Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.604701
Title: Computational detection of non-coding RNAs in genomes
Author: Huang, Y.-H.
Awarding Body: University of Cambridge
Current Institution: University of Cambridge
Date of Award: 2009
Availability of Full Text:
Full text unavailable from EThOS. Please contact the current institution’s library for further details.
Abstract:
This thesis is devoted to assessing available approaches and trying new solutions for finding ncRNAs in genomes. In the first half of this thesis, reasons that may contribute to the slow progress of genome-wide ncRNA finding are explored. A comprehensive analysis on a genome-wide scale of the credibility of currently used signals for classifying ncRNAs is conducted. Two factors, conservation of ncRNAs in human-mouse synthetic regions and abundance of covariations between human-mouse synteny-conserved ncRNAs, are evaluated. The result reveals that current comparative-genomics-based methods may not be able to find ncRNAs effectively in mammalian genomes. In addition, possible genomic features that could distinguish real ncRNAs from pseudogenes are investigated. Two different criteria, distribution of bit scores and physical clustering in genomes, are applied to filter out tRNA pseudogenes and to enrich bona-fide tRNA genes. Physiological roles of the tRNA genes in human-mouse synteny-conserved clusters are discussed and the degradation patterns of tRNA pseudogenes are analyzed. In the second half of this thesis, computational techniques are applied to model signals that may be potentially useful for genome-wide ncRNA finding. A sparse Bayesian learning algorithm, Eponine, is applied to model the transcription start sites of mammalian ncRNA genes that are transcribed by RNA polymerase III. In addition to modelling cis-regulatory elements for transcription, a new computational module, which extends the capability of Eponine to learn motifs consisting of both primary sequences and RNA secondary structures, is created. The capability of this new module is demonstrated by applying it to analyze several known cases of ncRNA motifs. The strength and the weakness of applying this new computational approach for finding ncRNAs are discussed.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.604701  DOI: Not available
Share: