant characterization of haplotypes, including parental-of-origin structural variant information. 65、66 .

[0073] In some embodiments, the methods can be used to determine the parental origin of a haplotype, for example, to determine risk haplotypes, to determine ancestry, or to examine complex traits.

[0074] In some embodiments, the methods can be used to phase haplotypes used in tests to determine parent-of-origin effects, avoiding the need for trio-based genotyping or trio-based sequencing, as is currently required. 67 .

[0075] One particular embodiment of the present invention implements the following steps. a) Obtain nanopore long read and single-cell Strand-seq sequence data from a subject. The nanopore data is used to call variants, some of which are phased by Strand-seq in an inversion-aware manner. These phased variants are then used to phase nanopore reads, which are used to phase more variants and DNA methylation in an iterative manner. b) Strand-seq and nanopore reads are combined to construct chromosome-length haplotypes without examining DNA methylation, but the assignment of each homolog (i.e., chromosome-length haplotype) to haplotype 1 or 2 is random with respect to its parental origin (PofO). c) The DNA methylation status of iDMRs is used to identify the PofO for each chromosomal homolog.

[0076] In various other embodiments, devices and / or kits are provided for carrying out the above-mentioned methods. For example, in one embodiment, a device is provided that includes a first sequencing device for performing a first sequencing technique to provide long-range phase information, a second sequencing device for performing a second sequencing technique to provide a methylation variable status of at least one iDMR associated with a haplotype, subhaplotype, or allele, and a processor for analyzing the long-range phase information and the methylation variable status of the at least one iDMR to determine a uniparental origin of the haplotype, subhaplotype, or allele associated with the haplotype. In one embodiment, a kit is provided that includes instructions for performing the first sequencing technique to provide long-range phase information, instructions for performing the second sequencing technique to provide a methylation variable status of at least one iDMR associated with the haplotype, subhaplotype, or allele, and instructions for combining the long-range phase information and the methylation variable status of the at least one iDMR to determine a uniparental origin of the haplotype, subhaplotype, or allele associated with the haplotype. EXAMPLES

[0077] Further embodiments are described with reference to the following examples, which are intended to be illustrative and not limiting in scope.

[0078] Example 1.0 Parental Origin Detection and Chromosome-Scale Haplotyping Using Long-Read DNA Methylation Sequencing and Strand-seq We integrated methylation-detecting nanopore sequencing with long-range phase information in Strand-seq data to determine the uniparental origin (PofO) of chromosome-length haplotypes for both DNA sequence and DNA methylation in five trios with diverse genetic backgrounds. Uniparental origin was correctly inferred for all autosomes with an average mismatch error rate of 0.31% for SNVs and 1.89% for indels. Because our method can determine whether inherited disease alleles are maternally or paternally derived, we can confidently predict that implementation of this method will improve the diagnosis, treatment, and management of many genetic diseases.

[0079] The examples herein demonstrate that when nanopore methylation and iDMR are integrated with Strand-seq chromosome-length haplotypes, alleles along the entire length of each autosome can be assigned to the maternal or paternal homolog (Figure 4). This method does not require parental sequence data (trio information) or SNP linkage analysis, but instead relies on the fact that all human autosomes have at least one imprinted methylation variable region. The only input required is a sample of fresh whole blood or other viable cells that can be cultured. We have validated PofO assignment for heterozygous SNVs and indels against five gold standard trios from the Genome in a Bottle Consortium (GIAB), Human Genome Structural Variation Consortium (HGSVC), and 1000 Genomes Project (1KGP). 21~23 By tracing pathogenic variants throughout families with sequencing efforts directed at selected family members, the method described has the potential to transform cascade genetic testing and improve screening for genetic diseases.

[0080] Example 1.1 - Nanopore and Strand-seq enable chromosome-scale haplotyping The inventors have demonstrated the described approach using five human genomes, including NA12878, HG002 and HG005 from GIAB, HG00733 from HGSVC, and NA19240 from 1KGP. 21~23 For all samples, we used nanopore sequencing data at coverage depths ranging from 24 to 38X, and 42 to 220 Strand-seq libraries with combined coverage depths ranging from 2.78 to 9.46X per sample. Nanopore raw signals were base called and mapped to the human reference genome GRCh38, and SNVs and indels ("variants") were called from the nanopore reads using Clair3. 24 .

[0081] Although it is possible to phase all called variants for each sample using only nanopore reads, the resulting phase blocks are relatively short (N50 M±SD=4.85±3.66 Mb, "M" mean, "SD" standard deviation) and do not span entire chromosomes (Figure 5).

[0082] Therefore, we first applied inversion-aware Strand-seq phasing to nanopore SNVs to construct sparse chromosome-length haplotypes. Strand-seq phased 61.03%–95.02% of the common heterozygous SNVs between the ground truth and nanopore call sets with mismatch error rates (number of variants phased incorrectly / number of all variants phased) between 0.14%–1.36%, with each chromosome spanning a single phase block (Table 1, Figure 5). The Strand-seq phased SNVs were then used to phase the nanopore reads (fraction of reads with at least MAPQ 20 successfully phased M ± SD = 71% ± 9.6%), which were then used to re-phase all variants, achieving dense chromosome-scale haplotypes containing nearly all heterozygous SNVs and most indels (Table 1). Combining Strand-seq and Nanopore in this way enabled us to phase 99.37%-99.91% of heterozygous SNVs and 96.29%-98.77% of heterozygous indels that were present in both the ground truth and Nanopore cell sets, with mismatch error rates of 0.07%-0.54% for SNVs and 1.33%-2.43% for indels (Table 1).

[0083] Figure 5 shows a comparison of nanopore-only phasing and Strand-seq phasing. Panel (a) shows that sub-chromosomal nanopore phase blocks on chromosome 1 contain more than 99% of the called SNVs and more than 96% of the called indels. However, using nanopore-only phasing for PofO assignment results in M±SD=42.37%±7.13% PofO errors for SNVs and M±SD=42.82%±6.83% for indels per chromosome, because this means that any phase switch between phase blocks effectively assigns PofOs randomly to any phase block. To phase both indels and SNVs, we used WhatsHap v1.2.1 with options--indels--ignore-read-groups. Panel (b) shows, in contrast, that phasing nanopore-detected variants using Strand-seq results in chromosome-scale haplotypes with consistent PofO across each haplotype, as shown here for chromosome 1.

[0084] FIG. 4 shows an overview of the PofO phasing method used in this example. The top panel shows that the input for the workflow is nanopore long reads and data from a single-cell Strand-seq library. The nanopore data is used to call variants, some of which are phased by Strand-seq in an inversion-aware manner. These phased variants are then used to phase nanopore reads, which are used to phase more variants and DNA methylation. Finally, the DNA methylation status of the iDMRs is used to identify the PofO for each chromosome homolog. The bottom panel shows that Strand-seq and nanopore reads can be combined to construct chromosome-length haplotypes without examining DNA methylation, but the assignment of each homolog (i.e., chromosome-length haplotype) to HP1 or HP2 (haplotype 1 or haplotype 2) is random for its PofO (upper part of the illustration). However, as shown by the lower part of the figure, the iDMRs can be used to distinguish between maternal and paternal homologs. Lollipop marks the positions of all 149 maternal iDMRs (methylated on the maternally homologous) and all 56 paternal iDMRs used in this study. See Table 2 for iDMR names and positions shown relative to the cytobands.

[0085] [Table 1]

[0086] [Table 2-1]

[0087] [Table 2-2]

[0088] [Table 2-3]

[0089] [Table 2-4]

[0090] [Table 2-5]

[0091] [Table 2-6]

[0092] Example 1.2 - PofO detection using iDMR PofO-specific DNA methylation at iDMRs provides a unique source of information for determining the PofO of homologs represented by chromosome-length haplotypes without using parental sequence data. We compiled a list of 205 iDMRs from previous genome-wide studies. 25~29Chromosome X was ignored because it does not have any known iDMRs. We combined DNA methylation information from phased nanopore reads with known PofO information in imprinted intervals to assign PofOs to each homolog. On average, six iDMRs (median=5, SD=5.8, range 1-32) were used for PofO assignment of each chromosome, and each chromosome was assigned to its parental origin with an average 96.3% confidence score (median=99.2%, SD=6.4%, range 60.7%-100%) (see Figure 6 showing representative data for CpG methylation data used for PofO assignment in HG0002; parallel data was obtained for HG005, HG00733, NA12878, and NA19240 but not shown). On average, 6.9% of the iDMRs were in conflict with the majority of PofO assignments. However, because iDMRs are weighted by different degrees of methylation in each sample, competing iDMRs represented only 2.5% of the PofO contribution value (Example 1.4 - x as described in Materials and Methods below).

[0093] Figure 6 shows CpG methylation of paternal and maternal iDMRs used for parental origin assignment in HG002. Maternally methylated iDMRs extend upwards and paternally methylated iDMRs extend downwards. Bars represent the proportion of CpGs with methylation difference > 0.35 (variable methylation) between haplotypes (HP1-HP2 for haplotype 1, and HP2-HP1 for haplotype 2) in each iDMR for each haplotype.

[0094] We examined 220 autosomal homologs across five individuals in this study (5 individuals × 22 autosomes × diploid) and compared the inferred PofOs to the trio-assigned PofOs in the ground truth phased variant call set. All 220 homologs were correctly assigned to PofOs, i.e., chromosome-length haplotypes were correctly characterized as either maternal or paternal, with few phasing errors (chromosome-level mismatch error rates for SNVs: M±SD=0.34%±0.53%, range 0.03%-4.86% (Figure 7); for indels, M±SD=1.93%±0.58%, range 0.98%-5.35%). Figure 7 shows the chromosome-wise results for PofO assignment of heterozygous SNVs in each of the five individuals for chromosomes 1-22 represented from left to right. PofOs could be assigned to all homologs. The bars, from left to right, show (i) all heterozygous SNVs in ground truth, (ii) common het-SNVs between ground truth and nanopore, (iii) het-SNVs with assigned PofO, and (iv) het-SNVs with correct assigned PofO. Only a small fraction of the variants with incorrect PofO are sporadic phasing errors in Strand-seq or nanopore data.

[0095] To further confirm that PofO phasing extracts reliable parental information, we calculated the Mendelian error rates between each child's inferred parental haplotypes and the ground truth variant genotypes for the children's parents. For the HG005 genome, the Mendelian error rates for maternal-maternal and paternal-paternal comparisons were low (M±SD=0.27%±2.69%, calculated for non-overlapping bins of 1000 variants), whereas for maternal-paternal and paternal-maternal comparisons they were high (M±SD=25.75%±14.14%, representing misassigned PofOs). For maternal-maternal and paternal-paternal comparisons, the highest average error rate for any chromosome was 2.29% for chromosome 8 in HG002. This is less than eight times lower than the lowest average error rate for any chromosome in the maternal-paternal and paternal-maternal comparisons (19.69% for chromosome 21 in NA12878), suggesting that the PofO assignments are correct for all chromosomes. Similar results were demonstrated for the Mendelian error rates for the HG002, HG00733, NA19240 and NA12878 genomes.

[0096] Example 1.3 - Analysis of experimental data The experiments described herein show that chromosomal homologs represented by chromosome-length haplotypes of SNVs and indels can be assigned PofOs without the use of parental sequence data. Long nanopore reads provide DNA sequence information along with PofO information in the form of DNA methylation differences between maternal and paternal alleles at known iDMRs. Strand-seq libraries provide sparse global haplotype information that phases variants and nanopore reads to reconstruct individual homologs. The PofO of each homolog can then be determined based on the consensus of one or more embedded iDMRs (Figure 4).

[0097] PofO phasing has the potential to address immediate clinical needs in the diagnosis and management of genetic diseases. These include improving variant curation and disease penetrance estimation by cosegregation of variants into each side of family members with and without the associated disease phenotype, determining which parents may be at risk of mosaicism in the context of a de novo variant, and establishing appropriate screening recommendations for pathogenic variants in genes with known PofO effects, such as those seen in SDHD and SDHAF2. Furthermore, PofO phasing offers considerable advantages over current clinical trials in facilitating cascade genetic testing that allows for opportunities for intervention in treatable genetic diseases. 31 Contacting, counseling, and testing relatives represents a significant logistical and financial burden for patients and the health care system, especially when considering adult-onset conditions where parental testing is often not possible. Cascade genetic testing may be hindered by limited intrafamilial communication and disrupted family structures, leading to low uptake in minority populations. 20 PofO phasing allows for a focused approach to cascade genetic testing across families, bringing the goal of optimal cascade genetic testing rates within reach. 32 Importantly, the ability of PofO phasing to infer with high certainty the pathogenic variant status of a patient's parents is likely to place a great emphasis on on-the-job alerting of individuals at risk for actionable genomic findings that may have been sought primarily or secondarily through the genetic testing process. Similar issues are already familiar to clinical genetics in the obligate carrier setting, but because this approach only requires testing a single individual to reconstruct the full genomic contribution from each parent, ethical considerations exist if PofO phasing is to be incorporated into mainstream clinical genetic testing due to its unprecedented scale.

[0098] We used a set of well-validated known iDMRs that are reported in at least two studies or confirmed in 179 WGBS datasets from 119 blood and 60 tissue samples. Using this set of iDMRs, we were able to assign PofO for all tested samples in all autosomes. Paternal or maternally derived methylation at iDMRs is always concordant when only one allele is methylated, whereas imprinted methylation can vary in the sense that two parental alleles may have similar amounts of DNA methylation in some tissues and individuals. 27、33 This may make it impossible to assign PofOs in some chromosomes in some individuals. However, with the exception of chromosome 17, which has a single iDMR, and chromosome 2, which has two iDMRs, all autosomes have at least three iDMRs, which should allow PofO assignment even in the presence of limited inter-individual and inter-tissue variability. In principle, this redundancy would also make PofO phasing more robust against epimutations and genomic imprinting disorders that may alter DNA methylation at iDMRs. 34 .

[0099] Furthermore, in a few iDMRs in some samples, such as maternally methylated TRPC3 on chromosome 4 in NA12878, we detected hypermethylation on alleles that were reported to be unmethylated. This explains the low confidence scores for PofO assignments for a few chromosomes, such as chromosome 4 in NA12878, which has the lowest confidence score (60.7%). Such discrepancies may be due to inaccuracies in the methylation calling or phasing of nanopore reads, or may reflect random allelic DNA methylation. Improvements to the current iDMR list will potentially reduce such errors in the future. DNA methylation-based (canonical) imprinting has been described in all placental mammals, and genomic maps of iDMRs have been established for many species, especially mice and primates. 7、35~37 Thus, the approach described herein can potentially be extended to other mammals.

[0100] Even if a homolog is assigned the correct PofO overall, local phasing errors can cause incorrect PofO assignments for some variants. The chromosome-length haplotypes constructed in the examples described herein are highly accurate, but have an average mismatch error rate of 0.31% for SNVs and 1.89% for indels. We identified only 61.3% of indels in the ground truth dataset, which reflects the limitations of current nanopore technology and could be improved by the addition of short Illumina reads. 24、38We observed rare switch errors, primarily for SNVs and indels at centromeres and inversions (e.g., an inversion on chromosome 8 in HG002 caused the highest mismatch error rate, FIG. 7), which generally contain small numbers of variants. Phasing errors at centromeres are likely due to misaligned reads in repetitive sequences, whereas errors in inversions are due to changes in sequence orientation that destroy the directional information that Strand-seq exploits for phasing. 8 The phasing errors associated with inversions can be partially addressed by the novel StrandPhaseR function, which rephasing variants inside a known inversion. 39 This is essential when the iDMR falls inside an inversion that may support an erroneous PofO if the phasing is not corrected (e.g., iDMRs RIMBP3 and CDRT15P6) or when the gene of interest falls inside an inversion (e.g., PMS2 in inversion 7 gene:5850673-6795880).

[0101] Sequencing costs for PofO phasing are relatively low, with only 24X nanopore and 3X Strand-seq coverage used in the examples described herein. DNA methylation information underlying PofO assignment is robust and can be easily extracted from nanopore sequence data, while previously rare Strand-seq libraries can now be generated in large numbers (>1000) at reduced cost. 40 In principle, one could also exploit genomic regions that are identical by descent in distant relatives to partially assign PofOs to large SNV datasets, using either parental sex chromosomes or ethnicity, but such bioinformatics approaches require that parents differ substantially in genetic background and are subject to well-known ethnic bias in genomic datasets. 41Given the simplicity and accuracy of PofO phasing, the current lack of trio-free alternatives for extracting PofO information from genomic data, and the notable clinical applications of the method, PofO phasing has the potential to become a routine component of genomic analysis.

[0102] Example 1.4 - Overview of Experimental Technique Nanopore sequencing and data We sequenced native DNA from Ashkenazi son (GM24385 or HG002) at 32-fold coverage on a nanopore PromethION instrument using library preparation and sequencing protocols previously described. 6 In addition to HG002, we used public nanopore data for HG005, HG00733, NA12878, and NA19240. Raw nanopore fast5 files for HG005 and HG00733 were downloaded from the Human Pangenome Reference Consortium. 42 , NA12878, taken from Jain et al. 2018. 43 , NA19240, was obtained from De Coster et al. 44 For HG002, HG005, and NA12878, paternal and maternal variant data and ground truth phased variants were obtained by GIAB v4.2.1, and for NA19240 and HG00733, parental phased variants were obtained from 1KGP shapeit2 v2a. 22、23 .

[0103] Nanopore data analysis Base calling and mapping: Nanopore signal level data were base called using Oxford Nanopore Technologies Guppy base caller version 6.0.1 and the ultra-precision model (dna_r9.4.1_450bps_sup) with default settings. Base-called nanopore reads were mapped to the human reference genome (GRCh38) using minimap2 version 2.24 with the --MD and -L options selected. 45 .

[0104] Variant calling: During alignment, variants were called from the nanopore alignment data using Clair3 version 0.1-r10 with trained model r941_sup_g5014 and default settings. 24 High quality variant calls from Clair3 (marked as "PASS" by the software) were then used for Strand-seq phasing (see next section).

[0105] Methylation calling: To call DNA methylation from nanopore data and obtain per-read CpG methylation information, we used nanopolish version 13.3 with default settings. 5 Read-by-read methylation call data were then preprocessed using NanoMethPhase v1.0 with --callThreshold 1.5 for downstream analysis and PofO phasing. 6、46 .

[0106] Strand-seq data processing, phasing, and inversion correction The inventors of the present application used 45 public strand-seq libraries for HG005 and 66 for HG002 from GIAB. 22、47 , and 230 libraries for HG00733 and 234 libraries for NA19240 from HGSVC. 21We used 96 previously described high-depth OP-Strand-seq libraries for NA12878 (clusters 5 and 6). 40 .

[0107] We trimmed adapters from paired-end FASTQ files and removed short reads (<30 bp) and low-quality bases (<15) using Cutadapt. 48 We aligned the reads to the GRCh38 human reference genome using Bowtie2 and discarded reads that had a MAPQ below 10 or did not map to chromosomes 10-22, X, and Y. 49 We used Picard (from the Broad Institute of MIT and Harvard, available on GitHub) to mark duplicate reads, then ran ASHLEYS QC with default settings and window sizes of 5,000,000, 2,000,000, 1,000,000, 800,000, 600,000, 400,000, and 200,000 to discard libraries with Strand-seq quality scores below 0.5. 50 .

[0108] We ran BreakpointR (BreakpointR from daewoooo of the Department of Genome Sciences at the University of Washington, commit 58cce0b09d01040892b3f6abf0b11caeb403d3f5, available on GitHub) with background set to 0.1, chromosomes set to autosomes, and maskRegions set to the blacklist previously described. 30、51We used 8 Mb bins because we found that they make phasing across difficult regions such as inversions easier to link and prevent large switch errors. We used the function exportRegions with default settings to identify regions of the genome that have both Watson and Crick reads suitable for phasing. We phased biallelic heterozygous SNVs called from the nanopore data for each sample using num.iterations set to 3, splitPhasedReads and assume.biallelic set to TRUE, using R v4.0.5 with StrandPhaseR (commit bb19557235de3d82092abdc11b3334f615525b5b from the devel branch of StrandPhaseR from daewoooo of the Department of Genome Sciences at the University of Washington, available on GitHub) with dependency rlang v1.0.1 or higher. 11 .

[0109] Inversion destroys the directional phase information of Strand-seq. We called the inversion for each sample using the R package InvertypeR (vincent-hanlon, in InvertypeR / commits / main, commit a5fac3b6b8264db28de1a997ad0bc062badea883, available on GitHub). 51Briefly, we used nanopore SNV to create a pair of composite files for each sample and added the genomic coordinates in the blacklist, chromosome 8:8231088-12039415, to ensure that common large inversions at these coordinates were correctly represented. We genotyped the catalog of published inversion coordinates using priors as previously described with adjust_method set to "all" and identified a list of de novo sample-specific strand switches by running BreakpointR three times on the composite files with different bin sizes. 30、51 For the latter, we used prior probabilities of 0.9, 0.05, and 0.05 for the reference, heterozygous, and homozygous genotypes, respectively. We combined inversions with posterior probabilities above 0.95 from the two call sets by discarding any inversions from the catalog call set that were intersected with the de novo call set (bedtools intersect -v -r -f 0.1). We did not remove misoriented reference contigs, which appear as homozygous inversions in all samples, because this would destroy phasing in the same way that inversions do.

[0110] The function correctInvertedRegionPhasing in the StrandPhaseR package switches the phase of heterozygous SNVs within homozygous inversions and rephases SNVs within heterozygous inversions. 39We used sample-specific inversion calls larger than 10 kb with nanopore sample-specific SNV locations, recall.phased and assume.biallelic set to TRUE, het.genotype set to "lenient", lookup.bp set to 1000000, background set to 0.1, and lookup.blacklist set to blacklist above. The resulting chromosome-length inversion-corrected SNV haplotypes were used to phase nanopore reads against each other.

[0111] iDMR, chromosome-scale haplotype, and PofO detection iDMR Verification We compiled a list of previously reported iDMRs from five previous genome-wide studies. 25~29 iDMRs with overlaps between two or more studies were merged. This resulted in 102 merged iDMRs and 326 iDMRs reported in only a single study. We previously investigated imprinted methylation genome-wide using 12 nanopore sequenced cell lines together with their trio sequencing information from 1KGP. 29We used the same cell lines to examine the 326 iDMRs from the single study mentioned above. For each CpG site with coverage >4 within the iDMR, we calculated the methylation frequency (percentage of reads methylated at the CpG). We then calculated the difference between the average methylation frequency for the paternal and maternal alleles for each iDMR in each cell line. We excluded 94 iDMRs with |methylation difference| ≥ 0.25 between alleles and with conflicting PofO between any of the 12 cell lines and the corresponding prior study. To further validate the 232 remaining iDMRs reported in single studies, we used WGBS datasets for 119 blood samples from 87 individuals in the Blueprint consortium and 60 tissue samples for 29 tissue types in the ENCODE and RoadMap consortia. 52~54In iDMRs, only one allele is methylated, and therefore the aggregated methylation frequency from both alleles in these regions is expected to be approximately 50% (partial methylation). Therefore, we investigated partial methylation in 232 iDMRs in the WGBS dataset. For each WGBS sample, we used CpGs with at least five mapped WGBS reads and counted the number of CpGs with partial methylation (methylation frequency of 0.35-0.65 among the mapped reads) in each iDMR. An iDMR is then considered to be partially methylated if it has at least five CpGs in the WGBS sample and more than 60% of the CpGs showed partial methylation. Of the 232 iDMRs, 129 iDMRs were excluded because they were partially methylated in less than 2 blood or tissue samples in which the iDMR could be interrogated (i.e., the iDMR had at least 5 CpGs with a coverage of ≧5) or in less than 5% of the blood or tissue samples. Overall, we compiled a list of 205 known iDMRs, of which 102 were reported in multiple studies and 103 (out of 326) were reported in a single study, the most probable of which are listed in Table 2.

[0112] Determination of parental origin We then integrated several steps to detect chromosome-scale haplotypes using their PofOs as follows. 1. Strand-seq phasing demonstrates sparse chromosome-scale haplotypes. Nanopore reads were phased into either HP1 or HP2 haplotypes using phased SNVs from strand-seq. A minimum mapping quality of 20 and a base quality of 7 were used to tag each read with a phased SNV. A read was tagged as HP1 if it had at least one phased SNV from HP1 and the ratio (number of SNVs from HP1 mapped to the read / total phased SNVs mapped to the read) was > 0.75, and vice versa. 2. All variants (SNVs and indels) were then re-phased for each haplotype using the phased nanopore reads from step 1. At least two phased reads are required to support a variant in order to assign it as HP1 or HP2. 3. The nanopore reads were then phased a second time using all phased variants from step 2 under the conditions mentioned in step 1. 4. Read-by-read methylation information for each nanopore read at known iDMRs was extracted and merged with its phase information from step 3. This allowed us to phase each CpG methylation in each read to either HP1 or HP2 and calculate the methylation frequency (number of methylated reads / number of total reads) at each CpG site for each haplotype. The methylation frequencies were then used to assign haplotypes to their PofO for each sample as follows:

[0113] In each of the 205 known iDMRs, we counted the CpGs with methylation frequency differences (methylation variables) between haplotypes of ≧0.35. We then calculated the contribution value of the iDMR to PofO detection for each haplotype as follows: x=ma / n where m is the average methylation frequency for the haplotype, a is the number of methylated variable CpGs that support the PofO for the haplotype, and n is the number of all CpGs in the iDMR. (HP1) -a (HP2) Only iDMRs were considered for PofO assignment. As an example, for a maternally methylated iDMR with 20 CpGs and an average methylation frequency of 0.8 in HP1 and 0.3 in HP2, if 12 CpGs show ≥ 0.35 methylation in HP1 compared to HP2 and 2 CpGs show ≥ 0.35 methylation in HP2 compared to HP1, x for HP1 as the maternal lineage and HP2 as the paternal lineage is x = 0.8 × 12 / 20, and x for HP1 as the paternal lineage and HP2 as the maternal lineage is x = 0.3 × 2 / 20.

[0114] If each chromosome for each haplotype is maternal or paternal,

[0115]

number

[0116]

number

[0117] 5-Finally, the phased variants from step 2 were assigned to their PofO using the results from step 4.

[0118] All steps are integrated into the workflow and tool PatMat and instructions are provided on GitHub (available as PatMat from vahidAK).

[0119] Mendel Error To validate the PofO assignments, we calculated the frequency of one Mendelian error between the PofO-assigned haplotypes and the parental genotypes. We obtained genotypes from GIAB for the parents of HG002 and HG005 (v4.2.1), from 1KGP for the parents of HG00733 and NA19240 (v2a), and from Byrska-Bishop et al. 2021 for the parent of NA12878. 22、23、47、55 For each parent-child pair, we examined the loci where we found the heterozygous genotype phased for the child and either the heterozygous or homozygous alternative genotype for the parent. If the child had the maternal reference allele and the mother was the homozygous alternative, we called it a Mendelian error (similarly for the paternal allele of the child and the father's genotype). We did this for non-overlapping bins of 1000 variants and calculated the error rate as the number of such Mendelian errors divided by the number of variants examined. We plotted the resulting error rates versus chromosomes using a RIdeogram. 56 .

[0120] While several exemplary aspects and embodiments have been described above, those skilled in the art will recognize certain modifications, permutations, additions, and subcombinations thereof. Accordingly, it is intended that the following appended claims, and any claims hereafter introduced, be interpreted as a whole to include all such modifications, permutations, additions, and subcombinations consistent with the broadest interpretation of the specification.

[0121] For example, while various example embodiments have been described with respect to determining chromosome-length haplotypes, and without being bound that determining chromosome-length haplotypes provides the most accurate results, alternative embodiments would assemble smaller portions of sequence data to generate at least partial haplotypes, so long as at least some variants associated with iDMRs are captured such that both the haplotype of the allele of interest and the methylation status of the iDMR associated with that allele determined determine a uniparental origin for that allele.

[0122] References The following references may be of interest to the subject matter described herein, each of which is incorporated herein by reference in its entirety. 1.Zheng,GXYet al.Haplotyping germline and cancer genomes with high-throughput linked-read sequencing.Nat.Biotechnol.34,303-311(2016). 2. Wenger, AMet al. Accurate consensus circular long-read sequencing improves variant detection and assembly of a human genome. Nat. Biotechnol. 37, 1155-1162 (2019). 3.Cheng,H.et al.Haplotype-resolved assembly of diploid genomes without parental data.Nat.Biotechnol.(2022)doi:10.1038 / s41587-022-01261-x。 4.H.,L.A.et al.Detection and mapping of 5-methylcytosine and 5-hydroxymethylcytosine with nanopore MspA.Proc.Natl.Acad.Sci.110,18904-18909(2013)。 5.Simpson,J.T.et al.Detecting DNA cytosine methylation using nanopore sequencing.Nat.Methods14,407(2017)。 6.Akbari,V.et al.Megabase-scale methylation phasing using nanopore long reads and NanoMethPhase.Genome Biol.22,68(2021)。 7.Gigante,S.et al.Using long-read sequencing to detect imprinted DNA methylation.Nucleic Acids Res.47,e46-e46(2019)。 8.Porubsky,D.et al.Direct chromosome-length haplotyping by single-cell sequencing.Genome Res.26,1565-1574(2016)。 9.Selvaraj,S.,R Dixon,J.,Bansal,V.& Ren,B.Whole-genome haplotype reconstruction using proximity-ligation and shotgun sequencing.Nat.Biotechnol.31,1111-1118(2013)。 10.Falconer,E.et al.DNA template strand sequencing of single-cells maps genomic rearrangements at high resolution.Nat.Methods 9,1107-1112(2012)。 11.Porubsky,D.et al.Dense and accurate whole-chromosome haplotyping of individual genomes.Nat.Commun.8,1293(2017)。 12.Porubsky,D.et al.Fully phased human genome assembly without parental data using single-cell strand sequencing and long reads.Nat.Biotechnol.39,302-308(2021)。 13.Van Der Mey,A.L.,Maaswinkel-Mooy,P.,Cornelisse,C.,Schmidt,P.& Van De Kamp,J.P.GENOMIC IMPRINTING IN HEREDITARY GLOMUS TUMOURS:EVIDENCE FOR NEW GENETIC THEORY.Lancet 334,1291-1294(1989)。 14.Knowles,J.W.,Rader,D.J.& Khoury,M.J.Cascade Screening for Familial Hypercholesterolemia and the Use of Genetic Testing.JAMA 318,381-382(2017)。 15.Richards,S.et al.Standards and guidelines for the interpretation of sequence variants:a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology.Genet.Med.17,405-424(2015)。 16.Hensen,E.F.et al.Somatic loss of maternal chromosome 11 causes parent-of-origin-dependent inheritance in SDHD-linked paraganglioma and phaeochromocytoma families.Oncogene 23,4076-4083(2004)。 17.Huai-Xiang,H.et al.SDH5,a Gene Required for Flavination of Succinate Dehydrogenase,Is Mutated in Paraganglioma.Science(80-.).325,1139-1142(2009)。 18.Hampel,H.Genetic counseling and cascade genetic testing in Lynch syndrome.Fam.Cancer 15,423-427(2016)。 19.Lee,P.W.C.et al.Evaluating the impact of universal Lynch syndrome screening in a publicly funded healthcare system.Cancer Med.9,6507-6514(2020)。 20.Braley,E.F.et al.Patient ethnicity and cascade genetic testing:a descriptive study of a publicly funded hereditary cancer program.Fam.Cancer(2021)doi:10.1007 / s10689-021-00270-0。 21.Chaisson,M.J.P.et al.Multi-platform discovery of haplotype-resolved structural variation in human genomes.Nat.Commun.10,1784(2019)。 22.Zook,J.M.et al.An open resource for accurately benchmarking small variant and reference calls.Nat.Biotechnol.37,561-566(2019)。 23.Auton,A.et al.A global reference for human genetic variation.Nature 526,68-74(2015)。 24.Zheng,Z.et al.Symphonizing pileup and full-alignment for deep learning-based long-read variant calling.bioRxiv 2021.12.29.474431(2021)doi:10.1101 / 2021.12.29.474431。 25.Court,F.et al.Genome-wide parent-of-origin DNA methylation analysis reveals the intricacies of human imprinting and suggests a germline methylation-independent mechanism of establishment.Genome Res.24,554-569(2014)。 26.Joshi,R.S.et al.DNA Methylation Profiling of Uniparental Disomy Subjects Provides a Map of Parental Epigenetic Bias in the Human Genome.Am.J.Hum.Genet.99,555-566(2016)。 27.Zink,F.et al.Insights into imprinting from parent-of-origin phased methylomes and transcriptomes.Nat.Genet.50,1542-1552(2018)。 28.Hernandez Mora,J.R.et al.Characterization of parent-of-origin methylation using the Illumina Infinium MethylationEPIC array platform.Epigenomics 10,941-954(2018)。 29.Akbari,V.et al.Genome-Wide Detection of Imprinted Differentially Methylated Regions Using Nanopore Sequencing.bioRxiv 2021.07.17.452734(2021)doi:10.1101 / 2021.07.17.452734。 30.Porubsky,D.et al.breakpointR:an R / Bioconductor package to localize strand state changes in Strand-seq data.Bioinformatics 36,1260-1261(2020)。 31.Miller,D.T.et al.ACMG SF v3.0 list for reporting of secondary findings in clinical exome and genome sequencing:a policy statement of the American College of Medical Genetics and Genomics(ACMG).Genet.Med.23,1381-1390(2021)。 32.Offit,K.et al.Cascading After Peridiagnostic Cancer Genetic Testing:An Alternative to Population-Based Screening.J.Clin.Oncol.38,1398-1408(2020)。 33.Prickett,A.R.& Oakey,R.J.A survey of tissue-specific genomic imprinting in mammals.Mol.Genet.Genomics 287,621-630(2012)。 34.Monk,D.,Mackay,D.J.G.,Eggermann,T.,Maher,E.R.& Riccio,A.Genomic imprinting disorders:lessons on how genome,epigenome and environment interact.Nat.Rev.Genet.20,235-248(2019)。 35.Renfree,M.B.,Hore,T.A.,Shaw,G.,Marshall Graves,J.A.& Pask,A.J.Evolution of Genomic Imprinting:Insights from Marsupials and Monotremes.Annu.Rev.Genomics Hum.Genet.10,241-262(2009)。 36.Cheong,C.Y.et al.Germline and somatic imprinting in the nonhuman primate highlights species differences in oocyte methylation.Genome Res.25,611-623(2015)。 37.Xie,W.et al.Base-Resolution Analyses of Sequence and Parent-of-Origin Dependent DNA Methylation in the Mouse Genome.Cell 148,816-831(2012)。 38.Shafin,K.et al.Haplotype-aware variant calling with PEPPER-Margin-DeepVariant enables high accuracy in nanopore long-reads.Nat.Methods 18,1322-1332(2021)。 39.Porubsky,D.et al.Haplotype-resolved inversion landscape reveals hotspots of mutational recurrence associated with genomic disorders.bioRxiv 2021.12.20.472354(2021)doi:10.1101 / 2021.12.20.472354。 40.Hanlon,V.C.T.et al.Construction of Strand-seq libraries in open nanoliter arrays.Cell Reports Methods 2,100150(2022)。 41.Ledford,H.Cancer geneticists tackle troubling ethnic bias in studies.Nature 154-155(2019)doi:10.1038 / d41586-019-01080-2.PMID:30967668。 42.Shafin,K.et al.Nanopore sequencing and the Shasta toolkit enable efficient de novo assembly of eleven human genomes.Nat.Biotechnol.(2020)doi:10.1038 / s41587-020-0503-6。 43.Jain,M.et al.Nanopore sequencing and assembly of a human genome with ultra-long reads.Nat.Biotechnol.36,338-345(2018)。 44.De Coster,W.et al.Structural variants identified by Oxford Nanopore PromethION sequencing of the human genome.Genome Res.29,1178-1187(2019)。 45.Li,H.Minimap2:pairwise alignment for nucleotide sequences.Bioinformatics 34,3094-3100(2018)。 46.Akbari,V.et al.NanoMethPhase.Zenodo(2021)doi:10.5281 / zenodo.4474430。 47.Wagner,J.et al.Benchmarking challenging small variants with linked and long reads.bioRxiv 2020.07.24.212712(2021)doi:10.1101 / 2020.07.24.212712。 48.Martin,M.Cutadapt removes adapter sequences from high-throughput sequencing reads.EMBnet.journal; Vol 17,No 1 Next Gener.Seq.Data Anal.-10.14806 / ej.17.1.200(2011)。 49.Langmead,B.& Salzberg,S.L.Fast gapped-read alignment with Bowtie 2.Nat.Methods 9,357-359(2012)。 50.Gros,C.,Sanders,A.D.,Korbel,J.O.,Marschall,T.& Ebert,P.ASHLEYS:automated quality control for single-cell Strand-seq data.Bioinformatics 37,3356-3357(2021)。 51.Hanlon,V.C.T.,Mattsson,C.-A.,Spierings,D.C.J.,Guryev,V.& Lansdorp,P.M.InvertypeR:Bayesian inversion genotyping with Strand-seq data.BMC Genomics 22,582(2021)。 52.Stunnenberg,H.G.et al.The International Human Epigenome Consortium:A Blueprint for Scientific Collaboration and Discovery.Cell(2016)doi:10.1016 / j.cell.2016.11.007。 53.Consortium,E.P.An integrated encyclopedia of DNA elements in the human genome.Nature 489,57(2012)。 54.Bernstein,B.E.et al.The NIH Roadmap Epigenomics Mapping Consortium.Nat.Biotechnol.28,1045-1048(2010)。 55.Byrska-Bishop,M.et al.High coverage whole genome sequencing of the expanded 1000 Genomes Project cohort including 602 trios.bioRxiv 2021.02.06.430068(2021)doi:10.1101 / 2021.02.06.430068。 56.Hao,Z.et al.RIdeogram:drawing SVG graphics to visualize and map genome-wide data on the idiograms.PeerJ Comput.Sci.6,e251(2020)。 57.Murphy,N.M.et al.Haplotyping the human leukocyte antigen system from single chromosomes.Sci Rep 6,30381(2016).https: / / doi.org / 10.1038 / srep30381。 58.Ramagopalan et al.Parent-of-origin of HLA-DRB1*1501 and age of onset of multiple sclerosis.”Journal of Human Genetics 54(2009):547-549。 59.Deighton and Walker.The familial nature of rheumatoid arthritis.Annals of the Rheumatic Diseases 1991,50:62-65。 60.Naranbhai V.et al.HLA-A*03 and response to immune checkpoint blockade in cancer:an epidemiological biomarker study.Lancet Oncol.2022 Jan;23(1):172-184.doi:10.1016 / S1470-2045(21)00582-9.Epub 2021 Dec 9。 61.de la Cruz,M.R.et al.Cis-acting factors causing secondary epimutations; impact on the risk for cancer and other diseases.Cancers(Basel).2021 Oct; 13(19):4807.Published online 2021 Sep 26.doi:10.3390 / cancers13194807。 62.Hitchins,M.P.et al.Dominantly inherited constitutional epigenetic silencing of MLH1 in a cancer-affected family is linked to a single nucleotide variant within the 5’UTR.Cancer Cell.2011 Aug 16;20(2):200-13.doi:10.1016 / j.ccr.2011.07.003。 63.Chan,T.et al.Heritable germline epimutation of MSH2 in a family with hereditary nonpolyposis colorectal cancer.Nat Genet 38,1178-1183(2006).https: / / doi.org / 10.1038 / ng1866。 64.Evans D.G.R.et al.A Dominantly Inherited 5’ UTR Variant Causing Methylation-Associated Silencing of BRCA1 as a Cause of Breast and Ovarian Cancer.Am J Hum Genet.2018 Aug 2;103(2):213-220.doi:10.1016 / j.ajhg.2018.07.002.PMID:30075112; PMCID:PMC6080768。 65.Dixon K.et al.Defining the heterogeneity of unbalanced structural variation underlying breast cancer susceptibility by nanopore genome sequencing.Eur J Hum Genet.2023 Feb 16.doi:10.1038 / s41431-023-01284-1.Epub ahead of print.PMID:36797466。 66.Thibodeau,M.L.,et al.Improved structural variant interpretation for hereditary cancer susceptibility using long-read sequencing.Genet Med.2020 Nov;22(11):1892-1897.doi:10.1038 / s41436-020-0880-8.Epub 2020 Jul 6.PMID:32624572; PMCID:PMC7605438。 67.Weinberg,C.R.Methods for detection of parent-of-origin effects in genetic studies of case-parents triads.Am J Hum Genet.1999 Jul; 65(1):229-235.doi:10.1086 / 302466。 68.Baysal,B.E.et al.Mutations in SDHD,a mitochondrial complex II gene,in hereditary paraganglioma.Science.Feb 4 2000;287(5454):848-51.doi:10.1126 / science.287.5454.848。 69.Bayley J.P.et al.Paraganglioma and pheochromocytoma upon maternal transmission of SDHD mutations.BMC Med Genet.Oct 10 2014; 15:111.doi:10.1186 / s12881-014-0111-8。 70.Hampel H.et al.Assessment of Tumor Sequencing as a Replacement for Lynch Syndrome Screening and Current Molecular Tests for Patients With Colorectal Cancer.JAMA Oncol.06 2018;4(6):806-813.doi:10.1001 / jamaoncol.2018.0104。 71.Beard V.K.et al.Genetic testing in families with hereditary colorectal cancer in British Columbia and Yukon:a retrospective cross-sectional analysis.CMAJ Open.2020 Oct-Dec 2020;8(4):E637-E642.doi:10.9778 / cmajo.20190167。 72.LaDuca H.et al.A clinical guide to hereditary cancer panel testing:evaluation of gene-specific cancer associations and sensitivity of genetic testing criteria in a cohort of 165,000 high-risk patients.Genet Med.Feb 2020;22(2):407-415.doi:10.1038 / s41436-019-0633-8。 73.Schrader K.A.,et al.Germline Variants in Targeted Tumor Sequencing Using Matched Normal DNA.JAMA Oncology.2016;2(1):104-111.doi:10.1001 / jamaoncol.2015.5208。 74.Frey M.K.,et al.Prospective Feasibility Trial of a Novel Strategy of Facilitated Cascade Genetic Testing Using Telephone Counseling.J Clin Oncol.May 1 2020;38(13):1389-1397.doi:10.1200 / jco.19.02005。

Claims

1. 1. A method of assigning parental origin to a haplotype, subhaplotype, or an allele associated with said haplotype in a subject, said method comprising: generating chromosome length haplotypes of the subject's genome; determining the methylation variable status of at least one imprinted methylation variable region (iDMR) for each autosome of the subject; and correlating the determined methylation variable status of at least one iDMR for each autosome of the subject with each one of the chromosome length haplotypes to assign a uniparental origin for each one of the chromosome length haplotypes.

2. 2. The method of claim 1, wherein generating chromosome length haplotypes of the genome comprises determining the chromosome length haplotypes by at least one sequencing method.

3. 3. A method of assigning uniparental origin to a haplotype, subhaplotype, or allele associated with said haplotype, as described in claim 2, wherein the step of generating chromosome length haplotypes of the genome comprises: (i) performing a first sequencing method that allows for the determination of long-range phase information; (ii) performing a second sequencing method that allows accurate determination and assignment of at least short reads of the sequence to haplotypes; (iii) using the overlapping results of the first and second sequencing methods to generate chromosome length haplotypes; and (iv) determining the methylation status of said at least one iDMR associated with each autosome; and (v) using the determined methylation status of the at least one iDMR associated with each autosome, assigning a uniparental origin for each one of the chromosome length haplotypes based on the methylation status of the at least one iDMR associated with each autosome.

4. 4. The method of claim 3, wherein results of the first and second sequencing methods performed in steps (i) and (ii) are used iteratively to phase variants in step (iii), the variants optionally comprising single nucleotide polymorphisms (SNPs) and / or indels.

5. 5. The method of claim 3 or 4, wherein step (iv) is carried out simultaneously with step (ii) using the second sequencing method.

6. 3. A method for assigning a parental origin to a haplotype, a subhaplotype, or an allele associated with said haplotype, as claimed in claim 2, comprising the steps of: (i) determining at least a portion of a sequence of one of each of said haplotypes using a first sequencing method that allows the determination of long-range phase information; (ii) determining the methylation variable status of at least one iDMR for each autosome of the subject using a second sequencing method; (iii) using the results of the first and second sequencing methods to generate chromosome length haplotypes; (iv) using the results of the second sequencing method to assign a uniparental origin for each one of the chromosome length haplotypes based on the methylation variable status of at least one iDMR for each autosome of the subject.

7. obtaining a sample from the subject; and performing said at least one sequencing method on said sample.

8. The method of claim 7, wherein the sample comprises a blood sample or a sample of viable cells that can be cultured.

9. 9. The method of claim 3, wherein the first sequencing method comprises Strand-seq, Hi-C, 3C-Seq, 10X, continuous long read (CLR), high fidelity (HiFi), Pore-C, or a combination thereof.

10. The method of any one of claims 3 to 9, wherein the second sequencing method comprises nanopore sequencing, single molecule real-time (SMRT) sequencing, bisulfite sequencing, or a combination thereof.

11. The method according to any one of claims 3 to 10, wherein the first sequencing method and the second sequencing method are the same sequencing method.

12. 11. The method of any one of claims 3-10, wherein the first sequencing method comprises StrandSeq and the second sequencing method comprises nanopore sequencing, and wherein nanopore data is used to call variants, and at least some of the variants are phased by Strand-seq in an inversion-aware manner to construct sparse chromosome-length haplotypes.

13. 3. A method for assigning a parental origin to a haplotype or to an allele associated with said haplotype as claimed in claim 2, comprising the steps of: (i) generating sparse chromosome-scale haplotypes using a first sequencing method; (ii) using the sparse chromosome-scale haplotypes to phase sequencing data obtained using a second sequencing method to generate first and second haplotypes; (iii) using the phased sequencing data from step (ii), re-phasing all sequence variants identified by the first sequencing method into each of the first and second haplotypes; (iv) phasing the sequencing data obtained using the second sequencing method to a second time; and (v) extracting per-read methylation information for each read obtained by the second sequencing method, and integrating the per-read methylation information at the at least one iDMR for each autosome of the subject with the phase information obtained in step (iv) to assign each read to either the first or second haplotype, and determining the uniparental origin for each of the first or second haplotypes.

14. 14. The method of claim 13, wherein the first sequencing method comprises Strand-seq and the second sequencing method comprises nanopore sequencing, and wherein a plurality of phased single base variants (SNVs) determined using the first sequencing method are used to phase the sequencing data obtained using the second sequencing method in step (ii).

15. each one of the reads from the second sequencing method is tagged as being from the first haplotype if the read has a number of SNVs from the first haplotype mapped to it that is greater than a number of SNVs from the second haplotype mapped to it; 15. The method of claim 14, wherein each one of the reads from the second sequencing method is tagged as being from a second haplotype if the read has a number of SNVs from the second haplotype mapped to it that is greater than the number of SNVs from the first haplotype mapped to it.

16. The method of any one of claims 1 to 15, wherein the chromosome length haplotypes comprise maternal and paternal haplotypes.

17. 1. A method for assigning parental origin to a haplotype, subhaplotype, or an allele associated with said haplotype, said method comprising: generating haplotypes of the genome; determining the methylation variable status of at least one imprinted methylation variable region (iDMR) associated with each one of said haplotypes; and correlating the determined methylation variable status of at least one iDMR with each one of said haplotypes to assign a parental origin for each one of said haplotypes.

18. 18. The method of claim 17, wherein the haplotype is a partial haplotype or a chromosome-length haplotype.

19. A method for assigning parental origin to a haplotype, subhaplotype, or allele associated with said haplotype, using only a sample obtained from a subject, by assessing the methylation status of an imprinted methylated variable region (iDMR) associated with each one of the subject's haplotype pairs.

20. selecting undesirable subhaplotypes or alleles from one of said haplotypes; determining a parent associated with the undesirable subhaplotype or allele based on the determination of the uniparental origin of the haplotype associated with the undesirable subhaplotype or allele; The method of any one of claims 1 to 19, further comprising performing cascade genetic testing on parental family members associated with said undesirable sub-haplotypes or alleles.

21. selecting a subhaplotype or allele of interest from one of said haplotypes; determining a parent associated with the subhaplotype or allele of interest based on the determination of the uniparental origin of the haplotype associated with the subhaplotype or allele of interest; The method of any one of claims 1 to 19, further comprising performing cascade genetic testing on parental family members associated with the subhaplotype or allele of interest.

22. The undesirable alleles are SDHD, SDHAF2, MAX, BRCA1, BRCA2, MLH1, MSH2, MSH6, PMS2, EPCAM, ATM, PALB2, TP53, APC, ACTA2, ACTC1, ACVR1L, AIP, ALK, ANKRD26, APOB, ARMC5, ATP7B, ATR, Axin2, BAG3, BAP1, BARD1, BLM, BMPR1A, BRIP1, BTD, BUB1B, CACNA1S, CASQ2, CASR, CDC73, CDH1, CDK4, CDKN1B, CDKN1C, CDKN2A, CEBPA, CFTR, CHEK1, CHEK2, COL3A1, CPA1, CTC1, CTNN1A, CTRC, CYLD, DDB2, DDX41, DES, DICER1, DIS3L2, DKC1, DLST, DROSHA, DSC2, DSG2, DSP, EGFR, EGLN1, ENG, ERCC1, ERCC2, ERCC3, ERCC4, ERCC5, ETV6, EXT1, EXT2, EZH2, FAM175A, FAN1, FANCA, FANCB, FANCC, FANCD2, FANCE, FANCF, FANCG, FANCI, FANCL, FANCM, FBN1, FH, FLCN, FLNC, FOCAD, GAA, GALNT12, GATA2, GLA, GPC3, GREM1, HFE, HNF1A, HOXB13, HRAS, KCNH2, KCNQ1, KIF1B, KIT, LDLR, LMNA, LZTR1, MC1R, MEN1, MET, MITF, MLH3, MRE11, MRE11A, MSH3, MUTYH, MYBPC3, MYH11, MYH7, MYL2, MYL3, NBN, NF1, NF2, NHP2, NOP10, NTHL1, OTC, PAX5, PALLD, PCSK9, PDGFRA, PHOX2B, PIK3CA, PKP2, POLD1, POLE, POLH, POT1, PRKAG2, PRKAR1A, PRSS1, PTCH1, PTCH2, PTEN, RAB43, RABL3, RAD1, RAD50, RAD51C, RAD51D, RB1, RBM20, RECQL, RECQL4, RECQL5, REST, RET, RINT1, RPE65, RPS20, RUNX1, RYR1, RYR2, SAMD9, SAMD9L, SCN5A, SDHA, SDHB, SDHC, SLC45A2, SLX4, SMAD321. The method of claim 20, wherein the undesired allele is an allele of SMAD4, SMARCA4, SMARCB1, SMARCE1, SPINK1, SRP72, STK11, SUFU, TERC, TERT, TGFBR1, TGFBR2, TINF2, TMEM127, TMEM43, TNNC1, TNNI3, TNNT2, TP53I3, TPM1, TRDN, TRIP13, TSC1, TSC2, TTN, TTR, TYR, VHL, WRAP53, WRN, WT1, XPA, XPC, or XRCC2, and optionally the undesired allele is BRCA1 187delAG, BRCA1 5385insC, or BRCA2 6174delT. ,

23. 20. The method of any one of claims 1 to 19, further comprising determining whether the two or more variants are cis-phased or trans-phased.

24. if the allele is a mutant form of succinate dehydrogenase complex subunit D (SDHD) and the parental origin is paternal, evaluating the subject periodically for paraganglioma or pheochromocytoma and / or performing cascade genetic testing, or if the allele is a mutant form of SDHD and the parental origin is maternal, performing cascade genetic testing; if the allele is a mutant form of succinate dehydrogenase complex assembly factor 2 (SDHAF2) and the parental origin is paternal, evaluating the subject periodically for paraganglioma and / or performing cascade genetic testing; or 20. The method of any one of claims 1 to 19, comprising periodically evaluating the subject for pheochromocytoma and / or performing cascade genetic testing if the allele is a mutant form of the MYC-associated factor X gene (MAX) and the parental origin is paternal.

25. 20. The method of any one of claims 1-19, comprising initiating pancreatic cancer screening in the subject beginning at age 50 years or 10 years earlier than the earliest pancreatic cancer diagnosis in the family, whichever is earlier, if the allele is a mutant form of ATM, BRCA1, BRCA2, MLH1, MSH2, MSH6, EPCAM, PMS2, PALB2, TP53, CHEK2, BRIP1, RAD51C, RAD51D, or MUTYH and the parental origin is determined to be a parent with a family history of pancreatic cancer.

26. 20. The method of any one of claims 1 to 19, further comprising, in a subject determined to have a particular pharmacogenomic outcome, predicting a similar pharmacogenomic outcome for a parent carrying the identified undesirable haplotype, subhaplotype or allele by identifying the parental origin of the undesirable haplotype, subhaplotype or allele.

27. 20. The method of any one of claims 1 to 19, further comprising, in a subject determined to have a particular pharmacogenomic outcome, predicting a similar pharmacogenomic outcome for a parent carrying the identified desirable haplotype, subhaplotype or allele by identifying the parental origin of the desirable haplotype, subhaplotype or allele.

28. determining whether the haplotype, subhaplotype, or allele is a disease-risk associated HLA haplotype, subhaplotype, or allele that has a known parent-of-origin effect; 20. The method of any one of claims 1 to 19, further comprising: if the haplotype, subhaplotype, or allele is determined to be a disease-risk associated HLA haplotype, subhaplotype, or allele with a known parental-of-origin effect, periodically monitoring the subject for risk of disease based on the known parental-of-origin effect.

29. 20. The method of any one of claims 1 to 19, wherein the haplotype, subhaplotype or allele is an HLA haplotype, subhaplotype or allele, and the method comprises selecting a cancer vaccine or cancer therapy for the parent based on a determination that the parent is the parental origin of the HLA haplotype, subhaplotype or allele in the subject.

30. 20. The method of any one of claims 1 to 19, further comprising comparing a DNA sequence obtained from a crime scene sample with at least a portion of the haplotype, subhaplotype or allele and parental origin information obtained from a subject, and if the DNA sequence obtained from the crime scene sample matches a DNA sequence of one of the haplotypes, subhaplotypes or alleles obtained from the subject, concluding that the identified uniparental origin of the subject is associated with the crime scene.

31. 20. The method of any one of claims 1 to 19, wherein the haplotype, subhaplotype or allele comprises an apparent de novo germline variant, and a biological parent of the subject tests negative for the germline variant, and the method further comprises determining a parental origin for the de novo germline variant to identify a biological parent at potential risk for recurrence of the germline variant, or determining a risk to a biological parent of having postzygotic somatic mosaicism indicating the biological parent is at risk of developing a disease.

32. 20. The method of any one of claims 1 to 19, further comprising identifying founder mutations by comparing haplotypes from two individuals carrying corresponding mutations to identify founder haplotypes.

33. 20. The method of any one of claims 1-19, further comprising administering a therapy to a parent determined to be of uniparental origin of an undesirable haplotype, subhaplotype or allele, said therapy optionally comprising an antisense oligonucleotide targeted to the undesirable allele, an allele-specific targeting construct, or allele-specific gene editing.

34. 34. The method of claim 33, wherein the undesirable allele comprises a mutation involved in a disorder with a dominant-negative effect, optionally a mutation involved in a disorder with a dominant-negative effect as associated with a single nucleotide variant (SNV), a copy number variant (CNV), an indel, a methylation, or a triple repeat.

35. 20. The method of any one of claims 1 to 19, further comprising detecting an abnormal methylation pattern in at least one of said haplotypes, optionally wherein said abnormal methylation pattern is detected in MLH1, MSH2, or BRCA2.

36. The method of any one of claims 1 to 19, further comprising performing a structural variant characterization of at least one of said haplotypes.

37. The method of any one of claims 1 to 19, wherein the subject is a mammal, optionally a human.

38. 1. An apparatus for determining a haplotype, a subhaplotype, or a uniparental origin of an allele associated with said haplotype, said apparatus comprising: a first sequencing device for performing a first sequencing technique to provide long range phase information; a second sequencing device for performing a second sequencing technique to provide a methylation variable status of at least one iDMR associated with said haplotype; and a processor for analyzing the long-range phase information and the methylation variable status of the at least one iDMR to determine the haplotype, subhaplotype, or uniparental origin of an allele associated with the haplotype.

39. 1. A kit for determining a haplotype, a subhaplotype, or a uniparental origin of an allele associated with said haplotype, said kit comprising: instructions for performing a first sequencing technique to provide long range phase information; instructions for performing a second sequencing technique to provide a methylation variable status of at least one iDMR associated with said haplotype; and instructions for combining the long-range phase information and the methylation variable status of the at least one iDMR to determine the haplotype, subhaplotype, or parental origin of an allele associated with the haplotype.

Images