sequence data from each cluster are collected as short reads of
300 bp, which tend to have very accurate fidelity to the original
DNA sequence [21]. These individual short reads can then be
bioinformatically assembled into larger genomic contigs. However,
it is generally difficult to create fully finished genomes using shortread sequencing only, due to the inability of the short reads to span
the length of complex or repetitive regions in the genome.
Long-read sequencing produces individual reads that are of
much greater length. PacBio sequencing uses a technology called
Single Molecule, Real-Time (SMRT
® ) sequencing. After the genomic DNA is isolated, hairpin adapters are ligated to either side of
the double-stranded DNA, creating a genomic library that consists
of circular template molecules. These circular template molecules
then bind to a DNA polymerase inside a zero-mode waveguide
(ZMW). Because the molecule being read is circular, each polymerase sequences the same DNA strand (and its complement) repeatedly. On an individual read basis, the PacBio sequencing method
has a relatively high error rate (%12%) relative to the Illumina
platforms. However, the repeated sequencing of each circularized
template molecule allows the correction of sequencing errors by
generation of a consensus sequence from each of the individual
reads. This technique is known as circular consensus sequencing
(CCS) [19, 20, 22, 23]. The length of the consensus reads typically
reaches several tens of kilobases, allowing reads to span across most
complex or repetitive regions, which would be unresolvable using
only short-read sequencing. During the library preparation, the
DNA is sheared into fragments of %10 kb. Thus, it is important
to have high-molecular weight DNA as an input; low-quality DNA
that has been sheared into smaller fragments places an upper limit
on the length of reads that can be generated. Each PacBio SMRT
Cell contains millions of ZMWs, and so a large volume of sequence
data is produced, making it easy to generate WGS for bacteria.
The technology employed by the Oxford Nanopore systems is
capable of producing even longer reads. The ONT platforms are
based upon the translocation of a DNA strand through a genetically
modified bacterial nanopore made of pore-forming proteins. As the
DNA strand passes through the pore, the ionic current in the
channel is disrupted by the nucleotides. A sensor detects and
records these electrical disturbances created as the DNA strand
passes through the nanopore, creating a trace known as a “squiggle”. Because each of the four nucleotides produces a different
characteristic change in this electrical conductance, the squiggle
can be algorithmically mapped back to the nucleotides and the
order in which they passed through the pore, thus yielding the
sequence of the DNA strand. Because the strands of DNA are not
fragmented during library prep, ONT sequencing allows the generation of ultra-long sequencing reads. There is no theoretical limit
to the length of a read; in practice, read length is governed only by
406
Marco A. Riojas et al.
300 bp, which tend to have very accurate fidelity to the original
DNA sequence [21]. These individual short reads can then be
bioinformatically assembled into larger genomic contigs. However,
it is generally difficult to create fully finished genomes using shortread sequencing only, due to the inability of the short reads to span
the length of complex or repetitive regions in the genome.
Long-read sequencing produces individual reads that are of
much greater length. PacBio sequencing uses a technology called
Single Molecule, Real-Time (SMRT
® ) sequencing. After the genomic DNA is isolated, hairpin adapters are ligated to either side of
the double-stranded DNA, creating a genomic library that consists
of circular template molecules. These circular template molecules
then bind to a DNA polymerase inside a zero-mode waveguide
(ZMW). Because the molecule being read is circular, each polymerase sequences the same DNA strand (and its complement) repeatedly. On an individual read basis, the PacBio sequencing method
has a relatively high error rate (%12%) relative to the Illumina
platforms. However, the repeated sequencing of each circularized
template molecule allows the correction of sequencing errors by
generation of a consensus sequence from each of the individual
reads. This technique is known as circular consensus sequencing
(CCS) [19, 20, 22, 23]. The length of the consensus reads typically
reaches several tens of kilobases, allowing reads to span across most
complex or repetitive regions, which would be unresolvable using
only short-read sequencing. During the library preparation, the
DNA is sheared into fragments of %10 kb. Thus, it is important
to have high-molecular weight DNA as an input; low-quality DNA
that has been sheared into smaller fragments places an upper limit
on the length of reads that can be generated. Each PacBio SMRT
Cell contains millions of ZMWs, and so a large volume of sequence
data is produced, making it easy to generate WGS for bacteria.
The technology employed by the Oxford Nanopore systems is
capable of producing even longer reads. The ONT platforms are
based upon the translocation of a DNA strand through a genetically
modified bacterial nanopore made of pore-forming proteins. As the
DNA strand passes through the pore, the ionic current in the
channel is disrupted by the nucleotides. A sensor detects and
records these electrical disturbances created as the DNA strand
passes through the nanopore, creating a trace known as a “squiggle”. Because each of the four nucleotides produces a different
characteristic change in this electrical conductance, the squiggle
can be algorithmically mapped back to the nucleotides and the
order in which they passed through the pore, thus yielding the
sequence of the DNA strand. Because the strands of DNA are not
fragmented during library prep, ONT sequencing allows the generation of ultra-long sequencing reads. There is no theoretical limit
to the length of a read; in practice, read length is governed only by
406
Marco A. Riojas et al.