Graduate Alumni Research Award, Fan Zheng
Fan Zheng, a PhD student in the Molecular and Cellular Biology Program, explains below how the Graduate Alumni Research Award facilitated his research on protein design.
Proteins adopt certain three-dimensional structures in order to exert their diverse functions. Computational structural biology is an emerging field studying protein structures that combines work in computer science, biology, and chemistry. Protein design, one of the branches of computational structural biology, involves manipulating the amino acid sequences of proteins to make engineered proteins with optimized or novel structures and functions.
In order to make useful engineered proteins, the stability of proteins must be preserved when changing amino acids. Since the traditional approach of identifying valid mutations is often inefficient, it would be valuable if the stability changes upon mutations could be predicted with computational methods. However, for many years, there has not been much success in attempting to develop an efficient and interpretable computational method.
In this sponsored study, we are trying to target this problem with a new strategy. Rapid growth in the number of experimentally-determined protein structures collected in the PDB (Protein Data Bank) makes it possible to cluster proteins according to diverse features and analyze their sequences. Previously in our lab, we have developed MaDCaT, a software which searches for similar structural pattern in the database for any given protein and produces valuable statistics to be analyzed. Combining MaDCaT and other computational tools, we are trying to develop a framework for accurate prediction of stability change upon mutations.
The frequent usage of large databases and the inherent complexity of representing protein structures oftentimes make the applications used in computational structural biology very expensive. In general, most of the research in computational structural biology requires high-performance computing clusters. Such CPU clusters exist on campus, but the usage of such facilities is limited.
Thanks to the Graduate Alumni Research Award, I am able to explore the option of using commercial computing services such as Amazon EC2 (Amazon Elastic Compute Cloud). Amazon EC2 allows for computed resources to be scaled nearly instantaneously on demand, reducing the need to forecast traffic. This resource has greatly facilitated my project, and preliminary data have shown promising results. We expect our work will provide insight into structural biology and protein engineering in the near future.
by Fan Zheng