Prediction: Within five years, the fruits of proteomics will begin enabling biologists to truly understand the cell as a network system of interactions. It’s a forecast that would be no surprise coming from a proteomics guru. But this one was made by a bioinformatics software vendor.
Addressing the BioData 2002 conference in Geneva last week, Lion Bioscience CEO Friedrich von Bohlen branched out into proteomics as the subject of his keynote address. “Proteomics, alone but also in combination with expression profiles and the genome, will drive [biology] towards a connected, networked understanding of the cellular system,” von Bohlen told ProteoMonitor following the talk. “This is where the focus will be in five years.”
As an industry watcher, von Bohlen commented that many companies’ current efforts to study the differential expression profiles of various tissues and cells may not provide the necessary biological insight to design pharmaceuticals. “As with [mRNA] expression profiles, you’re never sure whether you see the reason for the disease or just the effect of something,” he said.
Instead, he pointed to recently published papers from scientists at MDS Proteomics and Cellzome in Nature as evidence of a growing trend toward a systems biology approach to proteomics. “This is an interesting approach,” he said. “It comes back to a basic knowledge of cell biology.”
Not that von Bohlen is advocating a purely hypothesis-free line of attack. Scientists must continue to confront the challenge of understanding disease using traditional, hypothesis-based experiments, as well as purely discovery-driven research, he said. “This is like building a tunnel from both sides. On the one is understanding the data, and on the other side is to come up with hypothesis models, with algorithms that try to simulate and predict what’s going on.”
When asked whether proteomics is in danger of promising more than the field can deliver, von Bohlen responded that proteomics researchers have taken their cue from the genomics community, and are more realistic about what they can provide. “Everyone understands that proteomics is part of the revolution, but I’m not sure I would even talk of a revolution [at this point],” he said.
Not surprisingly, von Bohlen also sees the trend toward increasing biological complexity as a boon to bioinformatics, given the computing power required to process the data from experiments designed to understand the cell as a system. “It took the largest civil supercomputer at Celera to capture, read, and annotate the human genome,” he said. “What type of computing power would you need for 200,000 proteins in 100 tissues, or taking it to the extreme, 1014 cells in man?
“I believe this will turn into the largest data information management issue ever,” he added.
— JSM