Meera Sitharam, an associate professor in the University of Florida Department of Computer & Information Science & Engineering (CISE), and Mavis Agbandje-McKenna, a professor in the UF Department of Biochemistry and Molecular Biology, were awarded $799,990 of a $1.13 million grant funded by the National Science Foundation (NSF).
The project, “Collaborative Research: Geometric Elucidation of Supramolecular Assembly and Allostery with Experimental Validation,” is a collaboration with Carnegie Mellon University.
A wide variety of supramolecular structures in nature and engineering–from viruses to protein crystals to nanomaterials–assemble rapidly and spontaneously at room temperature with remarkable efficacy. Many assembly processes incorporate the phenomenon of allostery, where intermolecular interaction is controlled by binding events at a remote site of one of the interacting molecules. Despite increasingly sophisticated in vivo, in vitro and in silico experimental efforts, assembly processes are poorly understood. A more mathematically rigorous, and mechanistically intuitive theory is crucial not only to predict and engineer assembly and allostery but also to guide further time-consuming experimentation. Deeper understanding of assembly, allostery, and the role of the latter in the former will help control infectious diseases, assemble viral vectors for gene therapy, design drugs and engineer materials at the nanoscale.
It is natural to expect that geometry and algorithmic complexity would play a key role in understanding the mechanisms underlying assembly, since the assembly process must crucially depend on the intricate shape and volume of the so-called assembly configuration space in which the molecules move relative to each other as they assemble. Conversely, it is also natural to expect that new mathematics, algorithms and software will result from the quest to understand intricate molecular configuration spaces and perform computations over them. The project’s goals include new theorems and algorithms, their hybridization with prevailing methods, and open-source software. Progress is expected on long open problems in rigidity, configuration spaces, distance geometry; algorithms for efficient atlasing, search, sampling, and volume computation for high dimensional and topologically intricate configuration spaces; hybrid methods that combine the new algorithms with prevailing energy-based Monte Carlo simulation; and most significantly, concrete experimental validation of predictions. The project combines expertise in geometry and algorithms, experimental structural biology, and computational chemistry, and is well-suited for bringing the three communities together, for providing interdisciplinary training for research students as well as for outreach to schools and the public.