For her research in modeling and parameter estimation of small gene networks, Bradley Fellow Rebecca Shelton needed a firm understanding of the strengths and weaknesses of the data involved. This led to spending a couple months in Jean Pecoud's "wet" laboratory at the Virginia Bioinformatics Institute (VBI).
Her task was to prepare DNA constructs for testing. Putting the constructs together involved using restriction enzymes to digest, or cut, the pieces first, she explained. The digestion was then inserted into a gel made with agar, merged in a buffer, and electrified &mdash a process called electrophoresis. Different sized DNA fragments, which are negatively charged, move at different speeds toward the anode. "We then compared the fragments with a constant to select the size we needed. We cut it from the gel and isolated it," she said. "We were hoping to get data for a system model."
She found some interesting differences between an EE laboratory and a biology wet laboratory. "In engineering, we don't set things up and wait overnight to see if they grow. We just turn equipment on." She also gained an appreciation for good laboratory technique. "Not only is it surprisingly hard," she said, "but I also had to get comfortable working around mutagens and what you think of as potential disease-causing organisms like E. coli."
Her biggest lesson, however, was a new appreciation for data. "I thought I knew where data comes from and what variables go into it. In the lab, however, you see where the errors can come from. You can see all the things that can happen between giving someone a design for an experiment and the data that come out."