Comparative RNA structure prediction beyond stochastic context free grammars

点击翻译按钮获取中文摘要

RNA molecules are at the core of numerous fundamental processes in any cell. They encode information on the set of proteins to be synthesized in a cell; they help synthesize proteins; they control protein-production rates (see Nobel Prize in Medicine 2006); and they also act as genes in their own right (so-called RNA genes). Many of these functions rely on the structure of the RNA molecule.RNA molecules are linear chains of variable lengths made of four elementary building blocks, the nucleotides denoted A, C, G and U. An RNA structure is defined by weak hydrogen bonds between pairs of non-consecutive nucleotides in the RNA sequence; the three possible pairs are {A, U}, {G, C} and {G, U}. The number of possible RNA structures that an RNA sequence can assume grows exponentially with the length of the sequence. The challenge to the theoretical scientist is to single out the RNA structure that confers the functional property to the RNA molecule.To address this challenge, most existing methods rely on the assumption that RNA molecules assume the most stable structure. Recent results show that this assumption does not hold in general. We propose to develop novel,significantly improved, theoretical methods which do not rely on this assumption and which overcome the major conceptual limitations of the existing methods in order to improve RNA structure prediction, RNA gene prediction and to address many of the most fundamental and exciting questions in molecular biology today.