Center for Cell Decision Processes (MIT)
Principal investigator: Peter Sorger, Ph.D., Massachusetts Institute of Technology and Harvard Medical School
Other key investigators: Douglas Lauffenburger (MIT), Scott Manalis (MIT), Elba Serrano (New Mexico State University).
The goal of the MIT Center for Cell Decision Processes us to understand how mammalian signaling networks regulate life-death decisions in cells and tissues and to apply this understanding to the design and use of therapeutic drugs. We use a modify-measure-mine-model approach involving iterative application of mathematical modeling and experimental measurement that yields quantitative understanding of cellular biochemistry in actual cells. It is a truism that detailed knowledge of the parts of a complex process such as signal transduction provides only limited insight into the operation of the process as a whole. Mathematics is therefore important to systems biology because only formal models have the power to describe the temporal and spatial dynamics of large sets of interacting proteins. Experiments are critical because the validity of mathematical models can only be established only by rigorous empirical analysis.
A primary focus of CDP research is to dissect the networks that mediate extrinsic apoptosis provoked by of TRAIL, TNF and FAS receptors and the pro-survival networks downstream of EGF, IGF and insulin receptors. T-cell receptors and, cell-cell signaling in immune cells in general, are a second area of interest. We apply a wide spectrum of modeling methods from physicochemical to logical to data-driven and experimental measurements from biophysical studies, mass spectrometry of cell extracts, live-cell imaging and analysis of living cells in vivo. CPD investigators are actively developing new methods for assembling models and for linking models to experimental data. For example, we were among the first to apply partial least squares regression (PLSR) and fuzzy logic to mammalian signaling and have recently pioneered methods for calibrating logic-based and kinetic models to diverse experimental data. The net result is a series of increasingly sophisticated mathematical representations of signal transduction in both normal and diseased cells, with a particular emphasis on cancer and immune dysfunction.
The CDP Center is also dedicated to advancing the interface between biology and chemistry, engineering, computation and physics; half its members were trained as engineers and the other half as natural scientists. CDP has active education and outreach programs that address multiple career stages from high school to faculty. The Center runs an annual international conference on the Systems Biology of Human Disease and managed the Council for Systems Biology in Boston (CSB2) which helps to coordinate systems biology programs throughout Eastern New England. To increase the breadth and depth of systems biology research we host sabbatical visitors from universities serving minority and disadvantaged student communities. These sabbatical visitors have taken Center faculty in new directions as well: Elba Serrano (2007-2008) leads an effort to introduce organ culture into CDP and Shubha Govind (2008-2009) focuses on relationships between chronic inflammation and tumor development.
CDP alumni hold prestigious positions in academic institutions across the U.S. including University of Virginia, Georgia Tech and Harvard Medical School and several others head industrial laboratories or have started their own companies. The CDP Informatics core is developing a suite of interoperable software tools designed to make modeling more widely available to investigators outside the Center. Some tools implement new approaches to modeling such as data-optimized Boolean logic or Bayesian parameter estimation whereas others are part of the SB-Pipeline project. Through SB-Pipeline we are implementing an approach to the management of complex experimental data that does not use familiar but highly inflexible relational databases. Instead SB-Pipeline is based on semantic web concepts and manipulation of structured data arrays. Its streamlined architecture promises to make model-driven analysis of biochemical and pharmacological data accessible to a wide audience of non-specialists, particularly those involved in translational cancer research.
Linked media including by Peter Sorger is licensed under a Creative Commons Attribution-Share Alike 3.0 United States License.