Abstract

Systems biology represents an experimental approach to biology that attempts to study biological systems in a holistic rather than an atomistic manner. Ideally this involves gathering dynamic and global data sets as well as phenotypic data from different levels of the biological information hierarchy, integrating them and modeling them graphically and/or mathematically to generate mechanistic explanations for the emergent systems properties. This requires that the biological frontiers drive the development of new measurement and visualization technologies and the pioneering of new computational and mathematical tools—all of which requires a cross-disciplinary environment composed of biologists, chemists, computer scientists, engineers, mathematicians, physicists, and physicians speaking common discipline languages. The Institute for Systems Biology has aspired to pioneer and seamlessly integrate each of these concepts.

CONTEMPORARY SYSTEMS BIOLOGY FOCUSES ON SOLVING THE CHALLENGES OF COMPLEXITY IN BIOLOGY

As the past 40 years of biological research has revealed the nature and depth of complexity in biological systems, the major challenge for biology in the 21st century is dealing with this complexity. It is clear that this complexity cannot be understood by studying genes and proteins one at a time. Rather, biological systems must be studied as an integrated whole. An attempt to explain overall systems responses (phenotypes) by detailing dynamic changes in the full spectrum of informational molecules (DNA, RNA, proteins, metabolites) and their relationships (assemblies into complex molecular machines or biological networks), must be approached by integrating these dynamics into interactive models. This is the challenge of modern systems biology.

Systems approaches to biology have actually been around for more than 100 years—with physiologists recognizing and studying homeostasis at the beginning of the 20th century and subsequent efforts of neurobiologists, developmental biologists and immunologists recognizing the central role of the integrated system. What has made contemporary systems biology possible is (i) new high-throughput measurement (and visualization) technologies (DNA sequencing, DNA arrays, protein identification and concentration determinations by mass spectrometry, measurements of protein interactions, etc.), (ii) the convergence of systems thinking about biological systems with the view that biology is basically an informational science, (iii) the complete parts lists of an organism's genes (and hence mRNAs and proteins) provided by complete genome sequence analyses, (iv) the development of computational and mathematical tools to capture, store, analyse, integrate and model large data sets and their resulting phenotypes and (v) the ability of the World Wide Web to provide widespread access to the exponentially increasing databases, data sets and the results of their analyses.

Systems biology experiments can now be performed at a scale and degree of molecular completeness that was almost inconceivable even 15 years ago. Moreover, perhaps due to the influence of the World Wide Web and the effectiveness of social networks in science, knowledge acquisition and dissemination are increasingly viewed as a collective effort in which disciplinary boundaries are dissolved, and ideas and data are more freely and openly exchanged. The hope that pervades the relatively recent explosion of systems biology-oriented organizations is that a team-oriented, interdisciplinary approach to deciphering biological complexity will pay off not only in new conceptual insights but also in practical solutions to pressing problems in healthcare, alternative energy and agriculture.

MOLECULAR SYSTEMS BIOLOGY AT THE INSTITUTE FOR SYSTEMS BIOLOGY EMBODIES THE VIEW OF BIOLOGY AS AN INFORMATIONAL SCIENCE

The Institute for Systems Biology (ISB), created in 2000, was perhaps the first organization devoted entirely to contemporary systems biology. ISB was founded on the conviction that biology is an informational science that provides a conceptual framework for dealing with biological complexity. This notion is predicated on three basic assumptions.

First, there are two fundamental types of biological information—the digital information of the genome and the environmental information from outside the genome that impinges on and modifies the digital information. Indeed, it is the digital core of knowable information in the genomes of all organisms that distinguishes biology from all of the other scientific disciplines. Systems biology is focused on understanding the interaction and dynamic integration of the digital and environmental information in the three fundamental processes of life: evolution, development and physiology. Hence, the fundamental role of systems biology is to integrate and model information from the digital genome, the environment and the organism's phenotype to understand how life unfolds on three distinct time scales.

Second, biological information is captured, processed, integrated and transferred by biological networks—interacting sets of RNAs, proteins, the control regions of genes, and small molecules—to the simple and complex molecular machines that actually execute the functions of life. Thus, an understanding of the dynamical operation of biological networks in evolution, development and physiology, (including disease) is one of the central foci of ISB. One aspect of particular interest is in understanding how protein molecular machines are constructed, changed and dismantled dynamically as they function.

Third, biological information is encoded in a multi-scale information hierarchy: DNA, RNA, proteins, interactions, biological networks, cells, tissues and organs, individuals and, finally, ecologies. The important point is that the environment impinges upon each of these levels of the hierarchy and modulates the digital informational output from the genome. Thus, systems-level investigations demand the collection of data at each relevant level of the hierarchy between the phenotypic measurement (features of the cell) and the core digital genome. The information at each of these levels should then be integrated in such a manner that the environmental modifications are identified—understanding how these modifications impact the functioning of the system is an essential first step to understanding the system's function.

ISB'S VIEW OF THE UNIQUE FEATURES OF CONTEMPORARY SYSTEMS BIOLOGY

Contemporary systems biology as an experimental science has several characteristic features:

  • Global measurements. This means that all (or most) of the elements at a particular information level should be analysed—all genes, all mRNAs, all proteins, etc. The essential point is that biology is complex and one can never guess correctly all of the elements that might participate in the functioning of a particular system—hence, one should measure dynamical changes in all the informational elements of the type measured.

  • Dynamic measurements. Biological networks and molecular machines change their structures and often their compositions during the mediation of their functions—hence, one must measure the informational elements and the phenotypes that they mediate across the developmental or physiological time dimensions in their execution of their corresponding functions.

  • Integration of different data types. As noted above, different data types must be integrated together to capture the contributions of both the digital information and the environmental information to systems functions.

  • Quantitative measurements. All measurements must be as quantitative as possible—for the ultimate integration and modeling of this information is a quantitative process.

  • Combining discovery-driven and hypothesis-driven approaches to systems studies. Systems approaches employ two general approaches—discovery and hypothesis driven. Discovery-driven approaches define the elements of a particular system at a given level of the information hierarchy without testing a specific hypothesis. The human genome project is the prototypical example. Analyses of data from discovery-driven approaches lead to specific hypotheses. In hypothesis-driven approaches a model is tested by the capture, analysis, integration and modeling of, where possible, global data sets and phenotypic responses to carefully selected perturbations. In an iterative manner, data drive hypotheses generation and testing, which in turn, lead to the generation and analyses of new data sets.

  • Modeling. The ultimate objective of systems biology is to understand the mechanistic underpinnings of particular biological systems phenomena that emerge from the integrated operation of the information components (emergent properties). The idea is to perturb the system (at the molecular and/or environmental levels), gather global data sets (of as many relevant types of information as possible) and phenotypic information, integrate these data sets and phenotypes to capture the environmental information and begin to articulate a model that explains the systems behaviour of some of its emergent properties. Then a second round of perturbations, data capture and integration, and model comparison and revision occurs— this continues until the experimental data and the model are in agreement with one another. This hybrid process is the essence of hypothesis-driven systems biology. Data space is effectively infinite—so the perturbations must be carefully chosen to reveal relevant information about the system being studied, and the models must focus on specific assumptions about the relevant variables being studied.

EXAMPLES OF SYSTEMS BIOLOGY AT ISB

Characterization of regulatory networks in Halobacterium salinarum NRC-1

Halobacterium salinarum is an archael extremophile that lives in salt marshes. Baliga and collaborators have, over the past several years, systematically combined data generation, analysis, modeling and experimental validation using perturbations of environmental factors (e.g. transition metals, oxidative stress, light, UV and gamma irradiation, etc.) to produce what they call an ‘environmental and gene regulatory influence network (EGRIN)’[1]. The current network describes how at least 72 of the 130 putative transcription factors act in conjunction with 9 environmental factors to coordinate the regulation of expression of over 1900 of its 2400 genes. Their approach was to perturb the cells in a variety of ways (genetically and environmentally), characterize the growth and/or survival phenotype, quantitatively measure steady state and dynamic changes in mRNAs, assimilate these changes into a network model that can recapitulate all observations, and, finally, to experimentally validate the hypotheses formulated from the model.

Baliga and co-workers first annotated the H. salinarum genome with sequence and structure-based approaches to provide some clues about the putative functions of nearly 90% of proteins or protein domains in this organism. They then systematically perturbed the cells with well-characterized diverse environmental factors [2–5]. They also subjected the cells to several systematic genetic perturbations. For each of these perturbations they measured changes in gene expression system-wide at a molecular level. Together these experiments resulted in a huge amount of information, whose integrated analysis required the development of new algorithms and software. For instance, the cMonkey algorithm [6] was developed specifically for the discovery of genes that are putatively co-regulated in certain environmental conditions (biclusters) by analysing data from large and diverse systems biology measurements. These biclusters were then analysed by the Inferelator algorithm [7], which discovered instances wherein individual or combinatorial changes in the concentrations of certain transcription factors and/or environmental factors temporally preceded average transcriptional changes within a given bicluster or a gene. The Inferelator algorithm (i) selects parsimonious models (i.e. minimum number of regulatory influences for each bicluster) that are predictive; (ii) explicitly includes the time dimension to discover causal influences; and (iii) models combinatorial logic i.e. interactions between environmental factors and transcription factors and between pairs of transcription factors. The output of the Inferelator algorithm is the EGRIN model.

Exploring this complex network model required the development of new software that allows users to browse the relevant raw information from the millions of data points that were used to construct the EGRIN model and to conduct new ad hoc statistical analysis. This Gaggle software [8] enables both the seamless exploration of diverse databases as well as interoperability among an equally diverse set of software tools.

With the help of this software framework, Baliga and colleagues have been able to use the EGRIN model to recapitulate and extend previously known biological insights and, crucially for systems biology, to use the network to construct new, testable hypotheses that were confirmed by genetic analysis, ChIP-chip and new perturbation experiments. Perhaps the most important demonstrated property of the EGRIN model was its capability to predict gene expression changes in new environments and genetic backgrounds [1]. Their findings suggest that the intrinsic properties of biological and environmental networks will enable the construction of similar predictive models for physiological responses of more complex organisms, and shows the way that this challenge can be approached.

Regulation of peroxisome biogenesis and function

Peroxisomes are ubiquitous intracellular organelles that play many metabolic roles in eukaryotes, including humans. The cellular physiology affected includes β-oxidation of long chain fatty acids; the synthesis of cholesterol, bile acids and plasmalogens; and the decomposition of hydrogen peroxide and superoxides [9–11]. What makes peroxisomes remarkable is that their biogenesis responds considerably to a variety of stimuli. They are induced in metazoans in response to fats, hypolipidaemic agents and non-genotoxic carcinogens, and during normal physiological processes of organismal development and cellular differentiation [12]. In yeast, peroxisomes are dramatically induced by fatty acids, or depending on the species, different carbon sources such as methanol [13]. This response is controlled at the level of transcription [14–26].

J.D.A.'s laboratory applies systems biology approaches to understand the cellular responses governing peroxisome biogenesis and function in the model organism S. cerevisiae. Essentially, this involves identifying the factors involved in the response (signalling, transcriptional regulation, biogenesis and function) and using computation and modeling approaches to understand the dynamic interplay among the factors governing each process, with the ultimate goal of obtaining quantitative and predictive models of an integrative cellular response to peroxisome induction.

Focusing on the dynamics of gene regulatory networks activated by the addition of oleate to cells, microarrays have been used to quantify the temporal changes in gene expression upon exposure of cells to fatty acids. Transcript profiling of various mutant strains, ChIP-chip analysis and the development of a novel network topology-based clustering approach enabled the characterization of the core transcriptional network that is responsive to oleic acid. The interplay among these factors has been modelled using delay differential equations and stochastic simulations and has revealed the mechanisms controlling synchrony of different cellular responses (e.g. general stress and peroxisome protein production), and the motif structures (intersecting loop structures, transcription factor oligomerization, etc.) that control aspects of the response such as non-linearity of responses, transcriptional noise, memory, etc. [27–29]. This detailed, focused approach is complementary to large-scale global analyses of networks [30–33]. Mechanisms of regulation established by these approaches are likely to establish rules that are followed by other similar networks [34, 35]. The overall goals of this work are to develop a detailed quantitative model of the peroxisome and its responses in yeast and to establish the methods that will enable systems approaches to characterize and model organelles in mammalian cells.

To identify components of the peroxisome, signalling networks and additional factors controlling the cellular response to peroxisome-inducing conditions, J.D.A. and collaborators have also exploited a library of yeast strains containing single deletions of almost every yeast gene. Each strain has been subjected to fatty acids that are metabolized by functional peroxisomes, and assayed quantitatively for their ability to metabolize the fat, and by quantitative imaging techniques for their ability to proliferate normal, functional peroxisomes [36]. These studies have revealed global, coordinated communication networks involved in the cellular response.

Proteomics technologies have also been used to characterize the peroxisome itself. In this case, classic sub-cellular fractionation procedures were combined with isotope-coded affinity-tagged (ICAT) mass spectrometry to identify proteins that enriched with peroxisomes through two different isolation procedures. The data were analysed using an expectation algorithm to identify proteins likely to be peroxisomal. Of the ∼400 proteins identified, ∼70 were determined to be biochemically identified with the organelle. Importantly, many of these proteins are shared with other subcellular structures and dynamically associate with the organelle [37].

Immunity and inflammation

The immune response consists of two sub-systems, innate immunity that provides immediate protection against infection and the adaptive response that has exquisite specificity and memory. Innate immunity is a two-edged sword; it is absolutely required for host defence but, uncontrolled, can lead to a variety of diseases including autoimmune diseases and heart disease.

Macrophages represent a keystone of the innate immune system. They detect infectious organisms directly via a plethora of receptors; they phagocytose them, and then orchestrate an appropriate host response [38]. In order to define precisely the nature of an infectious threat, the immune cell reads the molecular bar code that is displayed by each specific pathogen. The family of Toll-like receptors (TLRs) plays a key role in defining the invading micro-organism [39]. Activation of TLR4, that detects LPS, and TLR5, that detects flagellin, for example, would indicate the presence of a Gram-negative, flagellated bacterium. This precise recognition triggers a specific, highly regulated, response to the pathogen by the host. Alan Aderem and collaborators are using genomic, proteomic and computational tools to define the molecular mechanisms of the innate systems response by which macrophages integrate the complex information emanating from multiple TLRs and formulate appropriate cellular responses.

For example, flagella, whip-like structures that enable bacterial motility by propeller-like rotation, are polymers consisting of a single protein, flagellin. The Aderem lab has used genomic and proteomic approaches to identify a dual detection system for flagellin. Extracellular flagellin is detected by TLR5, which induces expression of pro-inflammatory cytokines, while cytosolic flagellin is detected through a cytoplasmic protein, Ipaf, which activates post-translational processing of cytokines. Computational analysis has given insight into the cross-talk between these sensors, and the predictions have been validated in vitro and in vivo using TLR5 null and Ipaf null mice [40]. To further understand the molecular interactions between TLR5 and flagellin, they have used protein-folding algorithms to predict the structure of TLR5 and flagellin, and have used mutational analysis to confirm the computational predictions. This led to the identification of clades of bacteria that are capable of avoiding immune detection by TLR5 [41, 42].

By characterizing variations in the genes of specific components of this system, a number of significant links between mutations in TLRs and human susceptibility to a variety of infectious diseases have been identified. The results help explain why people have differing predispositions to infectious diseases and suggest new drug targets and vaccine strategies. Highlights of these studies include the demonstration that a mutation in TLR9 leads to rapid progression from HIV infection to the development of AIDS [43]; a mutation in TLR2 leads to increased susceptibility to genital herpes, tuberculosis and leprosy [44]; mutations in TLR4 leads to enhanced susceptibility to infections in transplantation recipients and predisposes people to meningococcal sepsis [45]; and a mutation in TLR5 leads to increased incidence of Legionnaires’ disease [46].

This focus on TLRs is part of a broader programme to understand the innate immunity system's inflammatory response. For example, the Aderem group and collaborators have used computational tools to integrate microarray, ChIP-chip and proteomic data to identify a key transcription factor, ATF3, which acts as a negative regulator of the inflammatory response system. Kinetic modeling predicted the molecular mechanisms underlying its action and these predictions were validated in macrophages derived from ATF3 null mice [47]. The Aderem and Ozinsky groups are using micro-fluidics approaches to build a device capable of detecting cytokines secreted from single cells. This device could be used in the field to determine the efficacy of vaccines, and in the clinic to measure specific biomarkers indicative of systems responses.

Predictive and personalized medicine: mouse liver response to acetaminophen

Systems biology not only promises to improve our ability to perform research on complex biological systems, it also provides the foundation for a transformation of medicine into a more predictive science. One of the cornerstones of predictive medicine will be a new approach to disease diagnosis—the measurement in the blood of multiple organ-specific blood protein markers—predicated on a systems view of disease. Since disease arises as a consequence of perturbed biological networks in the diseased organ and changes the patterns of genetic information expressed by these networks, some of this changed information results in changes in protein concentrations appearing in the blood. Changes in organ-specific markers that we can detect in the blood most likely will reflect network changes in the organ expressing those markers. We have used acetaminophen toxicity in the liver of mice as a model system to test these ideas. Acetaminophen (Tylenol) causes reversible damage to liver in sub-lethal doses. It presents a histological (phenotypic) picture of the liver where increasing damage is noted up to 48 h after an acute dose and this reverses to almost normal by 72–96 h. The Hood group is using this system to explore a coordinated approach to discovering liver-specific blood biomarkers. The approach begins with generation of a list of liver-specific expressed transcripts based on the comparative analysis of high-throughput transcriptional tag profiling data sets of most organs and major tissue types in the human and mouse [48]. To see which of the liver-specific proteins encoded by these genes are secreted into the blood upon toxin exposure, they are applying a number of different proteomics discovery technologies including mass spectrometry and antibody-based assays in time-course experiments. To date, they have identified about six liver-specific blood proteins whose levels follow the histological liver changes perfectly, peaking at about 48 h after exposure and returning to normal after 72–96 h. In collaboration with the Heath group at Caltech [49], the Hood group aims to produce a micro-fluidics protein chip capable of performing multiple measurements of such biomarker candidate levels in blood from a fraction of a droplet of blood in 10 min. Chips like this can allow the typing of bloods from millions of patients in the future as predictive and personalized medicine becomes a reality. This approach will, over the next 10–20 years, lead to the replacement of our current reactive medicine (wait until one is sick to treat) to a medicine that is predictive, personalized, preventive and participatory (P4 medicine) [50–53].

THE CULTURE AND PHILOSOPHY OF ISB

Some powerful new ideas that have begun to effectively integrate biology, medicine, technology and computation/mathematics are foundational to the development of ISB as an institution. The first three of these ideas are fundamentally scientific, the rest are strategic and guide the implementation of the development of the institute.

  • The frontier problems of biology will dictate what technologies are needed and should be developed. Likewise, the creation of new data sets from these technologies should drive the development of new computational and mathematical techniques for analysing them. New technologies and computational tools in turn enable the understanding of new levels of biological complexity.

Thus, the dynamics of emerging new technologies frame the rate at which new biological insights are generated. Fundamental to ISB's operational philosophy is the idea that biology, technology development, and the development of computational and mathematical tools must all be closely integrated to generate the necessary impedance match of this broad range of skills. Close collaborations with strategic partners enhance the collaborations by bringing other necessary biological, medical and technical skills.

  • This integration of biology, technology and computation necessitates the creation of a cross-disciplinary environment—bringing into close proximity biologists, chemists, computer scientists, engineers, mathematicians, physicians and physicists in an environment where they can learn one another's languages and viewpoints, work together in teams and the non-biologists can learn relevant biology in a deep manner.

This presents enormous challenges in communication, education and planning and ISB is constantly working to learn how best to achieve this. Our view is that the design and interpretation of experiments need and benefit from all of the insights and expertise that statisticians, mathematicians, chemists, physicists and biologists have to offer. The emphasis is on collaboration, both within and between the various faculty-directed research groups. The diversity of the scientific backgrounds represented at ISB is impressive. The 28 senior research scientists listed on our web site include 1 mathematician, 2 computer scientists, 6 physicists, 1 engineer, 5 chemist/biochemists, 10 biologists and 3 MDs. To facilitate knowledge exchange and interaction, we have established a weekly discussion group in addition to the usual research seminars and laboratory research meetings. The format of this group varies with the topic and the moderator, ranging from ‘journal clubs’ to career autobiographies to presenting an idea for a new grant to open-ended discussion of topics such as research integrity or women in science or the ethics of genetic testing.

  • General facilities with cutting-edge technologies to capture biological information need to be accessible to all ISB staff members. These include DNA sequencing, arrays, genotyping, mass spectrometry-based proteomics, imaging, extensive computational capability (including algorithm development) and single-cell analyses.

Each of these facilities at ISB has a director with the relevant expertise and is overseen by a faculty member whose research is in the relevant area to assure that the facilities are kept up-to-date.

  • Biology spans a spectrum of complexity, ranging from simpler model organisms—single-celled bacteria or yeast— to more complex model organisms (mice) and ultimately humans. Biological experimentation and the probing of systems complexity are clearly easier in the simpler species. Hence, ISB develops new tools and approaches using simpler model organisms and then learns how to apply these tools to higher model organisms—and ultimately to human complexity. Accordingly, we see simple model organisms as an essential tool for driving the development of new technologies and methods.

ISB is replete with examples of technological and computational tools developed for the analysis of simpler systems (e.g. Halobacterium and yeast). Application of these methods to more complex organisms will be challenging, but in many cases, the way is now well mapped out.

  • ISB is deeply committed to an open-source philosophy—making our data and our tools readily available to the scientific community, and taking advantage of the collective input of this community to improve those tools and further analyse our data.

The effective community-wide development of the graphical visualization program, Cytoscape (http://www.cytoscape.org) [54], which emerged from ISB is an example of the power of this approach. Similarly, a new open-source community is rapidly developing around the Gaggle software program (see http://gaggle.systemsbiology.org).

  • ISB is committed to transferring knowledge to society—this includes K-12 science education, spin-off companies employing ISB's novel biological or technology, or the education of society with regard to frontier issues in science and technology.

There are several examples of these efforts of which we will cite two. First, the Center for Inquiry Science at ISB has five full-time staff and aims to reform science education at the K-12 level by developing inquiry-based materials and procedures on the one hand, and conducting intensive teacher training workshops on the other. Over the past 12 years of these programmes, school systems in Seattle and some neighbouring cities have adapted innovative and effective hands-on, inquiry-based science curriculum as a result of this programme, with a corresponding rise in scores on the science portion of state-mandated tests. Second, the Accelerator (http://acceleratorcorp.com), founded in 2003, provides a unique environment for early-stage start-up companies. The Accelerator is a partnership between ISB, five venture capital companies and a real estate development company. The idea is to identify early-stage companies for platform development or new strategies for drug discovery or creation, companies that most venture capital, companies found uninteresting and support them for 2–4 year to prove (or fail to prove) their concepts—and then aid them to move to a second round of funding. To date, six companies have been supported and three have successfully moved on to the second stage of funding.

  • ISB has as one major strategic objective facilitating the development of a new type of medicine—one that will replace the current reactive mode of medicine with a medicine that is predictive, personalized, preventive and participatory (P4 medicine).

This is an enormous challenge that involves perhaps 12 major technical challenges as well as at least eight societal challenges. One approach to attacking big problems, such as P4 medicine, while keeping ISB relatively small, is to identify strategic partners that bring to ISB the scientific, technological, computational and the medical expertise that we lack. ISB searches for these partners among academics, industry and research institutions, as well as relevant institutions in foreign countries. ISB is also beginning to explore the possibility of a P4 medical consortium—bringing together academics, industry and government to attack the technical and societal challenges of P4 medicine.

We have now been at ISB for about 8 years and have had the opportunity to pioneer and apply systems approaches to biology, systems approaches to medicine, genomic technologies, proteomic technologies and single-cell technologies. We have also developed and applied a wide variety of computational tools essential for systems approaches. ISB has established effective facilities for high-throughput data generation (genomic, proteomic and single-cell) and computational facilities for storage and analyses. The various references for these advances can be found at the ISB web site (http://www.systemsbiology.org). Two papers of the Institute demonstrate the rapid advances in systems biology at ISB. At the beginning of ISB, Hood, Ideker and their coworkers first developed the network graphical program Cytoscape [54] and used it to take a systems approach to study the galactose metabolism system in yeast [55]. Most recently, Baliga and his coworkers described for the first time the relatively complete regulatory network of a simple organism, Halobacteria [1]. In both cases, understanding the biology drove the development of powerful new computational tools for visualization, analyses and integration of large data sets and phenotypic observations. The advancements in systems thinking over the past 8 years arose from the development and application of systems approaches to frontier challenges of biology and medicine by various teams drawn from the ISB faculty and research staff.

Key Points

  • Systems biology is predicated on the belief that biology is an informational science, that there are two fundamental types of biological information—digital genome and environmental—and that biological information is handled by biological networks.

  • Systems biology can only thrive in a cross-disciplinary environment and culture as the complexity of biology requires that the frontier problems of biology drive the development of new technologies and the data generated from these in turn force the development of new computational and mathematical tools.

  • Systems biology employs both discovery approaches to define the elements of systems and hypotheses-driven approaches to test mechanistic models.

  • Systems approaches lead to the formulation of big problems such as P4 medicine—and an effective approach to big problems requires formulating strategic partnerships from academia and industry to bring into focus all of the necessary requisite talents and skills to effectively move forward.

  • The development of new measurement and visualization technologies for biomolecules and single cells as well as powerful new computational and mathematical tools for capturing, storing, validating, analysing, integrating, visualizing and finally modeling, either graphically or mathematically, will be the key to the future growth of systems approaches to biology, medicine, agriculture, bio-energy, nutrition and the many other areas that systems approaches will revolutionize.

Acknowledgements

The authors would like to thank the entire faculty and staff of ISB each for contributing in unique ways to the vision and accomplishments of systems biology at ISB. We thank Nitin Baliga and Alan Aderem for the project descriptions of Halobacterium and innate immunity, respectively. The funding for the work has been provided by National Institutes of Health (R01 GM067228, U54 RR0222200 to J.D.A., P50 GM076547 to J.D.A., L.R. L.H.) and Department of Defense (W911SR-07-C-0101 to L.H.).

References

Bonneau
R
Facciotti
MT
Reiss
DJ
, et al.  . 
A predictive model for transcriptional control of physiology in a free living cell
Cell
 , 
2007
, vol. 
131
 (pg. 
1354
-
65
)
Schmid
AK
Reiss
DJ
Kaur
A
, et al.  . 
The anatomy of microbial cell state transitions in response to oxygen
Genome Res
 , 
2007
, vol. 
17
 (pg. 
1399
-
413
)
Whitehead
K
Kish
A
Pan
M
, et al.  . 
An integrated systems approach for understanding cellular responses to gamma radiation
Mol Syst Biol
 , 
2006
, vol. 
2
 pg. 
47
 
Kaur
A
Pan
M
Meislin
M
, et al.  . 
A systems view of haloarchaeal strategies to withstand stress from transition metals
Genome Res
 , 
2006
, vol. 
16
 (pg. 
841
-
54
)
Baliga
NS
Bjork
SJ
Bonneau
R
, et al.  . 
Systems level insights into the stress response to UV radiation in the halophilic archaeon Halobacterium NRC-1
Genome Res
 , 
2004
, vol. 
14
 (pg. 
1025
-
35
)
Reiss
DJ
Baliga
NS
Bonneau
R
Integrated biclustering of heterogeneous genome-wide datasets for the inference of global regulatory networks
BMC Bioinformatics
 , 
2006
, vol. 
7
 pg. 
280
 
Bonneau
R
Reiss
DJ
Shannon
P
, et al.  . 
The Inferelator: an algorithm for learning parsimonious regulatory networks from systems-biology data sets de novo
Genome Biol
 , 
2006
, vol. 
7
 pg. 
R36
 
Shannon
P
Reiss
D
Bonneau
R
, et al.  . 
Gaggle: an open-source software system for integrating bioinformatics software and data sources
BMC Bioinformatics
 , 
2006
, vol. 
7
 pg. 
176
 
Lazarow
PB
De Duve
C
A fatty acyl-CoA oxidizing system in rat liver peroxisomes; enhancement by clofibrate, a hypolipidemic drug
Proc Natl Acad Sci USA
 , 
1976
, vol. 
73
 (pg. 
2043
-
6
)
Lazarow
PB
Fujiki
Y
Biogenesis of peroxisomes
Annu Rev Cell Biol
 , 
1985
, vol. 
1
 (pg. 
489
-
530
)
van den Bosch
H
Schutgens
RB
Wanders
RJ
, et al.  . 
Biochemistry of peroxisomes
Annu Rev Biochem
 , 
1992
, vol. 
61
 (pg. 
157
-
97
)
Weller
S
Gould
SJ
Valle
D
Peroxisome biogenesis disorders
Annu Rev Genomics Hum Genet
 , 
2003
, vol. 
4
 (pg. 
165
-
211
)
Veenhuis
M
van der Klei
IJ
Titorenko
V
, et al.  . 
Hansenula polymorpha: an attractive model organism for molecular studies of peroxisome biogenesis and function
FEMS Microbiol Lett
 , 
1992
, vol. 
79
 (pg. 
393
-
403
)
Kal
AJ
van Zonneveld
AJ
Benes
V
, et al.  . 
Dynamics of gene expression revealed by comparison of serial analysis of gene expression transcript profiles from yeast grown on two different carbon sources
Mol Biol Cell
 , 
1999
, vol. 
10
 (pg. 
1859
-
72
)
Rachubinski
RA
Fujiki
Y
Lazarow
PB
Isolation of cDNA clones coding for peroxisomal proteins of Candida tropicalis: identification and sequence of a clone for catalase
Biochim Biophys Acta
 , 
1987
, vol. 
909
 (pg. 
35
-
43
)
Smith
JJ
Marelli
M
Christmas
RH
, et al.  . 
Transcriptome profiling to identify genes involved in peroxisome assembly and function
J Cell Biol
 , 
2002
, vol. 
158
 (pg. 
259
-
71
)
Devchand
PR
Keller
H
Peters
JM
, et al.  . 
The PPARalpha-leukotriene B4 pathway to inflammation control
Nature
 , 
1996
, vol. 
384
 (pg. 
39
-
43
)
Issemann
I
Green
S
Activation of a member of the steroid hormone receptor superfamily by peroxisome proliferators
Nature
 , 
1990
, vol. 
347
 (pg. 
645
-
50
)
Issemann
I
Prince
RA
Tugwood
JD
, et al.  . 
The peroxisome proliferator-activated receptor:retinoid X receptor heterodimer is activated by fatty acids and fibrate hypolipidaemic drugs
J Mol Endocrinol
 , 
1993
, vol. 
11
 (pg. 
37
-
47
)
Lee
S
Pineau
ST
Drago
J
, et al.  . 
Targeted disruption of the alpha isoform of the peroxisome proliferator-activated receptor gene in mice results in abolishment of the pleiotropic effects of peroxisome proliferators
Mol Cell Biol
 , 
1995
, vol. 
15
 (pg. 
3012
-
22
)
Lefebvre
P
Chinetti
G
Fruchart
JC
, et al.  . 
Sorting out the roles of PPAR alpha in energy metabolism and vascular homeostasis
J Clin Invest
 , 
2006
, vol. 
116
 (pg. 
571
-
80
)
Reddy
JK
Hashimoto
T
Peroxisomal beta-oxidation and peroxisome proliferator-activated receptor alpha: an adaptive metabolic system
Annu Rev Nutr
 , 
2001
, vol. 
21
 (pg. 
193
-
230
)
Gurvitz
A
Rottensteiner
H
The biochemistry of oleate induction: transcriptional upregulation and peroxisome proliferation
Biochim Biophys Acta
 , 
2006
, vol. 
1763
 (pg. 
1392
-
402
)
Platta
HW
Erdmann
R
Peroxisomal dynamics
Trends Cell Biol
 , 
2007
, vol. 
17
 (pg. 
474
-
84
)
Rottensteiner
H
Wabnegger
L
Erdmann
R
, et al.  . 
Saccharomyces cerevisiae PIP2 mediating oleic acid induction and peroxisome proliferation is regulated by Adr1p and Pip2p-Oaf1p
J Biol Chem
 , 
2003
, vol. 
278
 (pg. 
27605
-
11
)
Koerkamp
MG
Rep
M
Bussemaker
HJ
, et al.  . 
Dissection of transient oxidative stress response in Saccharomyces cerevisiae by using DNA microarrays
Mol Biol Cell
 , 
2002
, vol. 
13
 (pg. 
2783
-
94
)
Ramsey
SA
Smith
JJ
Orrell
D
, et al.  . 
Dual feedback loops in the GAL regulon suppress cellular heterogeneity in yeast
Nat Genet
 , 
2006
, vol. 
38
 (pg. 
1082
-
7
)
Ratushny
AV
Ramsey
SA
Roda
O
, et al.  . 
Modeling the dynamic behavior of an oleate-responsive gene regulatory network in Saccharomyces cerevisiae
In: The Eighth International Conference on Systems Biology
 , 
2007
Longbeach, CA
Smith
JJ
Ramsey
SA
Marelli
M
, et al.  . 
Transcriptional responses to fatty acid are coordinated by combinatorial control
Mol Syst Biol
 , 
2007
, vol. 
3
 pg. 
115
 
Ren
B
Robert
F
Wyrick
JJ
, et al.  . 
Genome-wide location and function of DNA binding proteins
Science
 , 
2000
, vol. 
290
 (pg. 
2306
-
9
)
Lee
TI
Rinaldi
NJ
Robert
F
, et al.  . 
Transcriptional regulatory networks in Saccharomyces cerevisiae
Science
 , 
2002
, vol. 
298
 (pg. 
799
-
804
)
Harbison
CT
Gordon
DB
Lee
TI
, et al.  . 
Transcriptional regulatory code of a eukaryotic genome
Nature
 , 
2004
, vol. 
431
 (pg. 
99
-
104
)
Luscombe
NM
Babu
MM
Yu
H
, et al.  . 
Genomic analysis of regulatory network dynamics reveals large topological changes
Nature
 , 
2004
, vol. 
431
 (pg. 
308
-
12
)
Longabaugh
W
Bolouri
H
Understanding the dynamic behavior of genetic regulatory networks by functional decomposition
Curr Genomics
 , 
2006
, vol. 
7
 (pg. 
333
-
41
)
Alon
U
An Introduction to Systems Biology: Design Principles of Biological Circuits
 , 
2006
London
Taylor & Francis
Niemisto
A
Selinummi
J
Saleem
R
, et al.  . 
Extraction of the number of peroxisomes in yeast cells by automated image analysis
Conf Proc IEEE Eng Med Biol Soc
 , 
2006
, vol. 
1
 (pg. 
2353
-
6
)
Marelli
M
Smith
JJ
Jung
S
, et al.  . 
Quantitative mass spectrometry reveals a role for the GTPase Rho1p in actin organization on the peroxisome membrane
J Cell Biol
 , 
2004
, vol. 
167
 (pg. 
1099
-
112
)
Ravetch
J
Aderem
A
Phagocytic cells
Immunol Rev
 , 
2007
, vol. 
219
 (pg. 
5
-
7
)
Ozinsky
A
Underhill
DM
Fontenot
JD
, et al.  . 
The repertoire for pattern recognition of pathogens by the innate immune system is defined by cooperation between toll-like receptors
Proc Natl Acad Sci USA
 , 
2000
, vol. 
97
 pg. 
1
 
Miao
EA
Andersen-Nissen
E
Warren
SE
, et al.  . 
TLR5 and Ipaf: dual sensors of bacterial flagellin in the innate immune system
Semin Immunopathol
 , 
2007
, vol. 
29
 (pg. 
275
-
88
)
Andersen-Nissen
E
Smith
KD
Bonneau
R
, et al.  . 
A conserved surface on Toll-like receptor 5 recognizes bacterial flagellin
J Exp Med
 , 
2007
, vol. 
204
 (pg. 
393
-
403
)
Andersen-Nissen
E
Smith
KD
Strobe
KL
, et al.  . 
Evasion of Toll-like receptor 5 by flagellated bacteria
Proc Natl Acad Sci USA
 , 
2005
, vol. 
102
 (pg. 
9247
-
52
)
Bochud
PY
Hersberger
M
Taffé
P
, et al.  . 
Polymorphisms in Toll-like receptor 9 influence the clinical course of HIV-1 infection
AIDS
 , 
2007
, vol. 
21
 (pg. 
441
-
6
)
Bochud
PY
Magaret
AS
Koelle
DM
, et al.  . 
Polymorphisms in TLR2 are associated with increased viral shedding and lesional rate in patients with genital herpes simplex virus Type 2 infection
J Infect Dis
 , 
2007
, vol. 
196
 (pg. 
505
-
9
)
Thuong
NT
Hawn
TR
Thwaites
GE
, et al.  . 
A polymorphism in human TLR2 is associated with increased susceptibility to tuberculous meningitis
Genes Immun
 , 
2007
, vol. 
8
 (pg. 
422
-
8
)
Hawn
TR
Berrington
WR
Smith
IA
, et al.  . 
Altered inflammatory responses in TLR5-deficient mice infected with Legionella pneumophila
J Immunol
 , 
2007
, vol. 
179
 (pg. 
6981
-
7
)
Gilchrist
M
Thorsson
V
Li
B
, et al.  . 
Systems biology approaches identify ATF3 as a negative regulator of Toll-like receptor 4
Nature
 , 
2006
, vol. 
441
 (pg. 
173
-
8
)
Lin
B
White
JT
Lu
W
, et al.  . 
Evidence for the presence of disease-perturbed networks in prostate cancer cells by genomic and proteomic analyses: a systems approach to disease
Cancer Res
 , 
2005
, vol. 
65
 (pg. 
3081
-
91
)
Bailey
RC
Kwong
GA
Radu
CG
, et al.  . 
DNA-encoded antibody libraries: a unified platform for multiplexed cell sorting and detection of genes and proteins
J Am Chem Soc
 , 
2007
, vol. 
129
 (pg. 
1959
-
67
)
Hood
L
Heath
JR
Phelps
ME
Lin
B
Systems biology and new technologies enable predictive and preventative medicine
Science
 , 
2004
, vol. 
306
 (pg. 
640
-
3
)
Heath
JR
Phelps
ME
Hood
L
NanoSystems biology
Mol Imaging Biol
 , 
2003
, vol. 
5
 (pg. 
312
-
25
)
Weston
AD
Hood
L
Systems biology, proteomics, and the future of health care: toward predictive, preventative, and personalized medicine
J Proteome Res
 , 
2004
, vol. 
3
 (pg. 
179
-
96
)
Price
ND
Edelman
LB
Lee
I
, et al.  . 
Ginsburg
G
Willard
H
Systems biology and the emergence of systems medicine
Genomic and Personalized Medicine: From Principles to Practice
 , 
2009
Amsterdam
Elsevier
Shannon
P
Markiel
A
Ozier
O
, et al.  . 
Cytoscape: a software environment for integrated models of biomolecular interaction networks
Genome Res
 , 
2003
, vol. 
13
 (pg. 
2498
-
504
)
Ideker
T
Thorsson
V
Ranish
JA
, et al.  . 
Integrated genomic and proteomic analyses of a systematically perturbed metabolc network
Science
 , 
2001
, vol. 
292
 (pg. 
929
-
34
)