Biology Reference
In-Depth Information
complexity: (1) viewing medicine as an informational
science, (2) creating a cross-disciplinary infrastructure in
which to implement systems medicine, (3) employing
experimental systems approach to disease that are holistic
and integrative, (4) driving the development of new tech-
nologies that permit the exploration of new dimensions of
patient data space, and (5) developing new analytical tools,
both computational and mathematical, for capturing, vali-
dating, storing, mining, integrating and finally modeling
data so as to convert data into knowledge.
These five pillars of systems medicine permit biological
complexity to be deciphered by providing a path forward
for both generating large amounts of data, integrating and
modeling these data in ways that reduce noise and delineate
biological mechanisms. They create
obvious. One cannot follow the disease from initiation
to the end in humans; one cannot usually easily sample
the diseased tissue at multiple different time points; nor
can one experimentally perturb the system with envi-
ronmental signals.
The need for systems dynamics data to deal with
noise and create models emphasizes the importance of
experimental animal disease models where the starting
point of the disease process can be known (e.g., by
genetic activation of the disease process or the experi-
mental initiation of disease such as an infection) and the
dynamics followed until death. The key point is that
animal models must closely mimic their human coun-
terpart diseases. Scientists must clearly identify those
aspects of the disease-perturbed systems that are
orthologous to human disease and those that are unique
to the animal
the
conceptual
framework for converting data into knowledge.
and use the former for gaining dynam-
ical insights into human disease. When the disease
process is translated into network dynamics, deter-
mining orthology between the animal model and human
disease becomes much simpler. Indeed, one can draw
inferences from model organism disease-perturbed
networks that are orthologous to their human counter-
parts and ignore the disease-perturbed networks that are
not orthologous. This approach enables animal studies to
be powerfully informative about human disease.
2.
Our belief is that a special infrastructure is required for
practicing systems medicine. This belief is driven by the
conviction that leading-edge biology must drive the
development of new high-throughput technologies to
explore new dimensions of patient data space. The data
arising from these technologies in turn require the
pioneering of new analytical tools for the integration
and modeling of diverse data types. We have termed this
the 'holy trinity' of biology
e
1.
Systems medicine views medicine as an informational
science, providing an intellectual framework for dealing
with complexity. Fundamentally, there are two types of
biological information: the digital information of the
genome and the environmental signals that come from
outside the genome. Together these two types of
information are integrated in the individual organism
(e.g., a human) to produce its phenotype, healthy or
diseased. These two types of information and the
phenotypes they produce are connected through bio-
logical networks that capture, transmit, integrate signals
and then pass the information to molecular machines
that execute the functions of life. It is the dynamics of
networks and molecular machines that constitute
a major focus of systems studies. The 'network of
networks' adds yet another multiscale challenge to
organizing and integrating information (
Figure 23.4
).
As noted above, systems medicine postulates that
disease arises from disease-perturbed networks (per-
turbed by genetic changes and/or environmental
signals). Altered molecular machinery encoded by the
disease-perturbed networks leads to the pathophysi-
ology of the disease. Thus following the dynamics of
the disease-perturbed networks gives deep insights into
disease mechanisms and provides a powerful tool for
dealing with the signal to noise challenges of big data
sets. The utility of this approach has been demonstrated
in two mouse models
e
mouse neurodegeneration
(prion infection)
[16]
and glioblastoma
biology drives tech-
e
nology drives analytical tools
and integrated them
together to revolutionize our understanding of medicine
(
Figure 23.5
).
This approach requires a cross-disciplinary environ-
ment where biologists, chemists, computer scientists,
engineers, mathematicians, physicists and physicians all
learn to speak the languages of the other disciplines and
work together in biology-driven teams to achieve this holy
trinity. To be effective, this cross-disciplinary environ-
ment requires the 'democratization' of data generation
and data analysis tools; that is, it is essential to make these
tools accessible to all individual scientists so that theymay
carry out either big science or small science projects.
Thus the infrastructure of systems medicine consists
of both the instrumentation to generate data for the diverse
'omic' technologies (genomics, proteomics, metab-
olomics, interactomics, cellomics, etc.) and a culture that
encourages scientists to learn to speak the languages of
e
from mice
genetically engineered in a combinatorial manner with
oncogenes and tumor suppressors). The prion model of
neurodegenerative is discussed later in this chapter.
To obtain the necessary information for systems
medicine it is also critical to integrate and model the
many diverse data types, including from animal models,
that follow disease progression. The reasons for this are
e
