I am presently a graduate student with Prof. Chris Bystroff, majoring in Biochemistry and Biophysics with a concentration in Computational Biology and Bioinformatics. We have focused on designing green fluorescent protein (GFP) - a naturally-occurring, intrinsically fluorescent molecule - to serve as a potential biosensor for pathogen detection. We have demonstrated previously at a proof-of-concept level that GFP can be reorganized to omit key elements of its structure that eliminate or severely weaken the fluorescent signal given off, that the reintroduction of these key structural features from an external source restore the protein's fluorescence, and that the remaining reorganized, edited GFP variant can be rationally redesigned to emit fluorescence in the presence of a different target sequence, ideally from another protein whose presence we want to screen.
To date, most of the work on this project has been largely computational and a team effort from almost everyone affiliated with our lab, past and present. Most recently, we have completed the steps necessary to automate the design process from an end user perspective. Our design process is computational and well-established, but the results it has given are either unsuccessful or very low affinity. We have acknowledged that to improve our designs, we need to overcome computational constraints. My job has largely been to write in auxiliary software programs to facilitate our management of these constraints, as well as communicating with other developers to ensure that the existing software components do not require heavy modification for interaction with my new software.
Using GFP as our base for design allows us the additional opportunity to study the protein folding problem at a more basic level. Our edited, non-fluorescent mutants have provided us insight into GFP's folding pathway, and allowed us to consider a number of putative starting points for target selection and design.
My future work is two-fold: first, we want to continue to refine our design “pipeline” and take advantage of all the computing resources available to us, either supercomputing or volunteer computing. This also entails taking computational outputs, generating and refining them experimentally, and using the information gained from the whole campaign to better improve our design process. Second, we have seen that these putative biosensors have a much higher signal-to-noise ratio when immobilized on a solid support, and we would like to try to utilize these immobilized GFP variants in quantitative experiments.
Personally, I chose to go into Biology because I am fascinated at how life works at the smallest levels, and how seemingly minor alterations at the sub-molecular levels lead to significant differences at the organismal level. Through my research and the influence of my adviser's structural biology background, I have determined that “the smallest levels” really means the structural level to me. As technology has progressed, knowledge of computers and programming has become a useful tool in all disciplines, and I have always had an interest in computers and problem solving, which made the decision to double major in Computer Science easy. Seeing as experimental work in structural biology can't be done easily without investing a lot of time, computers have become an essential tool both for solving structures and working with databases of existing structures. The work I am currently involved in now beautifully represents how well computational technology can complement other scientific endeavors.