Annual Review of Genomics and Human Genetics
Long-Read DNA Sequencing:
Recent Advances and
Remaining Challenges
Peter E. Warburton1,2
and Robert P. Sebra1,2,3,4
1
Department of Genetics and Genomic Sciences, Icahn School of Medicine at Mount Sinai,
New York, NY, USA; email: peter.warburton@mssm.edu, robert.sebra@mssm.edu
2
Center for Advanced Genomics Technology, Icahn School of Medicine at Mount Sinai,
New York, NY, USA
3
Black Family Stem Cell Institute, Icahn School of Medicine at Mount Sinai, New York, NY,
USA
4
Icahn Genomics Institute, Icahn School of Medicine at Mount Sinai, New York, NY, USA
Annu. Rev. Genom. Hum. Genet. 2023. 24:109–32
First published as a Review in Advance on
April 19, 2023
The Annual Review of Genomics and Human Genetics
is online at genom.annualreviews.org
https://doi.org/10.1146/annurev-genom-101722-
103045
Copyright © 2023 by the author(s). This work is
licensed under a Creative Commons Attribution 4.0
International License, which permits unrestricted
use, distribution, and reproduction in any medium,
provided the original author and source are credited.
See credit lines of images or other third-party
material in this article for license information.
Keywords
long-read sequencing, structural variants, repetitive DNA, pathogenic
mutations, epigenetic modifications
Abstract
DNA sequencing has revolutionized medicine over recent decades. However,
analysis of large structural variation and repetitive DNA, a hallmark
of human genomes, has been limited by short-read technology, with read
lengths of 100–300 bp. Long-read sequencing (LRS) permits routine sequencing
of human DNA fragments tens to hundreds of kilobase pairs in
size, using both real-time sequencing by synthesis and nanopore-based direct
electronic sequencing. LRS permits analysis of large structural variation
and haplotypic phasing in human genomes and has enabled the discovery
and characterization of rare pathogenic structural variants and repeat expansions.
It has also recently enabled the assembly of a complete, gapless human
genome that includes previously intractable regions, such as highly repetitive
centromeres and homologous acrocentric short arms. With the addition
of protocols for targeted enrichment, direct epigenetic DNA modification
detection, and long-range chromatin profiling, LRS promises to launch a
new era of understanding of genetic diversity and pathogenic mutations in
human populations.
109
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
OVERVIEW
DNA sequencing and the identification of disease-causing mutations and human genetic variation
has revolutionized medicine over recent decades. Since the first report of a sequenced human
genome in 2001 (52), the human reference genome has been the basis for functional annotation
and biomedical interpretation, providing an invaluable resource for basic science and clinical genetic
diagnostics today. The current human reference genome, hg38, is highly fragmented and
contains 349 discontinuous gaps and unresolved portions, and although incremental improvements
have been made by way of closing and spanning these gaps, adding alternate genomic loci
or patches, and improving the annotations (99), it continues to have at least ∼151 Mbp of unknown
sequence or gaps, including pericentromeric and subtelomeric regions, acrocentric short
arms, and large repeat arrays (81).
Over time, functional annotation and biomedical gene discovery have relied primarily on
comparison with the current reference genome build. The widespread use and relative affordability
of this class of next-generation sequencing technologies, which are dominated by short-read
sequencing-by-synthesis methods, have led to a large number of scientific discoveries in genetics,
evolution, and disease in humans. While extremely useful, the read length of short-read
sequencing (SRS), on the order of 100–300 bp per read, can result in the loss of resolution of
genomic regions that are not uniquely spanned by overlapping reads of this size. This limitation
is particularly true for low-complexity repetitive loci, duplicated regions, tandem arrays, and
complex structural variants, which collectively make up the majority of the gaps and missing
sequence. Long-read sequencing (LRS) can sequence through DNA fragments that are orders
of magnitude longer than the fragments sequenced using SRS (Figure 1a), with highly accurate
routine sequencing of 10–20-kbp fragments and lower-accuracy reads up to hundreds of kilobase
pairs using single-molecule, real-time (SMRT) sequencing by synthesis. Even longer reads of
up to 1–2 Mbp are available using nanopore-based electronic direct sequencing, although these
longer reads typically have lower accuracy.
LRS has emerged as critical in overcoming the limitations imposed by SRS technology and
has revealed a wide variety of previously underannotated genomic characteristics, such as repeat
arrays and structural variants. Long-read technology will continue to facilitate resolution of genomic
regions previously considered intractable until we collectively reveal the full spectrum of
human genetic variation (58). In this review, we concentrate on LRS technology development and
applications and reference a spectrum of studies that have applied LRS data to human genomics
and disease, although LRS has also been widely applied to bacterial genomes and other systems,
including plants, invertebrates, and other vertebrates (6). Collectively, LRS has provided major advancements
in human genomics, from the identification of rare disease-causing structural variants
such as insertions, deletions, and inversions to recently using a combination of LRS technologies
to provide the first examples of gapless telomere-to-telomere complete human chromosomes (59,
67) and a complete haploid human genome (81).
LONG-READ SEQUENCING TECHNOLOGY
There are currently two major long-read technologies: SMRT sequencing platforms (Figure 1b)
and nanopore-based sequencing platforms (Figure 1c). SMRT sequencing (developed primarily
by PacBio) is a single-molecule sequencing-by-synthesis technology. High-molecular-weight genomic,
double-stranded DNA is size selected and constructed into SMRTbell template libraries
with single-stranded closed hairpin adapters ligated to the ends. Primers annealed to the hairpin
adapters permit directed DNA polymerization around the single-stranded, topologically circular
SMRTbell. Individual SMRTbells are distributed into an array of up to 8 million zero-mode
110 Warburton • Sebra
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
waveguides (ZMWs) that each contain an immobilized DNA polymerase for real-time sequencing
within a restricted observation volume for improved signal-to-noise detection of each individual
real-time, fluorescently labeled nucleotide incorporation event. These circular SMRTbell templates
enable several passes on the immobilized DNA polymerase, giving continuous long reads
(CLRs) that are used to generate subreads that represent multiple reads of the genomic DNA
molecule in both directions. Earlier versions of SMRT CLR sequencing used very large insert
sizes of up to hundreds of kilobase pairs, and thus the circular template was sequenced only a few
times (on the order of 1–5 times, depending on the insert size and collection time), which resulted
in a higher error rate and sequencing accuracy of 85–92% and limited the early utility of this
technology.
cb
a
Polymerase
Up to 100-kb insert Up to 2-Mb fragment
Circular consensus read
Membrane
Nanopore
Electriccurrent
Time (s)
Base-calling algorithm
Meansignal(pA)
Excitation Emission
Base pairs (log scale)
Alu LINE
GeneGene
Repeat expansionRepeat expansion
NanoporeNanopore Ultralong readUltralong read
Composite repeat array (Hoyt et al. 2022)Composite repeat array (Hoyt et al. 2022)
Active alpha satellite array (Altemose et al. 2022)Active alpha satellite array (Altemose et al. 2022)
8q21VNTR array (Logsdon et al. 2021)8q21VNTR array (Logsdon et al. 2021)
Short readShort read
SMRT HiFiSMRT HiFi SMRT CLRSMRT CLR
rDNA array (Nurk et al. 2022)rDNA array (Nurk et al. 2022)
1 kb 10 kb 1 Mb 10 Mb100 kb10 bp
SMRTbell
template
Motor
protein
Sequencing
error
100 bp
(Caption appears on following page)
www.annualreviews.org • Long-Read DNA Sequencing 111
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
Figure 1 (Figure appears on preceding page)
LRS technologies and relative scales. (a) Relative read lengths across sequencing technologies and size
ranges of relevant variant structural elements, shown using a log scale. VNTR arrays are from 8q21.2 and
show the range of array sizes in the population. (b) SMRT HiFi sequencing. This technique uses dumbbellshaped
templates called SMRTbells, which are sequenced as single molecules within an array of ZMWs that
each contain a DNA polymerase. An observation volume restricted to the bottom surface of the ZMW
improves signal-to-noise detection of each nucleotide incorporation event using fluorescent signals specific
for each nucleotide. The circular templates are sequenced multiple times, and a circular consensus sequence
is obtained that is used for intramolecular error correction of the random polymerase errors while
processing each single pass of the SMRTbell library molecules. (c) Nanopore LRS, using Oxford Nanopore
Technologies as an example. This technique uses a linear DNA template loaded onto a motor protein and a
molecular nanopore. As the DNA unwinds and is driven through the nanopore, characteristic changes in
electric current allow the sequence of the resident bases to be determined using a trained base-calling
algorithm. Abbreviations: CLR, continuous long read; HiFi, high fidelity; LINE, long interspersed nuclear
element; LRS, long-read sequencing; SMRT, single molecule, real time; VNTR, variable number of tandem
repeats; ZMW, zero-mode waveguide. Panels b and c adapted from illustrations by Jill K. Gregory with
permission from Mount Sinai Health System; copyright 2022 Mount Sinai Health System.
A major improvement to SMRT sequencing came with the introduction of high-fidelity (HiFi)
sequencing, which represents the first LRS technology to provide sequence lengths on the order
of 10–20 kbp with an accuracy approaching 99.99% (Q40). Because the genomic DNA fragments
are selected to be between 15 and 20 kbp, the SMRTbell can progress through the DNA polymerase
multiple times (on the order of 7–12 times or more), providing multiple subreads that
are computationally combined via the circular consensus sequencing algorithm to provide highly
accurate sequencing of each insert (129) (Figure 1b). HiFi reads have provided improved structural
variant discovery and have deciphered some of the underannotated repetitive regions of the
genome, such as those found at centromeres and acrocentric short arms (59, 81).
The other major LRS technology is nanopore-based LRS (Figure 1c). A linear doublestranded
DNA template, which can be as long as several megabase pairs, is ligated to an adapter
that is preloaded with a motor protein and then loaded onto an array of nanopores. The motor
protein unwinds the DNA, and a single strand of the molecule is driven through the nanopore
by an electric current. The electric current is characteristically disrupted by combinations of
3–7 bp of DNA sequence as the templates pass through individual nanopores,producing a squiggle
(Figure 1c) that is translated into DNA sequence based on computational base-calling algorithms
(91). Standard nanopore (Oxford Nanopore Technologies) sequencing can sequence molecules
that are an order of magnitude longer than those sequenced by SMRT technology, routinely
tens to hundreds of kilobase pairs with an accuracy of approximately 87–98%, although improvements
in base-calling algorithms and chemistry are increasing this accuracy (91, 124), with current
advances achieving 99% or higher accuracy. Nanopore sequencing can also be used to generate
ultralong reads, which are generally much larger than 100 kbp and can occasionally be as long
as several megabase pairs (46) and are especially useful as genome scaffolds in the assembly of
large, difficult-to-sequence regions. Ultralong reads have permitted tremendous access to long
repetitive regions, as demonstrated for the Y chromosome centromeric region (47).
Another application is to use nanopores for the electrophoretic loading of ZMW arrays, which
greatly increase loading efficiency over the diffusion loading currently used in SMRT sequencing.
Thus, hybrid ZMW–nanopore LRS can reduce the DNA input required by SMRT sequencing by
orders of magnitude, alongside the reduction of any bias toward loading of smaller molecules into
ZMWs (29,53).An interesting combination of SMRT and nanopore-based technologies is LRS by
synthesis using a nanopore-coupled DNA polymerase with distinct polymer-tagged nucleotides,
which has allowed real-time electronic DNA sequencing without the need for complex optics and
112 Warburton • Sebra
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
more straightforward electronic base calling (31, 110). These well-established and emerging LRS
platforms exemplify the continuous improvements available for genomics research that provide
the opportunity to investigate genome organization and variation at an unprecedented scale.
IMPROVING HUMAN GENOME ASSEMBLIES USING LONG-READ
SEQUENCING TECHNOLOGIES
The most widely used reference genome (currently hg38) includes data constructed by sequencing
a collection of large-insert bacterial artificial chromosome (BAC) clones from several individuals;
thus, at any individual genomic location it represents a single human haplotype (58), although
an additional 434 patches and alternative haplotypes are available in some variable regions of
the genome (21, 78). An important consideration in these reference builds is that BAC cloning
leads to an underrepresentation of repetitive sequences and the use of multiple BAC libraries
from different individuals can result in mosaicism of haplotypes. While useful, this strategy results
in incompatible structural polymorphisms or large repetitive regions flanking many of the
unresolved sequence gaps (27).
LRS has been instrumental in improving genome resolution and understanding the full
spectrum of genetic variation.Long-read SMRT sequencing has been used along with sequencingby-synthesis
short-read data, genome-mapping scaffolds, and reference-based approaches to
improve assembly and understanding of the architecture of a diploid human genome, revealing
large structural differences from the reference genome, including 68 Mbp of novel content relative
to the NA12878 genome (87). In addition, LRS of trios has been used to analyze structural
variation in diploid genomes in a haplotype-resolved manner, resulting in a three-to-sevenfold
increase in structural variant detection over most other high-throughput sequencing studies (16,
26, 125). The Genome in a Bottle Consortium has used multiple platforms, including LRS, to
establish structural variant benchmarks in multiple reference genomes from diverse populations
in order to improve the reliability of variant calls across platforms to provide more robust references
for a multitude of haplotypes (125, 136, 137). The major structural variant alleles from 15
human genomes have also been characterized, which has allowed the discovery of many fixed and
common structural alleles that were missing from the human reference genome (7). Such work
has contributed to the Database of Genomic Variants, which provides an important reference for
the position, size, and population frequency of more common structural variations (60).
Long-read SMRT sequencing has also been applied to sequence haploid genomes, revealing a
large number of genetic variants and demonstrating the advantage of providing baseline haploid
genomes derived from hydatidiform mole cell lines, which retain only a single set of homologous
chromosomes due to either fertilization of an enucleated egg or the subsequent loss of the maternal
complement postfertilization (45). These hydatidiform mole references therefore represent a
functionally haploid equivalent of the human genome lacking allelic variation, in order to remove
the complexity of diploidy from sequencing and assembly of genomes (15, 42, 120).
To illustrate the range and depth of variants seen using current LRS data, we used the PacBio
structural variant (pbsv) caller to call variants on a publicly available 30-fold coverage NA24385
human HiFi dataset (Figure 2). The tracks on the illustrated Circos plots in Figure 2a,b are based
on the annotations called by pbsv. Tandem variants (Figure 2a,b) are those involving simple repeats,
such as those overlapping with Tandem Repeats Finder (12) and/or simple satellite repeats
in RepeatMasker (104) (Figure 2e). Other annotations easily observed are the insertion or deletion
of transposable elements, such as long interspersed nuclear elements (LINEs) (Figure 2d),
which are predominantly around 6 kbp in size (Figure 2a,b), and short interspersed nuclear elements
(SINEs) (Figure 2c), which are predominantly around 350 bp in size (Figure 2a,b). Precise
www.annualreviews.org • Long-Read DNA Sequencing 113
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
1
2
3
4
5
6
7
8910
11
12
13
14 15 16 17
18
19
20
21
22
X
Y
3,697 deletions
86 inversions
2,466 insertions
524 duplications
579 deletions 824 insertions
61 deletions 58 insertions
a Deletions
c ALU deletion
e Tandem repeat f Not annotated
d LINE deletion
b InsertionsTandem repeats
Unannotated variants
Inversions/duplications
Alu variants
LINE variants
1
2
3
4
5
6
7
89
10
11
12
13
14 15 16 17
18
19
20
21
22
X
Y
158,399,500
pbsv
Repeating elements by RepeatMasker
GENCODE v41
Database of Genomic Variants
343-bp deletion
RSRC1
essv6942554
essv6906790
essv6878123
essv6889571
essv6700341
essv6673743
Chr1: 218,010,000 218,015,000
pbsv
Repeating elements by RepeatMasker
Database of Genomic Variants
6,106-bp deletion
essv6537076
essv4521431
essv3564206
essv9745569
nssv4002947
essv6880551
essv6807519
essv6832740
Chr5: 178,395,000 pbsv
Repeating elements by RepeatMasker
Simple tandem repeats by
Tandem Repeats Finder
Database of Genomic Variants
GENCODE v41
1,672-bp deletion
GCATTCGGCTCTC...
COL23A1
essv9765830
essv7558951
essv6762533
esv2731230
essv6754423
esv2731232
essv6974575
Chr7:148,375,000 148,380,000
pbsv
Repeating elements by RepeatMasker
GENCODE v41
Database of Genomic Variants
3,460-bp deletion
CNTNAP2
esv3542618
nsv1074697
dgv3690n106
esv2735322
500bp
500bp
500bp
500bp
500bp
500bp
5kb5kb
5kb5kb
5kb5kb10kb10kb
5kb5kb
5kb5kb
500bp
500bp
6,825 deletions 9,158 insertions
Scale 200 bp
Chr3:158,399,000
SINE
LINE
LTR
DNA
Simple repeat
AluYa5
L1Hs
SINE
LINE
LTR
DNA
Scale 5 kb
Scale 2 kb Scale 2 kb
SINE
LINE
LTR
DNA
SINE
LINE
LTR
DNA
Simple repeat
500bp
500bp
500bp
500bp
500bp
500bp
5kb5kb
5kb5kb
5kb5kb
10kb10kb
5kb5kb
5kb5kb
500bp
500bp
(Caption appears on following page)
114 Warburton • Sebra
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
Figure 2 (Figure appears on preceding page)
Example variation detected using LRS. (a,b) Circos plots of deletions (panel a) and insertions (panel b) relative to the hg38 reference,
showing variants and annotations ranging from 50 bp to 15 kbp as called by the pbsv tool using SMRT HiFi sequence data (85). Variant
coordinates are shown relative to chromosomal location (outside track), and variant sizes are shown on a log2 scale within each Circos
track. From the outside to the inside, the colored tracks show variants involving tandem repeats (blue), unannotated variants (orange),
inversions (pink in panel a), duplications (pink in panel b), Alu insertions or deletions (green), and LINE insertions and deletions (red).
Counts for each type of variant are shown between the plots. (c–f ) Examples of deletions relative to the hg38 reference in the UCSC
Genome Browser. For each variant, tracks shown from top to bottom include the pbsv variant caller track, RepeatMasker track,
GENCODE v41 track (when applicable), simple repeat track (when applicable), and partial data from the Database of Genomic
Variants. Panel c shows a 343-bp deletion of a full-length AluYa5 element (green) in intron 6 of the RSRC1 gene; panel d shows a
6,106-bp deletion of a full-length L1Hs element (red); panel e shows a 1,672-bp deletion of 19 repeats from a 2,026-bp array of 23
tandem 88-bp G-rich repeats (blue); and panel f shows a 3,460-bp deletion (orange), not further annotated by pbsv, in intron 21 of the
CNTNAP2 gene. Abbreviations: HiFi, high fidelity; LINE, long interspersed nuclear element; LRS, long-read sequencing; LTR, long
terminal repeat; pbsv, PacBio structural variant; SINE, short interspersed nuclear element; SMRT, single molecule, real time; UCSC,
University of California, Santa Cruz.
insertions or deletions of full-length LINEs and SINEs with little or no surrounding nontransposon
sequence illustrate that these are likely to represent transposition events. A precise transposon
deletion observed in a demonstrative genome dataset represents a transposition event leading to
an insertion in the reference genome relative to the sequenced genome. This is supported by the
likelihood that most of these precise insertions and deletions are relatively young elements, such
as AluY and L1Hs or L1PA2.
Additional types of variants shown are inversions (Figure 2a) and duplications (Figure 2b),
both of which are often associated with known segmental duplications (8). Furthermore, there
are many insertions and deletions that are not annotated by pbsv (Figure 2a,b,f ), suggesting that
they are not associated with any of the above sequence characteristics. We should point out that
most of the variants detected by long-read HiFi whole-genome sequencing (WGS) demonstrative
data are common and have been compiled in previous databases, such as the Database of
Genomic Variants (60) (see Figure 2c–f ). However, this database was originally curated using
specialized, population-based, and often high-effort applications, including copy-number-variant
detection using comparative genomic hybridization or single-nucleotide polymorphism arrays
(19), size-selected and fosmid-based end-to-end sequencing (50, 51), and short-read sequencingby-synthesis
techniques such as read depth analysis, read pair analysis based on abnormally
mapping read end pairs, or sequence assembly with unmapped read pairs (1, 73). Recently, highercoverage
whole-genome SRS of 602 trios coupled with improved variant-calling algorithms has
expanded the discovery of insertions/deletions and structural variants in human populations (14).
However, the advent of LRS has greatly simplified the analysis, reliability, and annotation of these
structural variants and permitted their identification in individual genomes, especially in regions
with repetitive DNA or segmental duplications (135), expanding the combined knowledge base
while also informing any individual genome. Furthermore, rare structural variants that may be
protective or pathogenic are also amenable to discovery in individual genomes because of LRS
technologies that enable diagnostic and risk association legacies previously not achievable with
short-read WGS data.
Additional high-resolution technologies have been developed to resolve structural variants.
These include long-fragment-read (88) and linked-read sequencing technologies, both of which
use SRS to sequence the ends of longer genomic DNA molecules, enabling long-range sequence
and haplotype phasing over large segments. With Hi-C technology, protein/DNA complexes are
cross-linked in vitro, where DNA within complexes is digested and ligated, which links sequences
that are in close proximity in the nucleus. Since the contact frequency is related to distance, usually
on the same chromosome, the sequence of the ligated fragments provides long-distance and
www.annualreviews.org • Long-Read DNA Sequencing 115
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
haplotypic data that can be used to link contigs and provide long-range sequence information
(56, 83). Computational analysis methods have also been developed to use Hi-C data for
structural variant calling (22, 121, 123). Optical genome mapping is another technique where
high-molecular-weight genomic DNA is captured in a microfluidic chamber and subject to
sequence-specific nicking and subsequent fluorescent labeling, providing long-range readout of
large genome fragments with preserved genomic order that can permit resolution of up to several
megabase pairs when compared with a reference genome map. Optical genome mapping has been
used to anchor scaffolds for LRS or SRS WGS assemblies, identify large structural variations,
and even detect chromosomal aberrations (39, 131, 134). While neither Hi-C nor optical genome
mapping generates single-base-resolved long-read DNA sequence, their use provides scaffolding
and long-range sequence information that is critical for accurate genome assembly (55, 81).
TARGETED LONG-READ SEQUENCING APPROACHES
AND APPLICATIONS
While whole-genome LRS can be used to detect structural variations and complex sequence regions
that are not detectable by short-read technologies, comprehensive LRS of whole human
genomes remains prohibitively expensive, especially in a clinical setting. One approach to enhance
the utility of LRS is targeted enrichment of specific genomic loci. Targeted approaches permit
multiplexing of specific clinically and/or scientifically interesting regions from batches of samples
to increase the efficiency of LRS, with the goal of obtaining higher sequence coverage and/or
enabling localized assembly or phasing of relevant genomic variants that would otherwise be prohibitively
expensive using whole-genome LRS methods. One approach makes use of standard or
long-range PCR amplification of the targeted region to generate SMRTbell libraries from the resulting
amplicons for higher-throughput LRS. LRS amplicon sequencing has proven valuable for
phasing double mutations in cis in the PIK3CA gene as a regulatory target in breast cancer (118),
performing fully phased haplotyping of the NUDT15 gene (100), investigating CYP2D6 pharmacogenetic
genotypes (89), sequencing the MUC1 variable number of tandem repeats (VNTR)
region in autosomal dominant tubulointerstitial kidney disease (130), and analyzing a 72.8-kbp
deletion encompassing BBS9 (implicated in Bardet–Biedl syndrome) and RP9 (implicated in retinitis
pigmentosa) (93). An additional approach that can target larger regions of the genome uses
biotinylated oligonucleotides that are homologous to the region of interest, which can be designed
to tile across up to 1 Mbp of DNA or from multiple regions of the genome. The genome is
fragmented and ligated to barcoded adapters, PCR preamplified, and annealed to the hybridization
probes; the target DNA is purified using streptavidin beads and then PCR amplified again
to enrich the yield; and SMRTbell libraries are then sequenced (108). This targeted enrichment
approach has been used to investigate hepatitis B virus integrations in hepatitis B patients (116).
However,protocols that include PCR amplification steps pose some technical limitations,especially
when amplifying arrays of short tandem GC-rich repeats. Limitations to PCR amplification
also include target duplication, size limitations, template switching, and/or loss of epigenetic modifications.
Therefore, the Cas9-assisted targeting of chromosome segments (CATCH) approach
provides an amplification-free way to target genomic DNA and capture regions spanning hundreds
of kilobase pairs (32). CATCH uses Cas9-targeted fragmentation of high-molecular-weight
DNA in vitro, followed by separation and purification of the targeted region from the remaining
genome using pulsed field gel electrophoresis, followed by LRS. CATCH allows several-hundredfold
enrichment of the targeted region over WGS and has been developed for both SMRT and
nanopore-based LRS. CATCH has been used to target the 80-kbp BRCA1 gene and surrounding
regulatory regions on 200-kbp fragments, which were enriched 237-fold and enabled 70-fold
116 Warburton • Sebra
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
sequencing coverage. Although the low accuracy of the LRS used at the time reduced the ability
to confidently call single-nucleotide polymorphisms compared with SRS, the ability to phase
haplotypes and examine structural variants underscores the value of the technique (32). CATCH
is particularly valuable for unstable short repeat copy expansions and has been used to sequence
across an entire expanded 7,941-bp CCCCGG array in the C9orf72 gene of a patient with amyotrophic
lateral sclerosis (25, 36). Myotonic dystrophy type 1 is caused by an unstable CTG
expansion up to 4,000 repeats in the DMPK locus, which was initially sequenced using PCR amplification
and SMRT sequencing (61), but CATCH was subsequently used to improve the accuracy
of the repeat expansion analysis (113). Moreover, CATCH and nanopore LRS have recently been
used to isolate extrachromosomal DNA from human-derived cancer lines that contained amplified
select enhancers or oncogene coding sequences (43).
An alternative approach called targeted LRS uses adaptive sampling by selectively sequencing
DNA molecules that reside in a predefined genomic region. After approximately 500 bp of sequence
has been called in real time and computationally aligned to a reference genome, if the
region is within the predefined region, then the sequencing continues; if this criterion is not
met, then the molecule is ejected from the pore by reversing the current, and a new molecule
is selected and sequenced. Targeted LRS adaptive sampling has been used to identify previously
undetected pathogenic variants in several diseases, including Werner syndrome and pseudohypoparathyroidism
(70–72). Nanopore Cas9-targeted sequencing has also been used, where Cas9
guide RNAs cleave a targeted region in the genome, the adapters are ligated, and the library is
subjected to nanopore LRS (37). The targeted LRS and nanopore Cas9-targeted sequencing approaches
can be combined for further improved enrichment, such as for CYP2D6–CYP2D7 hybrid
allele genotyping (95). Collectively, these enrichment protocols for both SMRT and nanopore
LRS allow specific enrichment of genomic regions of interest, which will greatly reduce the persample
cost and enhance the utility for LRS in local genotyping and phasing of specific single or
ensemble structural variants for genetic diagnostics in cases where such rare structural variants
are the cause of disease or require downstream validation.
EPIGENOMICS USING LONG-READ SEQUENCING APPROACHES
LRS also has advantages in detecting epigenetic modifications found on native genomic DNA,
including chemical nucleotide modifications of cytosine or adenine methylation (38). DNA methylation
is most often examined using bisulfite deamination of DNA, where unmethylated cytosines
are converted to uracil followed by sequencing and informatic comparison with untreated DNA
controls. SMRT bisulfite sequencing has directly sequenced the products of bisulfite deamination
up to 1.5 kbp in size using LRS without the need for a cloning step (132). In more recent
LRS applications, modified or methylated nucleotides are detected directly in the native genomic
DNA because the epigenetic chemical moieties characteristically modify the sequencing signal to
provide a significant detection mechanism with sufficient signal to noise against the sequencing
baseline. For example, in SMRT LRS, methylated nucleotides perturb the kinetics of base-pair
incorporation by the polymerase, resulting in an increased interpulse duration and pulse width
of the fluorescent signal between nucleotides as they are sequenced. The methylation status of
cytosine and adenine nucleotides are effectively read in real time as the polymerase synthesizes
from the SMRTbell template and encounters modified bases (28, 30). The surrounding sequence
context affects the changes in interpulse duration and pulse width and is most accurate when multiple
circular consensus sequencing passes are achieved, impacting the maximum size that can be
sequenced, depending on the epigenetic target (11). 5-Methyl-CpG (5mCpG) is the most common
methylation site in mammalian genomes and important in gene regulation and development
www.annualreviews.org • Long-Read DNA Sequencing 117
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
(105). 5mCpG detection using SMRT sequencing has been the most explored computationally
(80, 114). When electronic nanopore LRS is used, modified bases introduce deviations to the
readout squiggle that can be decoded using a pretrained algorithm and hidden Markov models.
For example, Nanopolish (90, 101) and additional algorithms have been trained to detect 5mCpG
and other modified bases (38).
Bisulfite deamination treatment often results in fragmentation of DNA and thus is less suitable
for LRS applications sequencing DNA fragments tens of kilobase pairs in size. Therefore,
LRS has been coupled with other chemical treatments that do not cause DNA fragmentation
and convert additional modifications that are difficult to detect with altered readout of native
DNA. TET enzyme oxidation has been used to convert both 5-methylcytosine (5mC) and
5-hydroxymethylytosine (5hmC) to dihydrouracil, which is recognized as thymine, thus permitting
sufficient detection signal for 5hmC (57). Another treatment, called long-read enzymatic
modification sequencing, uses selective enzymatic protection of both 5mC and 5hmC before
deamination with APOBEC3A (112). Importantly, both of these techniques allow PCR amplification
of the treated DNA with preservation of the deamination signal, reducing the required
amount of starting material for LRS, although off-target modifications can be possible.
Emerging LRS-based chromatin-profiling methods, which present the opportunity to map
chromatin organization across kilobase-length genomic regions and can include repetitive DNA
elements such as centromeres and telomeres (23), have also recently been used as an epigenomics
tool. Fiber-seq and single-molecule adenine methylated oligonucleosome sequencing
assay (SAMOSA) techniques use a nonspecific DNA N6-adenine methyltransferase to reveal the
chromatin architecture of kilobase-sized chromatin fibers on the underlying DNA by methylating
accessible adenine residues, which are then detected using the changes in interpulse duration
in SMRT LRS (3, 109). Fiber-seq has been used to map transcriptional regulatory elements and
telomeric chromatin (23, 109). A related nanopore-based technique called nucleosome occupancy
and methylome nanopore sequencing (nanoNOMe) uses a GpC (not to be confused with a CpG)
methyltransferase to label endogenous GpCs in chromatin fibers and then leverages the Nanopolish
base-calling algorithm with additional training to detect GpC methylation from nanopore
squiggles. NanoNOMe has the advantage of simultaneously measuring both chromatin accessibility
and endogenous 5mCpG; it has been used to examine CTCF binding sites (54) and, more
recently, the phasing of gene regulatory and imprinting elements on fragments as large as 116 kbp
(10). Using LRS to detect base modifications and chromatin structure provides the critical advantage
of allowing mapping and phasing across kilobase-sized genomic alleles, which will continue
to provide important functional insight into these epigenetic signatures on chromosomes.
PATHOGENIC VARIANT DISCOVERY USING LONG-READ
SEQUENCING TECHNOLOGIES
The types of structural variation that have been revealed by LRS are usually those not easily
resolvable by SRS, such as insertions, deletions, and inversions (135). Although many commonly
occurring variants are included in structural variant databases (60), disease-causing pathological
structural variants are by definition rare in populations and not easily detected by SRS. Many
cases of rare large insertions and deletions relative to the human reference genome have been
discovered using both whole-genome LRS and targeted capture methods. Studies have found
375- and 395-kbp deletions in EYS in patients with retinitis pigmentosum (97),a 72.8-kbp deletion
encompassing BBS9 and RP9 in a patient with Bardet–Biedl syndrome (93), a 12.4-kbp deletion
in CLN6 in a patient with progressive myoclonic epilepsy (76), a 7.1-kbp deletion in one allele of
G6PC (together with a c.326G>A mutation in the other allele) in a patient with glycogen storage
118 Warburton • Sebra
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
disease type Ia (66), a 2.814-kbp deletion overlapping exons in PRKAR1a in a patient with multiple
neoplasia and cardiac myxomata (65), and a 4.6-kbp insertion in a family with benign adult familial
myoclonus epilepsy (77). Insertions of previously unmapped sequence not present in the reference
genome are particularly difficult to detect using SRS data mapping but are often plainly obvious
in LRS data because individual and/or ensemble long reads often contain the entire novel inserted
sequence in its genomic context. Yet another application for LRS is in human leukocyte antigen
(HLA) typing that can span the entirety of the HLA class 1 genes and allow accurate haplotyping in
this hypermorphic gene complex (63), which has contributed greatly to the expansion and known
diversity of HLA database haplotypes (94).
Another structural variant subtype that is often revealed by LRS is that of repeat expansions,
which are otherwise challenging to accurately sequence and characterize using conventional SRS
data (74). LRS can often span the entire expansion, which can be on the order of many thousands
of base pairs (33). Either WGS or a targeted enrichment approach can produce reads from specific
regions known to contain repeats. Repeat expansions are a hallmark of many neurodegenerative
diseases, such as expansion of the GGGGCC hexameric repeat in the c9orf72 gene associated with
5–7% of amyotrophic lateral sclerosis cases (25, 36, 117). Repeat expansions are also common in
spinocerebellar ataxia (98), neuronal intranuclear inclusion disease (106), and other diseases (18,
44). Pentanucleotide TTTCA and TTTTA expansions in SAMD12 have been shown to cause
familial cortical myoclonic tremor with epilepsy (133), while ATTCT and ATTCC expansions in
ATXN10 affect disease penetrance in spinocerebellar ataxia type 10 (79).Other pathogenic variants
revealed by LRS include insertion of a cytosine into a VNTR region of 60-bp repeats in the MUC1
gene, which has been identified as a causal mutation of autosomal dominant tubulointerstitial kidney
disease (130). Duplications include a 27-bp duplication in the polyalanine tract of the HoxD13
gene that leads to synpolydactyly (64) and a heterozygous 3.463-kbp duplication, including exon
30 of the LAMA2 gene, that causes merosin-deficient muscular dystrophy (13). Breakpoints of
a balanced chromosomal Xp11.1/20p13 translocation were precisely identified to the base pair
using LRS in an individual with developmental disabilities (24). Novel inversions detected using
LRS technology include a 5.9-Mbp inversion that disrupted exons 3–79 in the DMD gene of a
previously undiagnosed male patient (and was also present as a heterozygote in the mother) (13)
and a 12-kbp inversion that disrupted the first two exons of BRPF1 and the last four exons of
CPNE9 on chromosome 3 in a patient with intellectual disability (75). These pathogenic structural
variants were found primarily in an unbiased application of LRS. Indeed, the power of LRS
to identify the causative genetic variant in rare undiagnosed diseases is rapidly emerging, including
recent examples of large-cohort studies being deeply sequenced with SMRT HiFi whole-genome
LRS (20, 62). Thus, the value of LRS for rare mutation detection in undiagnosed disease patients
cannot be overstated (62). As application of LRS leads to improved reference genomes via the
bridging of large gaps and unresolved genomic regions across the range of diverse human haplotypes,
the ability to ascertain medically pathogenic variation will likely continuously reveal novel
human disease and resilience association in future medicine.
LONG-READ SEQUENCING ENABLES THE FIRST
TELOMERE-TO-TELOMERE HUMAN REFERENCE GENOME
A recent advance that demonstrates the true power of LRS is the first complete telomere-totelomere
sequence of several complete chromosomes (59, 67), which was then followed by the
first complete human genome sequence (81). For ease of assembly, the Telomere-to-Telomere
(T2T) consortium performed LRS data analysis of the haploid CHM13 genome, which is derived
from a hydatidiform mole line and therefore represents a functionally haploid equivalent of
www.annualreviews.org • Long-Read DNA Sequencing 119
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
the human genome that lacks diploid allelic variation. The use of LRS of a haploid cell line eliminated
the limitations of BAC-based assembly used in previously generated reference genomes and
the complexity of structural polymorphisms within complete diploid genomes. However, previous
technologies used to resolve human references for these genomes (SMRT CLR sequencing
and nanopore-based ultralong-read sequencing) collectively had an error rate of greater than 5%,
which did not permit assembly of larger, highly homologous tandem arrays. The increased accuracy
provided by whole-genome SMRT HiFi circular consensus sequencing reads enabled the
sequencing of subtly different but highly homologous tandem arrays. Nanopore-based ultralong
reads are also used to span repeat arrays as contiguous scaffolds for anchoring the highly accurate
HiFi WGS reads in fully resolved reference generation studies. This strategy was used to generate
gapless sequence assemblies of chromosome 8 (59) and the X chromosome (67), including
complete centromeres and other tandem arrays of previously unspanned repeats.
The T2T CHM13 haploid assembly was ultimately completed with a hybrid of data from
various next-generation sequencing technologies, including 100-fold SRS data, 30-fold wholegenome
HiFi LRS data, and 120-fold Oxford Nanopore Technologies–based ultralong nanopore
reads, alongside optical genome mapping, Hi-C data, and single-strand sequencing (67, 96). This
effort required a significant amount of manual curation by a large team over many months, with
different groups focused on each chromosome (48). Thus, despite improvements, a multitude of
integrated datasets across various LRS and SRS technologies is currently required to complete
even a single haplotypic reference. This limitation in the genomics field will motivate the additional
developments that are needed to automate assembly and phasing of diploid human genomes
at high quality and at scale. Ultimately, future improvements will be critical for the adoption
of LRS data in regulated discovery and validation of clinically relevant variants and a translated
understanding of human genetic variation in everyday medical practice.
The members of the T2T consortium were collectively determined to create a complete gapless
human genome that contained the large highly repetitive regions of the human genome,
including large interspersed arrays of tandem repeats (41), all segmental duplications (119), and,
remarkably, the complete span of the megabase-sized centromeric and pericentromeric repeat
array in each chromosome (5), as well as the highly homologous short arms of acrocentric chromosomes
(35). This research effort added nearly 200 Mbp (∼8%) of previously hidden sequence
to the human genome (4), including genomic regions representing large gaps that many thought
intractable.
One of the most challenging types of DNA structure to sequence is satellite arrays, which
consist of portions of the genome up to several kilobase pairs in size that have been tandemly
duplicated into arrays as large as several megabase pairs (128). These arrays are inherently challenging
to fully sequence and are poorly represented in reference genomes due to the size and high
homology of the repeat units,their instability in BAC libraries from which reference genomes were
constructed, and their high variability in repeat copy number and array size between individuals.
There are many classes of satellite DNA in the human genome, including satellite, simple, and
low-complexity repeats, as well as composite repeats that are more complex. Satellite repeats are
generally short tandem copies of several base pairs; for example, classical human satellites 2 and
3 (HSat2 and HSat3), totaling 28.7 and 47.6 Mbp in CHM13, respectively, are derived from the
simple repeat (CATTC)n and constitute the largest contiguous satellite arrays found in the human
genome, including a 27.6-Mbp array on chromosome 9 in CHM13 (81). Ten such simple sequence
satellite repeat types, including five previously unknown types, were classified in CHM13. Additional
tandem DNA classes include ∼68-bp beta satellite DNA, found in pericentromeric regions
and on the acrocentric short arms, and 171-bp alpha satellite DNA, which is the predominant
centromeric class of repetitive DNA.
120 Warburton • Sebra
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
A third class of tandemly repeated DNA is composite elements, which by definition contain
at least three types of other repetitive elements, such as transposable elements, satellite repeats,
or composite subunits. The T2T consortium described the repeat unit and array sizes for 19 of
these composite elements, as well as their distinct epigenetic and transcriptional patterns, in the
CHM13 genome. Most of these events are found in a single array, and eight contain proteincoding
annotations (41). While the repeat unit of many of these composite satellites had been
described more than a decade ago (128), several were found on either side of polymorphic gaps
in the hg38 reference genome. With the application of recent LRS and assembly protocols developed
by the T2T consortium, the exact size, copy number, repeat organization, and position
of these arrays have been determined for CHM13 and other genomes. A polymorphic composite
repeat on chromosome 8q21, referred to as the VNTR repeat (41), is found in a large array
of ∼12-kbp repeats, each of which includes the REXO1L gene and several types of interspersed
repeats detected by RepeatMasker (128) (Figure 3). This repeat array can be visible cytogenetically
(107, 115). In the hg38 reference genome, several repeats are visible spanning a gap
(Figure 3a), while in CHM13, this array has been fully sequenced and shows a complex arrangement
of 73 repeats spanning almost 1 Mbp of DNA (59) (Figure 3b). Genomic features like
composite arrays likely represent the product of natural chromosomal evolution processes, such
as unequal crossing over. Nonetheless, their functional significance remains unclear, especially
given that the 8q21 VNTR composite repeat array is the site of ectopic neocentromere formation
(40). The ability to sequence across and precisely locate these arrays may provide the means for
their targeted deletion using CRISPR technology, to assess their effect on chromosome stability,
homologous pairing, and epigenetics.
Moreover, the short arms of human acrocentric chromosomes 13, 14, 15, 21, and 22 are conspicuously
missing from the recent human reference genome hg38.These arms contain the 45-kbp
tandemly repeated rDNA genes, and during interphase they cluster in the nucleolus, where the
rDNA genes are expressed. The large amount of repetitive and satellite DNA, along with the
high degree of homology between acrocentric short arms, made them extremely challenging to
sequence and assemble. With the exception of a small amount of pericentromeric p-arm material,
these regions were largely excluded from reference genomes. With the advent of the approaches
developed by the T2T consortium, these regions have now been sequenced for each acrocentric
chromosome, which in CHM13 range in size from ∼10 to 17 Mbp. Although variable in size, each
acrocentric short arm follows a similar pattern, with inverted segmental duplications and arrays
of acrocentric, HSat3, beta satellite, and HSat1 repeats surrounding the rDNA arrays, most of
which are also seen in other regions of the genome. The organizational similarity between acrocentric
satellite repeats is likely a result of the dynamic evolution of intra- and interchromosomal
exchanges of acrocentric short arms in the nucleus (81).
Satellite arrays, including active alpha satellite arrays, HSat3 on chromosome 9, and beta satellite
arrays on chromosome 1 and in the acrocentric regions, contain many inversions in CHM13
that were confirmed in other sample genomes (5) (Figure 3). Inversions can form extruded
cruciform structures that may mimic regions of paired homologs, and such DNA that is both
tandemly repeated and has inversions could form complex structures that may play a role in sister
chromatid or homologous chromosome pairing (127). As more genomes are sequenced from
telomere to telomere, generalizable patterns of organization may emerge that could be candidates
for experimental functional testing using high-throughput CRISPR-based screening approaches.
For example, a systematic dissection of the repeats on acrocentric short arms using CRISPR
technology may reveal nucleolus-targeting sequences.
Human centromeres are characterized by megabase-sized arrays of tandemly repeated alpha
satellite DNA, which were previously represented by large gaps or reference models (68) in
www.annualreviews.org • Long-Read DNA Sequencing 121
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
Chr8
8q21.2
Chr8:
REXO1L REXO1L REXO1L REXO1L REXO1L REXO1LREXO1L
85,650,000 85,700,000 85,750,000 85,800,000 85,850,000
Repeating elements by RepeatMasker
GENCODE v41
Gap
12.192-kb repeat units
Chr8: 85,790,000 85,800,000 85,810,000
REXO1L
Gap
REXO1L
a hg38
b CHM13
Chr8:
CHM13 unique
86,500,00086,000,000 87,000,000
Gene annotations
Chr8:
RepeatMasker repetitive elements
SINE
LINE
LTR
DNA
Simple repeat
Low-complexity repeat
85,635,000 85,645,000 85,655,000
Scale 100 kb
SINE
LINE
LTR
DNA
Simple repeat
Low-complexity repeat
ii
i
Scale 10 kb Scale 10 kbiii iv
SINE
LINE
LTR
DNA
Simple repeat
Low-complexity repeat
SINE
LINE
LTR
DNA
Simple repeat
Low-complexity repeat
GENCODE v41 GENCODE v41
Scale 500 kb
ii
i
Scale 50 kb
Chr8:
CHM13 unique
86,400,000
iii
SINE
LINE
DNA
Low-complexity repeat
Gene annotations
SINE
LINE
DNA
Low-complexity repeat
Gene annotations
Scale 20 kb
Chr8:
CHM13 unique
86,700,000 86,750,000
iv
(Caption appears on following page)
122 Warburton • Sebra
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
Figure 3 (Figure appears on preceding page)
Gap filling using LRS at the 8q21.2 VNTR composite repeat. (a, i) UCSC Genome Browser hg38 chr8:85,600,000–85,900,000
(300 kbp), showing the region containing the VNTR composite repeat and an arbitrarily sized 50-kbp gap. (ii) Dot plot showing the
repeat structure in the region. Black lines indicate tandem orientation, and red lines indicate inverted orientation. (iii) Detail of several
12-kbp repeats in the area indicated by the light blue line between subpanels i and ii, which border the gap. (iv) Detail of several repeats
in the area indicated by the green line between subpanels i and ii, distal to the gap. (b, i) UCSC Genome Browser CHM13
chr8:86,000,000–87,000,000 (1 Mbp) showing the region containing the VNTR composite repeat, which has 73 copies of a 12-kbp
repeat. CHM13 unique is additional sequence in the T2T assembly relative to hg38. (ii) Dot plot of the VNTR repeat array. Black lines
indicate tandem orientation, and red lines indicate inverted orientation. (iii) Detail of several 12-kbp repeats and an inversion in the
area indicated by the light blue line between subpanels i and ii. (iv) Detail of several 12-kbp repeats and an inversion in the area
indicated by the green line between subpanels i and ii. Abbreviations: LINE, long interspersed nuclear element; LRS, long-read
sequencing; LTR, long terminal repeat; SINE, short interspersed nuclear element; T2T, Telomere-to-Telomere; UCSC, University of
California, Santa Cruz; VNTR, variable number of tandem repeats.
the human reference assembly. The lack of a linear assembly across human centromeres has
inhibited the ability to describe sequence and epigenetic elements that provide centromere
function (41). The sequencing across complete arrays of human centromeric alpha satellite arrays
using the methods developed by the T2T consortium has revealed a high degree of substructure
across the centromeric regions, alongside epigenetic activity that may regulate function. Human
centromeres are characterized by layered expansions of alpha satellite repeat arrays, with the
active kinetochore associated mainly with the most recent array. In addition, several unexpected
types of organization have been observed in CHM13 centromeres, such as inversions, splitting
of homogeneous alpha satellite arrays by insertions of other satellite DNAs, and deletions, which
were confirmed to be polymorphic in 16 additional draft diploid assemblies, demonstrating that
ongoing evolutionary forces are at work shaping the DNA at human centromeres (5, 69).
Centromeres are the critical chromosomal locus responsible for the correct segregation of
chromosomes, which is epigenetically determined by the presence of specialized chromatin containing
the histone H3 analog CENP-A (49, 111, 126). Ultralong nanopore reads were used to
investigate the DNA methylation status across the recently emerged active alpha satellite arrays
at centromeres in CHM13 (35, 67). This has revealed that most centromeres contain a small
region of tens to hundreds of kilobase pairs of decreased methylation, called a centromere dip
region (CDR), within the otherwise heavily methylated active alpha satellite array. The use of
native chromatin immunoprecipitation or the cleavage under targets and release using nuclease
(CUT&RUN) chromatin-mapping technique (103) showed that this methyl dip region is the location
of the CENP-A chromatin (5, 35). Thus, this hypomethylated CDR represents a second
epigenetic mark, in addition to the presence of CENP-A, that appears to be important for centromere
formation and/or maintenance. To rule out that the CDR was a consequence of CHM13
being a cell type from early human development, these CDRs were confirmed in human reference
genome HG002, a diploid terminally differentiated lymphoblastoid line. At least some chromosomes
in some individuals showed exceptions, such as multiple CDRs or that CENP-A did not
occupy the most recently emerged array of alpha satellite DNA. Studying these exceptions and
variation in centromere positioning across many samples, across families, and across different tissues
from the same individuals will eventually reveal the extent of this epigenetic plasticity in
centromere localization (5). The position and size of the CDR within and between chromosomes
may have profound implications for centromere function and chromosomal mitotic and meiotic
stability (35).
The ability to generate de novo centromeres from transfected DNA sequence, permitting the
construction of artificial human chromosomes, has long been a goal of centromere research. The
discovery that the CDR domain of hypomethylated alpha satellite DNA is the true location of
the CENP-A chromatin and kinetochore may suggest additional approaches to create artificial
www.annualreviews.org • Long-Read DNA Sequencing 123
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
chromosomes, perhaps by transfection of unmethylated or unmethylatable alpha satellite DNA
constructs.If the CENP-A chromatin can be epigenetically established on the unmethylated DNA
before it becomes methylated, then it will likely propagate and potentially lead to de novo centromere
and thus artificial chromosome formation. Thus, the LRS genome-wide sequencing and
associated epigenetic analyses across human centromeres have provided some important surprises
that open many avenues of centromere structure and chromosome stability research and illuminate
the power of LRS approaches in these otherwise perplexing regions of the human genome.
THE FUTURE PANGENOME
The Human Pangenome Reference Consortium (HPRC) has the goal of creating a collection
of human diploid reference genomes that represents the true diversity of human genetics. This
daunting task will be critical for biomedical research, where recognition of the importance of
diversity inclusion in medical genomic studies is growing (102, 137). Pangenomics efforts will
require a large team of experts in population genetics, computational biology, ethics, and especially
hybrid and long-read genome sequencing technologies (69). A major goal of the current
pangenome project is to provide complete telomere-to-telomere genomes using the sequencing
technology and computational framework developed for the T2T genome assemblies, as
these gapless assemblies will be the new standard in our complete understanding of variation in
population-level genomics. The realization of a complete, albeit haploid, human genome provides
a framework for better understanding some previously unknown gaps and difficult-to-sequence
regions such as centromeres, acrocentric short arms, segmental duplications, and satellite arrays,
and many findings have been confirmed in additional genomic data (81). However, the very nature
of a pangenome project suggests that we will continuously discover additional unknowns and
potentially novel genomic structure and organization that will reveal that the full range of human
genomic diversity is much broader than previously anticipated (122).
Although the LRS technical approaches used for the T2T consortium can certainly be applied
to diploid genomes, a major challenge will be the efficient haploid-aware assembly of these diploid
genomes, especially in the difficult-to-sequence satellite and repetitive regions of the genome
(122). Powerful graph-based approaches were used for the assembly of the complete T2T genome
(9, 17, 34, 81, 82, 92). The HPRC team determined which combination of genome sequencing
and automated assembly approaches gave the most accurate diploid genome assemblies and
provided the highest-quality haploid-aware complete diploid assembly of HG002 (48). These assemblies
were not quite telomere to telomere, with some missing assembly and unmapped contigs
in centromeres and telomeric repeats,thus requiring additional development in these regions.The
HPRC will initially use select samples from the 1000 Genomes Project, which provides a catalog
of human variation from 26 populations that were consented for unrestricted data usage (2), but
additional samples will be included from other sources to support the ambitious sampling and
diversity goals of the HPRC, especially for rare variants (122). The initial goal of the HPRC is to
produce high-quality genomes from 350 individuals (resulting in 700 diverse haplotypes) selected
to maximize global diversity on which to build the foundation for the complete pangenome in
the future. One important aspect of the HPRC is to address and incorporate research on ethical,
legal, and social implications in order to include divergent and Indigenous populations and obtain
correct permissions and inclusion of the populations both in the science and in the use and
interpretation of the sequence data.
The realization of a complete pangenome dataset could revolutionize human genomics and
understanding of variation and diversity on a scale not seen since the completion of the first human
reference genome in 2001. While LRS technologies continue to improve the resolution of
124 Warburton • Sebra
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
genomic features discovered and cataloged in the HPRC efforts, consideration of a global data
federation mechanism that promotes more seamless consent, sharing, and integration of diverse
genomics data will be necessary to empower the amount of data necessary to fully compile and
translate the true power of the future pangenome database.
REMAINING CHALLENGES
LRS has provided the ability to generate DNA sequence data with read lengths routinely in the
tens to hundreds of kilobase pairs and as high as 1 Mbp using single-molecule approaches at various
accuracies. Some primary challenges remain in improving read accuracy, variable throughput,
and cost, which collectively still impede the routine utility of LRS in a high-throughput clinical
setting. LRS throughput is highly sensitive to molecular damage during library preparation,
which requires high-quality, unnicked high-molecular-weight DNA. Accurate size selection of
high-molecular-weight DNA for library preparation is important as LRS technology can be biased
toward sequencing of smaller, more rapidly diffusing molecules, which can lead to uneven
read lengths due to loading biases. Furthermore, improvement of the read accuracy in some LRS
approaches is clearly important for accurate diploid assembly as well as detecting rare pathogenic
variants.
The manufacturers of both major LRS platforms have recently announced technical innovations
that purport to overcome some major limitations, although at the time of writing, neither has
been used in a peer-reviewed publication. PacBio has announced the Revio system (86), which increases
the number of ZMWs per SMRT cell from 8 million to 25 million, with four independent
25-million-ZMW arrays on each machine and the run times reduced to 24 hours. This theoretically
increases the throughput of SMRT LRS up to 12.5 times with a greatly reduced cost, making
highly accurate human HiFi WGS genomes and highly multiplexed comprehensive long-read diagnostic
panels widely available for basic research and potentially routine clinical applications.
Alternatively, Oxford Nanopore Technologies has announced the K14 chemistry (84) with an updated
enzyme and new nanopore design that routinely improves the sequencing accuracy to Q20
(99%), with options to reach Q24 (99.6%) or higher, partially overcoming the low-accuracy limitation
of nanopore LRS. Thus, the emerging demand for higher-accuracy, higher-throughput
LRS, as well as the competition between the two major LRS platforms, is continuously driving
innovations to overcome technological limitations.
CONCLUDING REMARKS
LRS has emerged as a critical tool for the discovery of the myriad of complex variation in human
genomes. It has driven the next generation of complete human reference genomes, which include
previously intractable regions such as polymorphic gaps and large arrays of repetitive elements,
especially at centromeres and acrocentric short arms. It also provides a straightforward method
to discover rare pathogenic structural variants in basic and clinical research, and as costs decrease
and throughput and accuracy increase, it is beginning to be adopted in regulated clinical settings.
The high-resolution, fully phased long-read genomic data continue to motivate a more extensive
database of pathogenic variants that assists in disease subtyping through novel genetic associations.
Coupled with the ability to phase variants alongside epigenetic modifications and chromatin elements
across large kilobase-sized regions, LRS will continue to provide unprecedented insight
into chromatin structure, gene regulation, and disease state across the human genome. Additional
developments in LRS-based, full-length, isoform-resolved bulk and single-cell transcriptomics
integrated with whole-genome LRS and epigenetic data collectively advance the potential for
www.annualreviews.org • Long-Read DNA Sequencing 125
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
biomarker discovery on the human,tissue,or even cellular level and inform downstream functional
genomics validation studies.
While LRS has transformed our understanding of the human genome, the LRS pangenome is
the future of personalized medicine. The pangenome is only realized once individuals and clinicians
have access to high-resolution next-generation genomes that will generate a data legacy
where healthy and pathogenic associations will be unlocked as they are discovered and validated
with time. Given the wide degree of complex variation among individual diploid genomes and
across diverse human populations, LRS will very likely continue to be used independently or coupled
with complementary hybrid scaffolding technologies for closing assembly gaps or revealing
novel genomic content. Recent momentum to hurdle physical and computational LRS technology
challenges will accelerate the eventual ability to produce a de novo assembled genome for any
individual, without the need for a reference genome, permitting unprecedented understanding of
personalized human genetics and genomics. We should also consider ways to build more extensive
databases of whole-genome LRS and SRS data to federate diverse genomics datasets. As the
pangenomics database grows, near-real-time functional associations will require both adoption of
deep learning algorithms and community guidelines on how to expedite the translation of vast
discoveries enabled by LRS-based omics technologies. Ultimately, the pace of LRS data generation
and the diversity of novel variants uncovered continue to enable a clearer database of the
human condition. The future will require this genomic resolution at the individual level to provide
a truly personalized framework that unlocks information in time from a single data source
for each individual condition.
DISCLOSURE STATEMENT
R.P.S. is a paid consultant with Sema4, Stamford, Connecticut, USA.
ACKNOWLEDGMENTS
The authors thank members of the Center for Advanced Genomics Technology for helpful
discussions and Aishwarya Mandava for help with the Circos plots in Figure 2.
LITERATURE CITED
1. 1000 Genomes Proj. Consort. 2010. A map of human genome variation from population-scale
sequencing. Nature 467:1061–73. Corrigendum. 2011. Nature 473:544
2. 1000 Genomes Proj. Consort. 2015. A global reference for human genetic variation. Nature 526:68–74
3. Abdulhay NJ, McNally CP, Hsieh LJ, Kasinathan S, Keith A, et al. 2020. Massively multiplex singlemolecule
oligonucleosome footprinting. eLife 9:e59404
4. Aganezov S, Yan SM, Soto DC, Kirsche M, Zarate S, et al. 2022. A complete reference genome improves
analysis of human genetic variation. Science 376:eabl3533
5. Altemose N, Logsdon GA, Bzikadze AV, Sidhwani P, Langley SA, et al. 2022. Complete genomic and
epigenetic maps of human centromeres. Science 376:eabl4178
6. Amarasinghe SL, Su S, Dong X, Zappia L, Ritchie ME, Gouil Q. 2020. Opportunities and challenges in
long-read sequencing data analysis. Genome Biol. 21:30
7. Audano PA, Sulovari A, Graves-Lindsay TA, Cantsilieris S, Sorensen M, et al. 2019. Characterizing the
major structural variant alleles of the human genome. Cell 176:663–75.e19
8. Bailey JA, Gu Z, Clark RA, Reinert K, Samonte RV, et al. 2002. Recent segmental duplications in the
human genome. Science 297:1003–7
9. Bankevich A, Bzikadze AV, Kolmogorov M, Antipov D, Pevzner PA. 2022. Multiplex de Bruijn graphs
enable genome assembly from long, high-fidelity reads. Nat. Biotechnol. 40:1075–81
126 Warburton • Sebra
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
10. Battaglia S, Dong K, Wu J, Chen Z, Najm FJ, et al. 2022. Long-range phasing of dynamic, tissue-specific
and allele-specific regulatory elements. Nat. Genet. 54:1504–13
11. Beaulaurier J, Zhang XS, Zhu S, Sebra R, Rosenbluh C, et al. 2015. Single molecule-level detection and
long read-based phasing of epigenetic variations in bacterial methylomes. Nat. Commun. 6:7438
12. Benson G. 1999. Tandem repeats finder: a program to analyze DNA sequences. Nucleic Acids Res. 27:573–
80
13. Bruels CC, Littel HR, Daugherty AL, Stafki S, Estrella EA, et al. 2022. Diagnostic capabilities of
nanopore long-read sequencing in muscular dystrophy. Ann. Clin. Transl. Neurol. 9:1302–9
14. Byrska-Bishop M, Evani US, Zhao X, Basile AO, Abel HJ, et al. 2022. High-coverage whole-genome
sequencing of the expanded 1000 Genomes Project cohort including 602 trios. Cell 185:3426–40.e19
15. Chaisson MJP, Huddleston J, Dennis MY, Sudmant PH, Malig M, et al. 2015. Resolving the complexity
of the human genome using single-molecule sequencing. Nature 517:608–11
16. Chaisson MJP, Sanders AD, Zhao X, Malhotra A, Porubsky D, et al. 2019. Multi-platform discovery of
haplotype-resolved structural variation in human genomes. Nat. Commun. 10:1784
17. Cheng H, Concepcion GT, Feng X, Zhang H, Li H. 2021. Haplotype-resolved de novo assembly using
phased assembly graphs with hifiasm. Nat. Methods 18:170–75
18. Chintalaphani SR, Pineda SS, Deveson IW, Kumar KR. 2021. An update on the neurological short
tandem repeat expansion disorders and the emergence of long-read sequencing diagnostics. Acta
Neuropathol. Commun. 9:98
19. Coe BP, Witherspoon K, Rosenfeld JA, van Bon BW, Vulto-van Silfhout AT, et al. 2014. Refining analyses
of copy number variation identifies specific genes associated with developmental delay. Nat. Genet.
46:1063–71
20. Cohen ASA, Farrow EG, Abdelmoity AT, Alaimo JT, Amudhavalli SM, et al. 2022. Genomic answers
for children: dynamic analyses of >1000 pediatric rare disease genomes. Genet. Med. 24:1336–48
21. Dennis MY, Harshman L, Nelson BJ, Penn O, Cantsilieris S, et al. 2017. The evolution and population
diversity of human-specific segmental duplications. Nat. Ecol. Evol. 1:69
22. Dixon JR, Xu J, Dileep V, Zhan Y, Song F, et al. 2018. Integrative detection and analysis of structural
variation in cancer genomes. Nat. Genet. 50:1388–98
23. Dubocanin D, Sedeno Cortes AE, Ranchalis J, Real T, Mallory B, Stergachis AB. 2022. Single-molecule
architecture and heterogeneity of human telomeric DNA and chromatin. bioRxiv 2022.05.09.491186.
https://doi.org/10.1101/2022.05.09.491186
24. Dutta UR, Rao SN, Pidugu VK, VS V, Bhattacherjee A, et al. 2019. Breakpoint mapping of a novel
de novo translocation t(X;20)(q11.1;p13) by positional cloning and long read sequencing. Genomics
111:1108–14
25. Ebbert MTW, Farrugia SL, Sens JP, Jansen-West K, Gendron TF, et al. 2018. Long-read sequencing
across the C9orf72 ‘GGGGCC’ repeat expansion: implications for clinical use and genetic discovery
efforts in human disease. Mol. Neurodegener. 13:46
26. Ebert P, Audano PA, Zhu Q, Rodriguez-Martin B, Porubsky D, et al. 2021. Haplotype-resolved diverse
human genomes and integrated analysis of structural variation. Science 372:eabf7117
27. Eichler EE, Clark RA, She X. 2004. An assessment of the sequence gaps: unfinished business in a finished
human genome. Nat. Rev. Genet. 5:345–54
28. Fang G, Munera D, Friedman DI, Mandlik A, Chao MC, et al. 2012. Genome-wide mapping of methylated
adenine residues in pathogenic Escherichia coli using single-molecule real-time sequencing. Nat.
Biotechnol. 30:1232–39
29. Farhangdoust F, Cheng F, Liang W, Liu Y, Wanunu M. 2022. Rapid identification of DNA fragments
through direct sequencing with electro-optical zero-mode waveguides. Adv. Mater. 34:e2108479
30. Flusberg BA, Webster DR, Lee JH, Travers KJ, Olivares EC, et al. 2010. Direct detection of DNA
methylation during single-molecule, real-time sequencing. Nat. Methods 7:461–65
31. Fuller CW, Kumar S, Porel M, Chien M, Bibillo A, et al. 2016. Real-time single-molecule electronic
DNA sequencing by synthesis using polymer-tagged nucleotides on a nanopore array. PNAS 113:5233–
38
www.annualreviews.org • Long-Read DNA Sequencing 127
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
32. Gabrieli T, Sharim H, Fridman D, Arbib N, Michaeli Y, Ebenstein Y. 2018. Selective nanopore sequencing
of human BRCA1 by Cas9-assisted targeting of chromosome segments (CATCH). Nucleic Acids Res.
46:e87
33. Gall-Duncan T, Sato N, Yuen RKC, Pearson CE. 2022. Advancing genomic technologies and clinical
awareness accelerates discovery of disease-associated tandem repeat sequences. Genome Res. 32:1–27
34. Garrison E, Siren J, Novak AM, Hickey G, Eizenga JM, et al. 2018. Variation graph toolkit improves
read mapping by representing genetic variation in the reference. Nat. Biotechnol. 36:875–79
35. Gershman A, Sauria MEG, Guitart X, Vollger MR, Hook PW, et al. 2022. Epigenetic patterns in a
complete human genome. Science 376:eabj5089
36. Giesselmann P, Brandl B, Raimondeau E, Bowen R, Rohrandt C, et al. 2019. Analysis of short tandem
repeat expansions and their methylation state with nanopore sequencing. Nat. Biotechnol. 37:1478–81
37. Gilpatrick T, Lee I, Graham JE, Raimondeau E, Bowen R, et al. 2020. Targeted nanopore sequencing
with Cas9-guided adapter ligation. Nat. Biotechnol. 38:433–38
38. Gouil Q, Keniry A. 2019. Latest techniques to study DNA methylation. Essays Biochem. 63:639–48
39. Hanlon VCT, Lansdorp PM, Guryev V. 2022. A survey of current methods to detect and genotype
inversions. Hum. Mutat. 43:1576–89
40. Hasson D, Alonso A, Cheung F, Tepperberg JH, Papenhausen PR, et al. 2011. Formation of
novel CENP-A domains on tandem repetitive DNA and across chromosome breakpoints on human
chromosome 8q21 neocentromeres. Chromosoma 120:621–32
41. Hoyt SJ, Storer JM, Hartley GA, Grady PGS, Gershman A, et al. 2022. From telomere to telomere: the
transcriptional and epigenetic state of human repeat elements. Science 376:eabk3112
42. Huddleston J, Chaisson MJP, Steinberg KM, Warren W, Hoekzema K, et al. 2017. Discovery and
genotyping of structural variation from long-read haploid genome sequence data.Genome Res.27:677–85
43. Hung KL, Luebeck J, Dehkordi SR, Colon CI, Li R, et al. 2022. Targeted profiling of human
extrachromosomal DNA by CRISPR-CATCH. Nat. Genet. 54:1746–54
44. Ishiura H, Shibata S, Yoshimura J, Suzuki Y, Qu W, et al. 2019. Noncoding CGG repeat expansions in
neuronal intranuclear inclusion disease, oculopharyngodistal myopathy and an overlapping disease. Nat.
Genet. 51:1222–32
45. Jacobs PA, Wilson CM, Sprenkle JA, Rosenshein NB, Migeon BR. 1980. Mechanism of origin of
complete hydatidiform moles. Nature 286:714–16
46. Jain M, Koren S, Miga KH, Quick J, Rand AC, et al. 2018. Nanopore sequencing and assembly of a
human genome with ultra-long reads. Nat. Biotechnol. 36:338–45
47. Jain M, Olsen HE, Turner DJ, Stoddart D, Bulazel KV, et al. 2018. Linear assembly of a human
centromere on the Y chromosome. Nat. Biotechnol. 36:321–23
48. Jarvis ED, Formenti G, Rhie A, Guarracino A, Yang C, et al. 2022. Semi-automated assembly of highquality
diploid human reference genomes. Nature 611:519–31
49. Karpen GH, Allshire RC. 1997. The case for epigenetic effects on centromere identity and function.
Trends Genet. 13:489–96
50. Kidd JM, Cooper GM, Donahue WF, Hayden HS, Sampas N, et al. 2008. Mapping and sequencing of
structural variation from eight human genomes. Nature 453:56–64
51. Korbel JO, Urban AE, Affourtit JP, Godwin B, Grubert F, et al. 2007. Paired-end mapping reveals
extensive structural variation in the human genome. Science 318:420–26
52. Lander ES, Linton LM, Birren B, Nusbaum C, Zody MC, et al. 2001. Initial sequencing and analysis of
the human genome. Nature 409:860–921
53. Larkin J, Henley RY, Jadhav V, Korlach J, Wanunu M. 2017. Length-independent DNA packing into
nanopore zero-mode waveguides for low-input DNA sequencing. Nat. Nanotechnol. 12:1169–75
54. Lee I, Razaghi R, Gilpatrick T, Molnar M, Gershman A, et al. 2020. Simultaneous profiling of chromatin
accessibility and methylation on human cell lines with nanopore sequencing. Nat. Methods 17:1191–99
55. Levy-Sakin M, Pastor S, Mostovoy Y, Li L, Leung AKY, et al. 2019. Genome maps across 26 human
populations reveal population-specific patterns of structural variation. Nat. Commun. 10:1025
56. Lieberman-Aiden E, van Berkum NL, Williams L, Imakaev M, Ragoczy T, et al. 2009. Comprehensive
mapping of long-range interactions reveals folding principles of the human genome. Science 326:289–93
128 Warburton • Sebra
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
57. Liu Y, Cheng J, Siejka-Zielinska P, Weldon C, Roberts H, et al. 2020. Accurate targeted long-read DNA
methylation and hydroxymethylation sequencing with TAPS. Genome Biol. 21:54
58. Logsdon GA, Vollger MR, Eichler EE. 2020. Long-read human genome sequencing and its applications.
Nat. Rev. Genet. 21:597–614
59. Logsdon GA, Vollger MR, Hsieh P, Mao Y, Liskovykh MA, et al. 2021. The structure, function and
evolution of a complete human chromosome 8. Nature 593:101–7
60. MacDonald JR, Ziman R, Yuen RK, Feuk L, Scherer SW. 2014. The Database of Genomic Variants: a
curated collection of structural variation in the human genome. Nucleic Acids Res. 42:D986–92
61. Mangin A, de Pontual L, Tsai YC, Monteil L, Nizon M, et al. 2021. Robust detection of somatic mosaicism
and repeat interruptions by long-read targeted sequencing in myotonic dystrophy type 1. Int. J.
Mol. Sci. 22:2616
62. Marwaha S, Knowles JW, Ashley EA. 2022. A guide for the diagnosis of rare and undiagnosed disease:
beyond the exome. Genome Med. 14:23
63. Mayor NP, Robinson J, McWhinnie AJ, Ranade S, Eng K, et al. 2015. HLA typing for the next
generation. PLOS ONE 10:e0127153
64. Melas M, Kautto EA, Franklin SJ, Mori M, McBride KL, et al. 2022. Long-read whole genome
sequencing reveals HOXD13 alterations in synpolydactyly. Hum. Mutat. 43:189–99
65. Merker JD, Wenger AM, Sneddon T, Grove M, Zappala Z, et al. 2018. Long-read genome sequencing
identifies causal structural variation in a Mendelian disease. Genet. Med. 20:159–63
66. Miao H, Zhou J, Yang Q, Liang F, Wang D, et al. 2018. Long-read sequencing identified a causal
structural variant in an exome-negative case and enabled preimplantation genetic diagnosis. Hereditas
155:32
67. Miga KH, Koren S, Rhie A, Vollger MR, Gershman A, et al. 2020. Telomere-to-telomere assembly of a
complete human X chromosome. Nature 585:79–84
68. Miga KH, Newton Y, Jain M, Altemose N, Willard HF, Kent WJ. 2014. Centromere reference models
for human chromosomes X and Y satellite arrays. Genome Res. 24:697–707
69. Miga KH, Wang T. 2021. The need for a human pangenome reference sequence. Annu. Rev. Genom.
Hum. Genet. 22:81–102
70. Miller DE,Hanna P,Galey M,Reyes M,Linglart A,et al.2022.Targeted long-read sequencing identifies
a retrotransposon insertion as a cause of altered GNAS exon A/B methylation in a family with autosomal
dominant pseudohypoparathyroidism type 1b (PHP1B). J. Bone Miner. Res. 37:1711–19
71. Miller DE, Lee L, Galey M, Kandhaya-Pillai R, Tischkowitz M, et al. 2022. Targeted long-read sequencing
identifies missing pathogenic variants in unsolved Werner syndrome cases. J. Med. Genet.
59:1087–94
72. Miller DE, Sulovari A, Wang T, Loucks H, Hoekzema K, et al. 2021. Targeted long-read sequencing
identifies missing disease-causing variation. Am. J. Hum. Genet. 108:1436–49
73. Mills RE, Walter K, Stewart C, Handsaker RE, Chen K, et al. 2011. Mapping copy number variation by
population-scale genome sequencing. Nature 470:59–65
74. Mitsuhashi S, Frith MC, Mizuguchi T, Miyatake S, Toyota T, et al. 2019. Tandem-genotypes: robust
detection of tandem repeat expansions from long DNA reads. Genome Biol. 20:58
75. Mizuguchi T, Okamoto N, Yanagihara K, Miyatake S, Uchiyama Y, et al. 2021. Pathogenic 12-kb copyneutral
inversion in syndromic intellectual disability identified by high-fidelity long-read sequencing.
Genomics 113:1044–53
76. Mizuguchi T, Suzuki T, Abe C, Umemura A, Tokunaga K, et al. 2019. A 12-kb structural variation in
progressive myoclonic epilepsy was newly identified by long-read whole-genome sequencing. J. Hum.
Genet. 64:359–68
77. Mizuguchi T, Toyota T, Adachi H, Miyake N, Matsumoto N, Miyatake S. 2019. Detecting a long insertion
variant in SAMD12 by SMRT sequencing: implications of long-read whole-genome sequencing
for repeat expansion diseases. J. Hum. Genet. 64:191–97
78. Morales J, Pujar S, Loveland JE, Astashyn A, Bennett R, et al. 2022. A joint NCBI and EMBL-EBI
transcript set for clinical genomics and research. Nature 604:310–15
www.annualreviews.org • Long-Read DNA Sequencing 129
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
79. Morato Torres CA, Zafar F, Tsai YC, Vazquez JP, Gallagher MD, et al. 2022. ATTCT and ATTCC
repeat expansions in the ATXN10 gene affect disease penetrance of spinocerebellar ataxia type 10. HGG
Adv. 3:100137
80. Ni P, Zhong Z, Xu J, Huang N, Zhang J, et al. 2023. DNA 5-methylcytosine detection and methylation
phasing using PacBio circular consensus sequencing. bioRxiv 2022.02.26.482074. https://doi.org/10.
1101/2022.02.26.482074
81. Nurk S, Koren S, Rhie A, Rautiainen M, Bzikadze AV, et al. 2022. The complete sequence of a human
genome. Science 376:44–53
82. Nurk S, Walenz BP, Rhie A, Vollger MR, Logsdon GA, et al. 2020. HiCanu: accurate assembly
of segmental duplications, satellites, and allelic variants from high-fidelity long reads. Genome Res.
30:1291–305
83. Olivares-Chauvet P, Mukamel Z, Lifshitz A, Schwartzman O, Elkayam NO, et al. 2016. Capturing
pairwise and multi-way chromosomal conformations using chromosomal walks. Nature 540:296–300
84. Oxford Nanopore Technol. 2022. The power of Q20+ chemistry. Oxford Nanopore Technologies. https://
nanoporetech.com/q20plus-chemistry
85. PacBio. 2022. pbsv. GitHub. https://github.com/PacificBiosciences/pbsv
86. PacBio. 2022. Revio system. PacBio. https://www.pacb.com/revio
87. Pendleton M, Sebra R, Pang AW, Ummat A, Franzen O, et al. 2015. Assembly and diploid architecture
of an individual human genome via single-molecule technologies. Nat. Methods 12:780–86
88. Peters BA, Kermani BG, Sparks AB, Alferov O, Hong P, et al. 2012. Accurate whole-genome sequencing
and haplotyping from 10 to 20 human cells. Nature 487:190–95
89. Qiao W, Yang Y, Sebra R, Mendiratta G, Gaedigk A, et al. 2016. Long-read single molecule real-time
full gene sequencing of cytochrome P450-2D6. Hum. Mutat. 37:315–23
90. Rand AC, Jain M, Eizenga JM, Musselman-Brown A, Olsen HE, et al. 2017. Mapping DNA methylation
with high-throughput nanopore sequencing. Nat. Methods 14:411–13
91. Rang FJ, Kloosterman WP, de Ridder J. 2018. From squiggle to basepair: computational approaches for
improving nanopore sequencing read accuracy. Genome Biol. 19:90
92. Rautiainen M, Marschall T. 2020. GraphAligner: rapid and versatile sequence-to-graph alignment.
Genome Biol. 21:253
93. Reiner J, Pisani L, Qiao W, Singh R, Yang Y, et al. 2018. Cytogenomic identification and long-read
single molecule real-time (SMRT) sequencing of a Bardet-Biedl Syndrome 9 (BBS9) deletion. npj Genom.
Med. 3:3
94. Robinson J, Barker DJ, Georgiou X, Cooper MA, Flicek P, Marsh SGE. 2020. IPD-IMGT/HLA
Database. Nucleic Acids Res. 48:D948–55
95. Rubben K, Tilleman L, Deserranno K, Tytgat O, Deforce D, Van Nieuwerburgh F. 2022. Cas9 targeted
nanopore sequencing with enhanced variant calling improves CYP2D6-CYP2D7 hybrid allele
genotyping. PLOS Genet. 18:e1010176
96. Sanders AD, Falconer E, Hills M, Spierings DCJ, Lansdorp PM. 2017. Single-cell template strand
sequencing by Strand-seq enables the characterization of individual homologs. Nat. Protoc. 12:1151–76
97. Sano Y, Koyanagi Y, Wong JH, Murakami Y, Fujiwara K, et al. 2022. Likely pathogenic structural variants
in genetically unsolved patients with retinitis pigmentosa revealed by long-read sequencing. J. Med.
Genet. 59:1133–38
98. Sato N, Amino T, Kobayashi K, Asakawa S, Ishiguro T, et al. 2009. Spinocerebellar ataxia type 31 is
associated with “inserted” penta-nucleotide repeats containing (TGGAA)n. Am. J. Hum. Genet. 85:544–
57
99. Schneider VA, Graves-Lindsay T, Howe K, Bouk N, Chen HC, et al. 2017. Evaluation of GRCh38
and de novo haploid genome assemblies demonstrates the enduring quality of the reference assembly.
Genome Res. 27:849–64
100. Scott ER, Yang Y, Botton MR, Seki Y, Hoshitsuki K, et al. 2022. Long-read HiFi sequencing
of NUDT15: phased full-gene haplotyping and pharmacogenomic allele discovery. Hum. Mutat.
43:1557–66
101. Simpson JT, Workman RE, Zuzarte PC, David M, Dursi LJ, Timp W. 2017. Detecting DNA cytosine
methylation using nanopore sequencing. Nat. Methods 14:407–10
130 Warburton • Sebra
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
102. Sirugo G, Williams SM, Tishkoff SA. 2019. The missing diversity in human genetic studies. Cell 177:26–
31
103. Skene PJ, Henikoff S. 2017. An efficient targeted nuclease strategy for high-resolution mapping of DNA
binding sites. eLife 6:e21856
104. Smit AF. 1999. Interspersed repeats and other mementos of transposable elements in mammalian
genomes. Curr. Opin. Genet. Dev. 9:657–63
105. Smith ZD, Meissner A. 2013. DNA methylation: roles in mammalian development. Nat. Rev. Genet.
14:204–20
106. Sone J, Mitsuhashi S, Fujita A, Mizuguchi T, Hamanaka K, et al. 2019. Long-read sequencing identifies
GGC repeat expansions in NOTCH2NLC associated with neuronal intranuclear inclusion disease. Nat.
Genet. 51:1215–21
107. Song XH, Hsu HK, Su MT, Chang TS, Su PY, et al. 2017. Euchromatic variants of 8q21.2 in twins.
Taiwan J. Obstet. Gynecol. 56:227–29
108. Steiert TA, Fuss J, Juzenas S, Wittig M, Hoeppner MP, et al. 2022. High-throughput method for
the hybridisation-based targeted enrichment of long genomic fragments for PacBio third-generation
sequencing. NAR Genom. Bioinform. 4:lqac051
109. Stergachis AB, Debo BM, Haugen E, Churchman LS, Stamatoyannopoulos JA. 2020. Single-molecule
regulatory architectures captured by chromatin fiber sequencing. Science 368:1449–54
110. Stranges PB, Palla M, Kalachikov S, Nivala J, Dorwart M, et al. 2016. Design and characterization of a
nanopore-coupled polymerase for single-molecule DNA sequencing by synthesis on an electrode array.
PNAS 113:E6749–56
111. Sullivan BA, Karpen GH. 2004. Centromeric chromatin exhibits a histone modification pattern that is
distinct from both euchromatin and heterochromatin. Nat. Struct. Mol. Biol. 11:1076–83
112. Sun Z, Vaisvila R, Hussong LM, Yan B, Baum C, et al. 2021. Nondestructive enzymatic deamination
enables single-molecule long-read amplicon sequencing for the determination of 5-methylcytosine and
5-hydroxymethylcytosine at single-base resolution. Genome Res. 31:291–300
113. Tsai YC, de Pontual L, Heiner C, Stojkovic T, Furling D, et al. 2022. Identification of a CCG-enriched
expanded allele in DM1 patients using amplification-free long-read sequencing. J. Mol. Diagn. 24:1143–
54
114. Tse OYO, Jiang P, Cheng SH, Peng W, Shang H, et al. 2021. Genome-wide detection of cytosine
methylation by single molecule real-time sequencing. PNAS 118:e2019768118
115. Tyson C, Sharp AJ, Hrynchak M, Yong SL, Hollox EJ, et al. 2014. Expansion of a 12-kb VNTR containing
the REXO1L1 gene cluster underlies the microscopically visible euchromatic variant of 8q21.2.
Eur. J. Hum. Genet. 22:458–63
116. van Buuren N, Ramirez R, Soulette C, Suri V, Han D, et al. 2022. Targeted long-read sequencing reveals
clonally expanded HBV-associated chromosomal translocations in patients with chronic hepatitis
B. JHEP Rep. 4:100449
117. van der Ende EL,Jackson JL,White A,Seelaar H,van Blitterswijk M,Van Swieten JC.2021.Unravelling
the clinical spectrum and the role of repeat length in C9ORF72 repeat expansions. J. Neurol. Neurosurg.
Psychiatry 92:502–9
118. Vasan N, Razavi P, Johnson JL, Shao H, Shah H, et al. 2019. Double PIK3CA mutations in cis increase
oncogenicity and sensitivity to PI3Kα inhibitors. Science 366:714–23
119. Vollger MR, Guitart X, Dishuck PC, Mercuri L, Harvey WT, et al. 2022. Segmental duplications and
their variation in a complete human genome. Science 376:eabj6965
120. Vollger MR, Logsdon GA, Audano PA, Sulovari A, Porubsky D, et al. 2020. Improved assembly and
variant detection of a haploid human genome using single-molecule, high-fidelity long reads. Ann. Hum.
Genet. 84:125–40
121. Wang S, Lee S, Chu C, Jain D, Kerpedjiev P, et al. 2020. HiNT: a computational method for detecting
copy number variations and translocations from Hi-C data. Genome Biol. 21:73
122. Wang T, Antonacci-Fulton L, Howe K, Lawson HA, Lucas JK, et al. 2022. The Human Pangenome
Project: a global resource to map genomic diversity. Nature 604:437–46
123. Wang X, Luan Y, Yue F. 2022. EagleC: a deep-learning framework for detecting a full range of structural
variations from bulk and single-cell contact maps. Sci. Adv. 8:eabn9215
www.annualreviews.org • Long-Read DNA Sequencing 131
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
124. Wang Y, Zhao Y, Bollas A, Wang Y, Au KF. 2021. Nanopore sequencing technology, bioinformatics and
applications. Nat. Biotechnol. 39:1348–65
125. Wang YC,Olson ND,Deikus G,Shah H,Wenger AM,et al.2019.High-coverage,long-read sequencing
of Han Chinese trio reference samples. Sci. Data 6:91
126. Warburton PE.2001.Epigenetic analysis of kinetochore assembly on variant human centromeres.Trends
Genet. 17:243–47
127. Warburton PE, Giordano J, Cheung F, Gelfand Y, Benson G. 2004. Inverted repeat structure of the
human genome: the X-chromosome contains a preponderance of large, highly homologous inverted
repeats that contain testes genes. Genome Res. 14:1861–69
128. Warburton PE,Hasson D,Guillem F,Lescale C,Jin X,Abrusan G.2008.Analysis of the largest tandemly
repeated DNA families in the human genome. BMC Genom. 9:533
129. Wenger AM,Peluso P,Rowell WJ,Chang PC,Hall RJ,et al.2019.Accurate circular consensus long-read
sequencing improves variant detection and assembly of a human genome. Nat. Biotechnol. 37:1155–62
130. Wenzel A, Altmueller J, Ekici AB, Popp B, Stueber K, et al. 2018. Single molecule real time sequencing in
ADTKD-MUC1 allows complete assembly of the VNTR and exact positioning of causative mutations.
Sci. Rep. 8:4170
131. Yang H, Garcia-Manero G, Sasaki K, Montalban-Bravo G, Tang Z, et al. 2022. High-resolution structural
variant profiling of myelodysplastic syndromes by optical genome mapping uncovers cryptic
aberrations of prognostic and therapeutic significance. Leukemia 36:2306–16
132. Yang Y, Sebra R, Pullman BS, Qiao W, Peter I, et al. 2015. Quantitative and multiplexed DNA methylation
analysis using long-read single-molecule real-time bisulfite sequencing (SMRT-BS). BMC Genom.
16:350
133. Zeng S, Zhang MY, Wang XJ, Hu ZM, Li JC, et al. 2019. Long-read sequencing identified intronic
repeat expansions in SAMD12 from Chinese pedigrees affected with familial cortical myoclonic tremor
with epilepsy. J. Med. Genet. 56:265–70
134. Zhang S, Pei Z, Lei C, Zhu S, Deng K, et al. 2023. Detection of cryptic balanced chromosomal
rearrangements using high-resolution optical genome mapping. J. Med. Genet. 60:274–84
135. Zhao X, Collins RL, Lee WP, Weber AM, Jun Y, et al. 2021. Expectations and blind spots for structural
variation detection from long-read assemblies and short-read genome sequencing technologies. Am. J.
Hum. Genet. 108:919–28
136. Zook JM, Catoe D, McDaniel J, Vang L, Spies N, et al. 2016. Extensive sequencing of seven human
genomes to characterize benchmark reference materials. Sci. Data 3:160025
137. Zook JM, Hansen NF, Olson ND, Chapman L, Mullikin JC, et al. 2020. A robust benchmark for
detection of germline large deletions and insertions. Nat. Biotechnol. 38:1347–55
132 Warburton • Sebra
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
GG24_TOC ARjats.cls July 5, 2023 17:48
Annual Review of
Genomics and
Human Genetics
Volume 24, 2023
Contents
A Journey from Blood Cells to Genes and Back
Lucio Luzzatto p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1
Meiotic Chromosome Structure, the Synaptonemal Complex,
and Infertility
Ian R. Adams and Owen R. Davies p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p35
The p-Arms of Human Acrocentric Chromosomes Play
by a Different Set of Rules
Brian McStay p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p63
RNA Crossing Membranes: Systems and Mechanisms Contextualizing
Extracellular RNA and Cell Surface GlycoRNAs
Peiyuan Chai, Charlotta G. Lebedenko, and Ryan A. Flynn p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p85
Long-Read DNA Sequencing: Recent Advances
and Remaining Challenges
Peter E. Warburton and Robert P. Sebra p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 109
Padlock Probe–Based Targeted In Situ Sequencing: Overview
of Methods and Applications
Anastasia Magoulopoulou, Sergio Marco Salas, Katarína Tiklová,
Erik Reinhold Samuelsson, Markus M. Hilscher, and Mats Nilsson p p p p p p p p p p p p p p p p p p p 133
DECIPHER: Improving Genetic Diagnosis Through Dynamic
Integration of Genomic and Clinical Data
Julia Foreman, Daniel Perrett, Erica Mazaika, Sarah E. Hunt,
James S. Ware, and Helen V. Firth p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 151
The Genetic Determinants of Axial Length: From Microphthalmia
to High Myopia in Childhood
Daniel Jackson and Mariya Moosajee p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 177
The SWI/SNF Complex in Neural Crest Cell Development
and Disease
Daniel M. Fountain and Tatjana Sauka-Spengler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 203
TGF-β and BMP Signaling Pathways in Skeletal Dysplasia
with Short and Tall Stature
Alice Costantini, Alessandra Guasto, and Valérie Cormier-Daire p p p p p p p p p p p p p p p p p p p p p p p p 225
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.
GG24_TOC ARjats.cls July 5, 2023 17:48
Sickle Cell Disease: From Genetics to Curative Approaches
Giulia Hardouin, Elisa Magrin, Alice Corsia, Marina Cavazzana,
Annarita Miccio, and Michaela Semeraro p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 255
Methods and Insights from Single-Cell Expression Quantitative
Trait Loci
Joyce B. Kang, Alessandro Raveane, Aparna Nathan, Nicole Soranzo,
and Soumya Raychaudhuri p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 277
Methods for Assessing Population Relationships and History
Using Genomic Data
Priya Moorjani and Garrett Hellenthal p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 305
Avoiding Liability and Other Legal Land Mines in the Evolving
Genomics Landscape
Ellen Wright Clayton, Alex M. Tritell, and Adrian M. Thorogood p p p p p p p p p p p p p p p p p p p p p p 333
Federated Analysis for Privacy-Preserving Data Sharing:
A Technical and Legal Primer
James Casaletto, Alexander Bernier, Robyn McDougall, and Melissa S. Cline p p p p p p p p p p p 347
Open Data in the Era of the GDPR: Lessons from the Human
Cell Atlas
Bartha Maria Knoppers, Alexander Bernier, Sarion Bowers, and Emily Kirby p p p p p p p p p p 369
Return of Results in Genomic Research Using Large-Scale
or Whole Genome Sequencing: Toward a New Normal
Susan M. Wolf and Robert C. Green p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 393
Errata
An online log of corrections to Annual Review of Genomics and Human Genetics articles
may be found at http://www.annualreviews.org/errata/genom
Annu.Rev.Genom.Hum.Genet.2023.24:109-132.Downloadedfromwww.annualreviews.org
Accessprovidedby212.89.239.228on09/24/23.Seecopyrightforapproveduse.