That wish—echoed by virtually everyone interviewed
for this article—will eventually be fulfilled. Andy
Felton, PhD, head of product management for Life
Technologies’ Ion Torrent Business Unit, said, the
data formats for NGS platforms are already “pretty
standard.” He said vendors are developing software
applications “to help people make sense of the
massive datasets NGS generates. Tools for doing
that analysis will get simpler and simpler... It’s a
question of evolution of the software.”
If history is any guide, the wheels of evolution are
spinning ever faster when it comes to genetics-related innovation.
13,330 bases per penny
Ever since the DNA double helix was first described
in 1953, the pace of technological change has been
growing. The first sequencing technologies were
developed in the 1970s. The most notable of these was
the brainchild of British biochemist, Frederick Sanger,
who shared the Nobel Prize for Chemistry the year his
method debuted. But although the original Sanger
sequencing technique was a breakthrough, it is a
tedious and time-consuming affair.
According to the method, a single-stranded DNA
template is replicated and divided into four sequencing reactions. Each of these requires specially
designed primers, polymerase enzymes, the four DNA
nucleotides (adenine, thymine, guanine, and cytosine
or A, T, G and C) and one modified, radioactively-labeled “dideoxy” nucleotide that terminates chain
elongation. The end result is four pools of variously
sized DNA fragments, terminated at every occurrence
of A, T, G or C. The four pools are then heat denatured
and run in four columns using gel electrophoresis to
separate the fragments by size. Finally, photographic
film is applied to the gel in a dark room to imprint
the radioactive DNA bands. The genetic sequence can
then be read, bottom up, from the film, based on the
relative positions of the bands. But the process is more
art than science. (It can also, of course, be used to
sequence RNA.)
Victor Waddell, PhD, head of the Arizona Public Health
Laboratory remembers using the method as a graduate student in Ireland. He said there were “a lot of
failures.” Even under the best of circumstances, it took
a week to get anywhere from 300 to 1,200 base pairs
of data.
The first innovation was to use four differently colored
fluorescent labels in place of the radioactive label,
enabling all four reactions to run as one and eliminating the film step. The next improvement, in the late
1980s, was the use of capillary electrophoresis in place
of gel electrophoresis. Finally, arrays of capillaries
were employed to speed the process. An automated
version of Sanger sequencing is now available as a
lab-on-a-chip using nano-scale volumes. Even
today, this is considered the gold standard sequencing
method, used as the basis for the $2.7 billion Human
Genome Project.
The problem is that Sanger sequencing is not scalable
for high-throughput applications.
The first NGS methods were introduced in 2005,
based on massively parallel sequencing techniques.
The basic idea is to massively replicate a DNA (or
RNA) sample; break it into random, short, segments;
sequence those segments; and then feed the resulting data into a software program that pieces together
the original long sequence by identifying overlapping
regions on the segments.
Daryl Lamson, who helps manage special projects
at New York’s Wadsworth Center, the state PHL, said,
“You can get hundreds of replicate sequences over a
region. The software sorts that out. But you have to be
able to understand what the software is doing—what
it is capable of doing and what it is not capable of
doing—and what you’re going to do about gaps in the
sequences. Definitely there can be gaps.”
The sequencing itself is performed by various
methods, depending upon the test platform. Roche’s
GS FLX 454—the first commercial NGS platform—
uses a sequencing-by-synthesis method called
pyrosequencing. The instrument tries A, T, G and C
in turn at each position on a single-stranded DNA
segment. Each time a nucleotide successfully bonds
to a strand, a burst of light is emitted and recorded by
the instrument, which can monitor hundreds of thousands of flashes at once. The Ion Torrent uses a similar
technique, but detects the hydrogen ion released as
a byproduct of the sequencing reaction instead of
flashes. Several other methods are also in use.
Ion 314 Chip
The advantages of NGS techniques are obvious. A
Sanger capillary platform will sequence 1,000 base
pairs in an hour for $10, or a penny a base. The Ion
Torrent will sequence a billion base pairs in 2-3 hours
for $750 or roughly 13,330 bases per penny.
