Megabase-scale methylation phasing using nanopore long reads and NanoMethPhase

Akbari, Vahid; Garant, Jean-Michel; O’Neill, Kieran; Pandoh, Pawan; Moore, Richard A.; Marra, Marco A.; Hirst, Martin; Jones, Steven J.M.

doi:10.1186/s13059-021-02283-5

Cited by 55 publications

(86 citation statements)

References 54 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…5A). Similar predicted methylation patterns and performance of Nanopolish, Megalodon, and DeepSignal for NA12878 were observed by recent research [43]. In contrast, the DNA methylation-level histogram of Tombo and METEORE showed multiple peaks between 0 and 100% methylation levels, rather than two peaks.…”

Section: Per-site Performance Of 5mc Predictionsupporting

confidence: 84%

“…Therefore, the accuracy of methylation callers likely differs among the different types of genomic regions within which the CpGs are located. Recent benchmarking work on methylation calling tools for nanopore sequencing either compared only three such tools and considered very few genomic contexts [43], or restricted the comparisons to E. coli and 1743 CpGs of the human genome [38]. Hence, there is no published guideline and systematic comparison of all current DNA methylation-calling tools for nanopore sequencing using natural human DNA [44], especially at the whole-epigenome scale.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

DNA methylation-calling tools for Oxford Nanopore sequencing: a survey and human epigenome-wide evaluation

et al. 2021

View full text Add to dashboard Cite

Background Nanopore long-read sequencing technology greatly expands the capacity of long-range, single-molecule DNA-modification detection. A growing number of analytical tools have been developed to detect DNA methylation from nanopore sequencing reads. Here, we assess the performance of different methylation-calling tools to provide a systematic evaluation to guide researchers performing human epigenome-wide studies. Results We compare seven analytic tools for detecting DNA methylation from nanopore long-read sequencing data generated from human natural DNA at a whole-genome scale. We evaluate the per-read and per-site performance of CpG methylation prediction across different genomic contexts, CpG site coverage, and computational resources consumed by each tool. The seven tools exhibit different performances across the evaluation criteria. We show that the methylation prediction at regions with discordant DNA methylation patterns, intergenic regions, low CG density regions, and repetitive regions show room for improvement across all tools. Furthermore, we demonstrate that 5hmC levels at least partly contribute to the discrepancy between bisulfite and nanopore sequencing. Lastly, we provide an online DNA methylation database (https://nanome.jax.org) to display the DNA methylation levels detected by nanopore sequencing and bisulfite sequencing data across different genomic contexts. Conclusions Our study is the first systematic benchmark of computational methods for detection of mammalian whole-genome DNA modifications in nanopore sequencing. We provide a broad foundation for cross-platform standardization and an evaluation of analytical tools designed for genome-scale modified base detection using nanopore sequencing.

show abstract

Section: Per-site Performance Of 5mc Predictionsupporting

confidence: 84%

Section: Introductionmentioning

confidence: 99%

DNA methylation-calling tools for Oxford Nanopore sequencing: a survey and human epigenome-wide evaluation

et al. 2021

View full text Add to dashboard Cite

show abstract

“…One of the significant features of our single-molecule epigenetic data is our ability to investigate patterns of epigenetics; in the past, we and others have used this feature to phase contiguous blocks of the human epigenome (17,53). We have taken this a step further, using methylation alone to cluster reads in repetitive areas devoid of heterozygous polymorphisms, including the DXZ4 array where the methylation signature is important to X chromosome inactivation (55,69).…”

Section: Discussionmentioning

confidence: 99%

“…Long-read sequencing also makes it possible to robustly assign reads to haplotypes, allowing us to generate a fully phased human epigenome. The longer the read, the greater the chance of encountering one or more heterozygous single-nucleotide polymorphisms (SNPs), which can be used to phase the reads into maternal or paternal origin (17,52,53). Because HG002 is a diploid cell line, we could evaluate allele-specific epigenetic states of HG002 across the T2T-CHM13 reference using variant data on HG002 and both parents (54).…”

Section: Single Molecule Epigenetics Reveals Haplotype Specific Regulationmentioning

confidence: 99%

Epigenetic Patterns in a Complete Human Genome

Gershman

Sauria

et al. 2021

Preprint

View full text Add to dashboard Cite

The completion of the first telomere-to-telomere human genome, T2T-CHM13, enables exploration of the full epigenome, removing limitations previously imposed by the missing reference sequence. Existing epigenetic studies omit unassembled and unmappable genomic regions (e.g. centromeres, pericentromeres, acrocentric chromosome arms, subtelomeres, segmental duplications, tandem repeats). Leveraging the new assembly, we were able to measure enrichment of epigenetic marks with short reads using k-mer assisted mapping methods. This granted array-level enrichment information to characterize the epigenetic regulation of these satellite repeats. Using nanopore sequencing data, we generated base level maps of the most complete human methylome ever produced. We examined methylation patterns in satellite DNA and revealed organized patterns of methylation along individual molecules. When exploring the centromeric epigenome, we discovered a distinctive dip in centromere methylation consistent with active sites of kinetochore assembly. Through long-read chromatin accessibility measurements (nanoNOMe) paired to CUT&RUN data, we found the hypomethylated region was extremely inaccessible and paired to CENP-A/B binding. With long-reads we interrogated allele-specific, long-range epigenetic patterns in complex macro-satellite arrays such as those involved in X chromosome inactivation. Using the single molecule measurements we can clustered reads based on methylation status alone distinguishing epigenetically heterogeneous and homogeneous areas. The analysis provides a framework to investigate the most elusive regions of the human genome, applying both long and short-read technology to grant new insights into epigenetic regulation.

show abstract

“…1) For the Nanopore sequencer, several tools were developed for methylation calling [67] , [68] , [69] , and several benchmarking results have been published [70] , [71] , [72] . For example, Nanopolish, which is pioneering and one of the most applicable to cancer samples, was developed by using a hidden Markov model to detect 5 mCs in CpG sites according to the differences in signals [67] .…”

Section: Dna Methylation and Svsmentioning

confidence: 99%