Systems biology
How Will Big Pictures Emerge From a Sea of Biological Data?
Every year since the article with this title appeared, the question becomes more compelling. We are now accumulating information about biological sequences, structures, and interactions faster than we have the power to make sense of them. For hundreds of years prior to this, practical considerations coerced biological research into reductionism. There are simply too many components in a biological system for a biologist to examine the whole picture with the tools formerly available. Over the past decade this has rapidly changed as biological information has become cheap and plentiful due to the advent of high-throughput tools, making it possible for the first time to ask questions on time and length scales that were previously intractable. The relaxation of the practical limitations on systems-level analysis has also brought a change in the philosophy of how we regard biology, moving towards a holistic method of research and interpretation.
This places Systems biology in stark contrast to traditional biological research, and for good reason. In the words of Denis Noble, "Systems biology is about putting together rather than taking apart, integration rather than reduction. It starts with what we have learned from the reductionist approach; and then it goes further." This shift from reductionism is essential, for as we know from studying complex systems, the whole is greater than the sum of the parts. With this new approach we are able to explore scientific territory that has previously been untouched due to physical impossibility and philosophical differences.
The complexity of the tangled web of nonlinear interactions between genes, proteins, and the environment necessitates the development of simplified models to illuminate biological functions. Merely generating networks of interactions is not enough, providing us with far too much information in a single view without emphasizing the important features of the map. When we use Google maps and look at a picture of the United States it doesn't show us every city, we would never see Evanston, Illinois being shown at that level of detail. Only large and recognizable cities are shown to help us orient the map. Once we zoom in other smaller cities and features become visible, giving us more relevant information in a manner that is usable. Simply generating networks without any type of analysis or visualization is akin to showing a map of the United States with every state, city, and town marked on it. For this reason we emphasize the development of simple biological models that provide insight into the function of biological molecules and dynamics of the network, answer the question in a manner that is amenable to biological interpretation, and are valuable to the community at large by emphasizing usability.
It is essential that we work closely with experimental biologists to develop rational models and assumptions and validate them with empirical data. To this end, we collaborate with other laboratories at Northwestern University (The Morimoto and Carthew Laboratories) and participate in the Chicago Consortium for Systems Biology, which all have an emphasis on experimental benchwork. Our current projects encompass a broad range of topics within Systems biology, ranging from the dynamics of the eukaryotic stress response to developing a process-motivated model for handling microarray data to modeling development of the drosophila eye.