Chemistry and Biochemistry Seminar: Amarda Shehu
All about Energy Landscapes: Generating and Analyzing them to Predict and Characterize Protein Structure, Dynamics, and Function.
Amarda Shehu (George Mason University, Computer Science)
Research in my laboratory focuses on the design of novel algorithmic frameworks for elucidating biomolecular structures and their rearrangements as fundamental to understanding (dys)function, cellular processes, our own biology, and disease. Inspiration comes from a combination of biology and science and engineering fields that model dynamic systems. Specifically, we are driven by the energy landscape view, which allows modeling and understanding the behavior of intrinsically-dynamic molecular systems interconverting between structures with varying energies. In this talk I will share two recent directions of our research, one where we utilize protein energy landscapes to address a well-known but open problem known as decoy selection in template-free protein structure prediction, and one where we construct detailed representations of energy landscapes of healthy and diseased variants of a protein molecule of interest to human health. On the latter, I will demonstrate that computing and mining landscapes is allowing us to discover and categorize mechanisms via which pathogenic mutations alter protein dynamics and function in human disorders. This is bringing us closer to machines revealing part of the complex puzzle on how mutations alter biological activities.