miRNA detection and analysis from high-throughput small RNA sequencing data

Paicu, Claudia (2016) miRNA detection and analysis from high-throughput small RNA sequencing data. Doctoral thesis, University of East Anglia.

[img]
Preview
PDF
Download (10MB) | Preview

Abstract

Small RNAs (sRNAs) are a broad class of short regulatory non-coding
RNAs. microRNAs (miRNAs) are a special class of -21-22 nucleotide
sRNAs which are derived from a stable hairpin-like secondary structure.
miRNAs have critical gene regulatory functions and are involved
in many pathways including developmental timing, organogenesis and
development in both plants and animals. Next generation sequencing
(NGS) technologies, which are often used for identifying miRNAs,
are continuously evolving, generating datasets containing millions of
sRNAs, which has led to new challenges for the tools used to predict
miRNAs from such data. There are several tools for miRNA detection
from NGS datasets, which we review in this thesis, identifying a
number of potential shortcomings in their algorithms.
In this thesis, we present a novel miRNA prediction algorithm, miRCat2.
Our algorithm is more robust to variations in sequencing depth
due to the fact that it compares aligned sRNA reads to a random uniform
distribution to detect peaks in the input dataset, using a new
entropy-based approach. Then it applies filters based on the miRNA
biogenesis on the read alignment and on the computed secondary
structure.
Results show that miRCat2 has a better specificity-sensitivity trade-off than similar tools, and its predictions also contains a larger percentage
of sequences that are downregulated in mutants in the miRNA
biogenesis pathway. This con�rms the validity of novel predictions,
which may lead to new miRNA annotations, expanding and contributing
to the field of sRNA research.

Item Type: Thesis (Doctoral)
Faculty \ School: Faculty of Science > School of Computing Sciences
Depositing User: Katie Miller
Date Deposited: 13 Jun 2017 13:26
Last Modified: 31 Dec 2019 01:38
URI: https://ueaeprints.uea.ac.uk/id/eprint/63738
DOI:

Actions (login required)

View Item View Item