Third Generation DNA Sequencing with a Nanopore - Nanopores: Sensing and Fundamental Biological Interactions - page 289

de-phasing, where bases become out of step with the rest in the cluster, which
increases fluorescence noise and causes base-calling errors forcing shorter reads [ 4 ].

12.1.1 Single Molecule Sequencing

On the other hand, de-phasing is not an issue with single-molecule templates and so
the requirement for cycling efficiency is relaxed. However, single molecule tem-
plates are still susceptible to multiple nucleotide additions in a cycle. Deletion
errors occur due to either quenching effects between adjacent dye molecules or
incorporation of dark nucleotides. And finally, achieving the signal-to-noise ratio
(SNR) required for single molecule detection still remains a challenge. Pacific
Biosystems is developing a SMRT (single molecule real time) sequencing method
that relies on enzymatic incorporation of a fluorescently labeled nucleotide through
a DNA-polymerase, applying zero-mode waveguide (ZMW) technique to suppress
the ambient radiation so that one molecule can be identified. While this strategy
boasts the potential for 5-10 Gbp/s from an array of ZMWs, it is still error prone
and suffers polymerase-dependent shortcomings [ 5 ].
The polymerase-dependent methods favor short reads due to the nature of a
processive sequencing process; because errors accumulate over time, the quality of
the base-calling deteriorates with read length. As a result of the short reads, a
bioinformatic bottleneck develops, which makes alignment and assembly of the
genome an especially vexing proposition [ 6 ]. It has been shown that the assembly
quality deteriorates rapidly as the read length decreases [ 7 , 8 ]. For Sanger sequencing
(750 bp) an assembly of Neisseria meningitidis resulted in 59 contigs, 48 of which
were 1,800 contigs. The

assembly was still fragmented (296 contigs) even for relatively long reads (200 bp)
within reach of NGS [ 7 ]. Irrespective of the sequencing technology, de novo assembly
becomes difficult whenever the length of the read is shorter than the repetitive
segments in the sequence because the sequences cannot be aligned unambiguously.
For such cases assembly falls into a class of NP-hard problems with no efficient
computational solution [ 6 ]. This is especially problematic with a large number of
repeats - which can be used for genotyping or relevant in mobile genetic element
location [ 9 ]. It becomes exponentially harder to assemble a genome as the number of
repeats grows.
Single molecule DNA sequencing represents the logical, end-of-the-line in
development sequencing technology, which extracts the maximum amount of
information from a minimum of material and pre-processing. The low material
requirement coupled with quick results would allow for easy sequencing of pre-
cious primary samples from human patients, e.g. allowing doctors to look for rare
diseases in the pre-clinical stage [ 10 ]. Among the emerging third generation
technologies, sequencing a single molecule of DNA with a nanopore seems to
have the brightest prospects [ 11 ], because it has the potential for very long read
lengths (

> 1 kbp, whereas at 70 bp, the assembly consisted of >

> 1 kbp), thereby reducing the time required for alignment of the reads and