'Mass' opportunities
Yi Zhang, PhD, works in the Institute's proteomics lab.
In some respects, proteomics represents both the future and past of cancer research. Scientists have been identifying proteins associated with cancer, one by one, for decades. The difference is that, today, technology has made it possible to study them en masse, as large groups whose members and composition change as cells pass from a normal state to a malignant one.
The challenge of tracking such alterations is many orders of magnitude more complex than tracing the changes in genes. At last count, there are at least 25,000 genes in human cells. Proteins produced from those genes, however, number in the hundreds of thousands. This discrepancy occurs because a single gene can give rise to dozens of different types of proteins. Such proteins usually are related to one another – one being a fragment of another, for example, or a slightly tweaked version of its siblings – but even similar proteins can perform widely different roles within a cell.
Complicating matters further is that proteins rarely work in isolation. They are constantly reacting with other proteins to form new compounds, breaking one another apart and reconnecting to some of the pieces, spinning off new proteins in the process. In addition, the protein picture is never static; the types and distribution of proteins within cells can vary radically from one part of the body to another, from childhood to adulthood, from one hour to the next, even from waking to sleeping.
Given all these considerations, how does one take a "snapshot" of the half-million or so different types of jostling, binding, breaking, shape-shifting proteins in a cell – especially when some are thousands of times more plentiful than others? The short answer: one doesn't (at least not yet). While modern technology, particularly "mass spectrometry," makes it possible to identify thousands of cell proteins and their relative abundance, proteomics experts need to narrow their search to those that seem particularly important in cancer.
While genomics is famous, or notorious, for generating mammoth amounts of data, they're dwarfed by the size of data sets associated with proteomics. For that reason, proteomics requires ever more advanced computer programs to crunch the numbers in a way that's useable for investigators. Computational biology, a field combining biology, mathematics, statistics, and computer science, is critical to proteomics. Without programs for cutting through statistical "clutter" to find nuggets of relevant information, all the advanced protein-sorting technology in the world wouldn't be of much help. At Dana-Farber, computational biology is part of a coordinated approach to proteomics.
- Next: Target selection
- Page: 1 | 2 | 3 | 4
