Browsing by Author "Gardner, Paul Phillip"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Simulating the RNA-world and computational ribonomics : a thesis presented for the degree of Doctor of Philosophy in Biomathematics at Massey University, Palmerston North, New Zealand
    (Massey University, 2003) Gardner, Paul Phillip
    Project 1: Experiments by Piccirilli et al (Nature, Lond. 343, 33-37 (1990)) have shown that the canonical RNA genetic alphabet, AUCG (or ATCG in DNA), is not the only possible nucleotide alphabet. In this work we address the question "Is the canonical alphabet optimal?" Computational tools are used to infer RNA secondary structures (shapes) from RNA sequences of various possible alphabets, and measures of RNA shape are gathered with respect to alphabet size. Then, simulations based upon replication and selection of fixed sized RNA populations are used to investigate the effect of alternative alphabets upon RNAs ability to evolve through a fitness landscape. Those results imply that for low copy fidelity the canonical alphabet is fitter than two, six and eight letter alphabets. Under high copy fidelity conditions, a six letter alphabet out-performed the four letter alphabets, which suggests that the canonical alphabet is indeed a relic of the RNA-world. Project 2: Non-coding RNA genes produce functional RNA molecules rather than proteins. One such family is the H/ACA snoRNAs. Unlike the related C/D snoRNAs, these have resisted automated detection until recently. We develop an algorithm for screening the Saccharomyces cerevisiae genome for novel H/ACA snoRNAs. To achieve this, we introduce some new methods to facilitate the search for non-coding RNAs in genomic sequences which are based on properties of predicted minimum free energy (MFE) secondary structures. The algorithm has been implemented and can be generalised to enable screening of other eukaryote genomes. We find that use of primary sequence data alone is insufficient for identifying novel H/ACA snoRNAs. The use of secondary structure filters reduces the number of candidates to a manageable size. On the basis of genomic location data, we identify three strong H/ACA snoRNA candidates. These together with a further 47 candidates obtained by our analysis are being screened experimentally and investigated (along with known H/ACA snoRNAs) using comparative genomic analysis.