Personalized Medicine - Translational Informatics: Realizing the Promise of Knowledge-Driven Healthcare - page 37

ing DNA, adding one base at a time and observing which base is incorporated in any
given step. In January 2014, Illumina announced that through the release of its
HiSeq X Ten, the long sought-after $1,000 genome barrier had been broken [ 5 ]. One
downside to NextGen sequencing is that the read lengths are relatively short (on the
order of 100 bases). This makes it impossible to accurately sequence some portions
of the genome, particularly highly repetitive regions.
Third generation sequencing is aimed at taking the next big leap: continuous
single-molecule sequencing. Instead of determining a very long DNA sequence
by sequencing millions of short reads and then aligning them, third generation
sequencing reads the individual based pairs from a single molecule [ 6 ]. Major
advantages to this approach include the small sample size required and increased
accuracy in highly repetitive regions of DNA. As the price of sequencing has
dropped, these techniques are increasingly being used not only for DNA but for
other assays too, e.g. RNA expression (RNA-seq) and transcription factor binding
assays (ChIP-seq).
With these technological advances for generating massive amounts of omics
data, biologists can no longer glean new scientifi c knowledge simply by looking at
a gel or a spreadsheet. Rather, computational tools are required to make sense of
tens of thousands of data points from a single assay. Visualization tools, statistical
methods, and machine learning techniques are all critical for turning genomic data
into knowledge. Thus was born the fi eld of bioinformatics, opening up a whole new
frontier of scientifi c inquiry. As described above, DNA microarrays enabled eluci-
dation of underlying pathways by showing which genes had similar expression pat-
terns. Genome-wide associated studies (GWAS) enable researchers to identify
genotypes that are associated with a given condition, providing hints regarding the
area of the genome where causal mutations are found. And whole genome and
whole exome sequencing have enabled fi ne-grained detection of rare, disease-
causing mutations.
It is worth noting that a signifi cant proportion of the fi eld of personalized medi-
cine, and hence the content of this chapter, is devoted to genomics, despite the fact
that the real action in biology tends to happen downstream of genes, at the level of
proteomics and metabolomics. Unfortunately, proteins and metabolites do not have
DNA and RNA's intrinsic base-pairing quality, so specifi c quantitation can be more
diffi cult. Protein microarrays have been developed with which proteins can be cap-
tured using antibodies and fl uorescent dyes. But by far the most popular approach
to quantifi cation of proteins, peptides, and metabolites is the use of mass spectrom-
etry (MS), a technology that has actually been in existence since the fi rst half of the
twentieth century. MS involves the separation, ionization, and detection of mole-
cules and their sub-components. This enables researchers to compare the observed
particle sizes with known molecular weights and thus deduce which molecules are
present in a given sample and their relative quantity. MS may be used in a discovery
fashion to detect all molecules above a minimum level of abundance in a given
sample, or specifi c molecules may be labeled with a radioisotope, enabling quanti-
fi cation of the molecule in question relative to the labeled molecule of known
concentration.