Exploring the machinery of life
How does systems biology differ from the conventional approach to biology?
Wiley: Systems biology attempts to understand how an organism works from an overall perspective. We're building this understanding starting from the molecular level. Conventional molecular biology is devoted to learning more about the parts--describing the structure, properties and interactions of the individual components. Systems biology looks at how the components work together as a system. Using the analogy of an automobile, systems biology is like understanding the different parts of a car and how the different parts work together to achieve motion.
Colson: It all starts with the word "system." Only recently have we had advances necessary to create models that could potentially be predictive at the level of a biological system. Systems could refer to a cell, to multiple cells, an organ or an organism. At all of those levels, the degree of complexity is substantial. We're really stretching the resources of biology, computer science and instrumentation to make progress in this area.
Rodland: To me, it is the opposite of reductionist biology. For a long time we made a lot of progress in biology by looking at one protein at a time and learning everything there was to know about it in isolation. Now we have to put those pieces together.
Dixon: As we integrate all of the data to describe a biological system, we also want to do a better job of predicting biological behavior. We need to see how proteins work together to form networks to make a living cell.
What problems can this approach eventually solve?
Rodland: My bias is health-related research. The only way to really understand the human body is from a systems biology standpoint. We learned from a reductionist approach to cancer research that cancer is not one disease with one cause. It is hundreds of diseases with hundreds of causes, not based on one gene or one cell, but dozens of genes acting in concert.
Wiley: I expect systems biology to allow us to both understand and productively manipulate biological systems. At this point, new drugs are tested on thousands of people. If they don't work, researchers go back to the drawing board because they don't know what needs to be changed. If we can make biology predictive, we can engineer cells in productive ways. You could help solve problems anywhere a living cell is involved--treating disease, understanding the implications of pollutants, removing carbon from the atmosphere and developing clean energy supplies.
Brockman: From my perspective, defining networks of carbon, energy and information flow between microorganisms will improve our ability to help solve environmental problems. In the long term, it will improve the predictive understanding of how microbial communities respond to and buffer changes in the natural environment. For example, we could influence microbial communities to subtly change the relative rates of chemical reactions relating to sulfur or carbon as a powerful way to reduce global warming. Systems biology also will speed the ability to carry out improved and new chemical processes using microbial communities in engineered environments such as bioreactors. For example, we could establish communities of microbes with the optimal structure and function to process waste streams, to produce chemicals from low-value raw materials or generate hydrogen as a source of clean energy for fuel cells.
Why is DOE interested?
Colson: DOE has big issues to solve. You can't address these problems one molecule at a time. We have to understand how systems work together if we are going to address global climate change or clean up complex contaminated environments or produce energy using microorganisms. We now have a much more powerful way of going after these issues.
Dixon: You can't really understand how to clean up contaminated sites if you don't fully understand the risk of low-dose radiation damage. It's not just one person doing experiments in a lab that will solve the real questions of systems biology, it's teams of scientists from many different disciplines working together. DOE and the national labs have the team approach and the tools--the big computers and the big instruments--to do the large-scale science necessary for systems biology.
How does PNNL intend to contribute to DOE's Genomes to Life program?
Wiley: PNNL would like to be a primary driver. We would like to help define the experimental systems, the computational infrastructure and the instrumentation necessary to do systems biology. We feel we have the ideas to help DOE become preeminent in this area. We partnered with Oak Ridge National Laboratory to submit a proposal to the Genomes to Life program to do global analysis of protein complexes in cells. In our second proposal, we put forward a systems biology and computational approach for understanding complex microbial communities.
Brockman: One aspect of our plan deals with the tight integration of theory, experimentation and measurements. Classically, science always has had these three components, but now the idea is to go through the process very rapidly with the help of computers. Computers can design a sequence of experiments to effectively test ambiguous hypotheses. A computer can show how and where models are inadequate and help define parameters to go into that part of the model.
How are advancements in computational science and instrumentation critical to systems biology research?
Brockman: It's really clear that the reason biological knowledge has exploded in the past decade is the ability to make measurements and develop instrumentation to take those measurements. The number of new instruments and new types of measurements is increasing almost exponentially. It's almost incomprehensible. In the future, we'll depend more and more on computers and computation to deal with the flood of data from these new instruments.
Rodland: At any given time, the ability to answer questions is dependent upon the tools available to study the question.
Why do you feel PNNL can be a leader in systems biology?
Wiley: We made a decision to do it. We made the investment and have the commitment behind the decision. We have real expertise in this area and we're building the teams to get the job done. The scientific community has known how to do this for 10 years. It simply requires long-term investments in instruments and computational infrastructure and faith in people.
Colson: PNNL has made substantial investments in new ventures for years. One brought the Environmental Molecular Sciences Laboratory (EMSL) here and it has some of the world's best instrumentation. Investments in systems biology allow us to go forward with DOE and other clients.
Dixon: We are one of the few DOE labs that understands how to run a user facility that is not built around a single instrument. EMSL allowed us to stick our foot into the waters of biology and now we can make more of a leap because of the investments we've made.
Wiley: And the organizational structure here is better for interdisciplinary research. Universities are bound more by departmental organizational structures. To get the engineering department to work with the life sciences department at a university is very difficult. The culture and structure here make it far easier to begin making progress.
Rodland: That's a real strength of PNNL. The communication and conversations between different components--computational science, optical experts, proteomics experts, toxicologists and biologists--have been going on here for years.
Where do you see this field in 10 years?
Dixon: It's my feeling that biology will look a lot more like high-energy physics. I think it will revolutionize how we solve biological problems, but it's going to be an interesting road. The transition will require individual scientists from disciplines that have been very individualistic to work together as a team. Biologists and chemists have never been known for working together to solve a problem, but physicists have.
Rodland: There also will be a new niche in data mining and data scavenging because so much data is generated by the systems biology approach. It will involve a whole new discipline of sorting, formatting and looking at data to figure out what it all means.
Dixon: Biology will become more of an information science. Advances will be made in gathering data and then synthesizing the information to tell the "biology story." How is the biological system working? What are the conceptual models that will allow us to design a new drug or modify an organism to do what we want it to do? These models will be on our computers, based on experimental data and validated with experimental data so we can take giant steps forward.
Brockman: In 10 years, I see systems biology as a dominant force in how biology is done. There is a tremendous opportunity for DOE to take a leadership role by drawing upon its ongoing investments in technology development and computation and the multidisciplinary scientific teams within the DOE national laboratory system.
For more detailed responses to these and other questions, see www.pnl.gov/breakthroughs/sum02/systemsbio.html