Exploring RNA structure space using multidisciplinary approaches with applications for novel RNA design

Over the past decade remarkable discoveries concerning RNA have reshaped our current understanding of this biopolymer. Synthetic RNAs, developed through in vitro selection experiments, have revealed a surprisingly variety of catalytic or binding activities with biotechnology and therapeutic applications. An intriguing feature of RNA is its capacity to form hierarchical and versatile network-like structures due to predictable base pair interactions. This makes RNA well suited for mathematical and computational approaches. In collaboration with the Schlick group, I have contributed to elucidate the universe of RNA structures using a multidisciplinary approach combining biological, mathematical and computational methods. Our work has led to the systematic analysis and design of RNA structures.;In order to explore the entire possible RNA structure space, we have developed a new method of analyzing and predicting RNA 2D topologies utilizing their graphical representations and graph-theoretical tools. We have constructed the RNA-As-Graphs database to catalogue and classify possible RNA graphs, as well as to offer a comprehensive searching capacity, which can predict novel RNA folds in both natural and synthetic RNAs. This allows us to cluster possible RNA topologies into two categories: RNA-like or non-RNA-like. RNA-like topologies are more likely to exist in Nature.;Another application for RNA-like graphs is modeling of RNA in vitro selection, a widely used experimental technology for finding active RNAs for biotechnology and therapeutic applications. Specifically, we have developed a computational framework for modeling key aspects of in vitro selection procedures (pool generation and design). The concept of the "mixing matrix'' was introduced to represent the mixing ratios of nucleotides in synthesis ports. This enables us to design sequence pools targeting RNA-like topologies that are more likely to be active when produced experimentally. To assist experimentalists, we have made these pool design procedures available through a web-server, RNA-As-Graph-Pools. To further aid experimentalists, we have also developed a computational methodology for screening large (10 9--1014 sequences) pools that is designed to simulate the essential selection process.;Finally, we have sought to design specific fluorescent riboswitches to visualize transcription elongation by combining two distinct modules of gene-regulating riboswitch and fluorescence-binding aptamer using structural and energetic analyses.