Biology Reference
In-Depth Information
complex biological systems: understanding what indi-
vidual genes or proteins do does not tell us how the bio-
logical systems in which they operate function. An
engineer would connect the radio parts into their circuits
and come to understand how the circuits worked indi-
vidually and then collectively to convert radio into sound
waves. So it is with living organisms: they employ bio-
logical circuits or networks to manage biological infor-
mation and convert it into phenotype and function, and to
understand these we must understand the dynamics of
biological networks in handing information.
To generate holistic data systems biology (and
therefore systems medicine) has three central elements.
(1) It is hypothesis driven, where a model (which is
a formally structured, precise and potentially complex) is
formulated from existing data. Hypotheses from model
predictions are then tested with systems perturbations
and the high-throughput acquisition of data. The data
(and metadata) are then reintegrated back into the model
with appropriate modifications, and this process is
repeated iteratively until new predictions from theory
and experimental data are in agreement. (2) It is based on
high-throughput data that should be (i) global (compre-
hensive), (ii) generated from different multi-scale data
types (e.g., DNA, RNA, protein, metabolites, interac-
tions, etc.), (iii) used to monitor networks dynamically,
(iv) employed to provide deep insight into biology, and
(v) integrated using proper statistics and bioinformatics
to handle the enormous signal-to-noise problems. (3)
Models may be descriptive, graphical or mathematical as
dictated by the amount of available data, but theymust be
predictive. For medical use, predictions made must be
actionable and useful for treating patients.
Boosting signal-to-noise in complex biology is
essential for deciphering complexity. To reduce noise
and to enhance statistical power, biologists have lever-
aged two fundamental ideas: filters and integrators [27] .
Filters are used to winnow down the number of candi-
dates based on the biological assumptions about
complexity (e.g., modularity, hierarchical organization,
complexity arising from evolution and inheritance).
Integrators leverage the availability of complementary
data of genome, transcriptome, miRNAome, proteome,
metabolome, and interactome. Successful application of
these strategies in disease will lead to a transformational
understanding of disease and therapeutics.
The framework for approaching these studies in
a holistic way is a systems approach to disease. As dis-
cussed above, the key idea is that disease arises as
a consequence of the perturbation of one or more bio-
logical networks in the relevant organ. This perturbation
alters the information the network encodes in a dynamic
manner that changes during the progression of the
disease (e.g., changing levels of mRNAs, miRNAs, or
even proteins)
and these altered levels explain the
pathophysiology of the disease and provide new insights
into diagnosis and therapy.
e
A systems approach to a neurodegenerative disease in
mice.Wewill illustrate this holistic systems approach as it
applies to neurodegenerative disease (prion disease) in
mice. This disease is initiated by the injection of 'infec-
tious prion proteins' into the brains of mice. An important
point is that we know precisely when the disease is
initiated (at injection), allowing us to follow the dynamics
of the disease process from initiation to termination. We
analyzed the brain transcriptomes of the infected mice at
10 time points across the approximately 22 weeks of
disease progression, in addition to the transcriptome of
their healthy littermates. This procedure identified 7400
differentially expressed genes (DEGs), which represented
a staggering signal-to-noise problem.
There are two types of noise: technical noise,which
comes from the instrumentation/procedures for handling
the data, and biological noise, which arises from biolog-
ical processes other than neurodegeneration contributing
to the phenotypic measurements. When you measure any
aspect of phenotype in an organism, often those pheno-
types are the sum of a number of different biologies.
Hence one must use a deep understanding of biology to
subtract from the biology of interest (neurodegeneration)
the signals resulting from the other biologies.
To overcome the noise we carried out this study in
eight different inbred-strain/prion-strain combinations of
infected mice. With more data and a deep biological
understanding of the disease process we were able to
subtract away noise. For example, in the double-
knockout mouse for the prion gene, after injection with
infectious prions the animals never develop the disease.
Thus, any changes in the brain transcriptomes of these
animals were irrelevant to the prion neurodegeneration
response and could be subtracted away. With seven
additional subtractions, we identified a core of about 333
differentially expressed genes that encoded the basic
prion neurodegeneration process. We mapped these
DEGs on to four major biological networks of the prion
disease process that had been defined by serial histopa-
thology of the diseased brains. We then integrated the
transcriptome data with (1) serial brain histopathological
analyses of these animals, (2) serial sagittal brain sections
stained for infectious prions, (3) clinical signs of the
disease and (4) blood biomarker analyses. Figure 23.6
illustrates one of the major dynamically changing
networks (prion replication and accumulation).
We drew the following conclusions from this study:
(1) The disease starts with one or a few networks being
Search WWH ::




Custom Search