Charlottesville, August 24, 2005
By: Charlie Feigenoff
Biologists are a bit like the blind men in the parable who try to describe an elephant. Each one has an impression based on his own observations of a portion of the beast—but no one has a complete and accurate picture. University of Virginia biomedical engineer Jason Papin’s goal is to help researchers attain that big picture.
In laboratories around the world, biologists of all stripes, from ecologists to molecular scientists, have been churning out terabytes of data , with no end in sight. Ten years ago, scientists first sequenced the genome of an organism. Today, 200 organisms have been sequenced, and genomics has ushered in proteomics, the study of the proteins our genes produce.
In effect, we’ve entered an era of “high-throughput” biology. The protein and metabolic composition of an entire cell can be catalogued, the dynamic expression programs of complete genomes can be measured, and the set of interactions between cellular components can be characterized. With such an onslaught of biological data, the need for computational approaches to integrate and analyze these data has become imperative. “Each data set that scientists produce represents a different perception,” Papin says. “We’re developing quantitative frameworks that can help researchers understand how this information comes together.”
Papin’s approach is to explore this problem at different scales. He is developing techniques to model the metabolic, regulatory, and signaling networks within cells. Moving a step up, he is building a framework describing how cellular signaling networks interact with each other in multicellular systems, a project that includes fellow biomedical engineering faculty members Tom Skalak and Shayn Peirce-Cottler among others. Furthermore, he is reconstructing the pathogen-host interactions that underlie many human diseases.
In each of these cases, his goal is the stoichiometric reconstruction and analysis of large cell-signaling networks. “One advantage of the approach we are using is that we don’t require a complete set of dynamic kinetic parameters,” Papin explains. “Because biological systems are constantly adapting, these parameters change. We use constraint-based analysis that enables us to make generalizations about the solution space, rather than about individual solution points.”
Papin and his colleag ues, as they develop a reliable, overall image of how a cell behaves, are working on genome-scale analysis that can account for the inherent unknowability of biological systems. “It’s not enough to analyze individual systems or to add these systems together,” he says. “You have to understand how these systems interact.”
“To illustrate Papin’s point, take the situation of a hypothetical pharmaceutical company: The drugmaker could release a medication knowing it’s superior in a particular way to other similar products. But if the company lacks a larger, more complete view, it may not understand that the same drug has the potential to be harmful in another way.
The pioneering work that Papin and his colleagues in the U.Va. School of Engineering and Applied Science, Department of Biomedical Engineering are conducting will remedy this deficiency. By applying the techniques of systems engineering to biological systems, their work will accelerate the translation of discovery into effective treatment—and in the process position the Engineering School at the crossroads of medical progress.
