Investigating the microRNA-mediated regulation of protein-coding transcripts in animals using RNA-Seq data

MicroRNAs (miRNA) are small non-coding RNAs, of approximately 22 nucleotides in length, which play an important role in the post-transcriptional regulation of gene expression. Post-transcriptional regulation by miRNAs is achieved by direct translational inhibition or decay of other RNA molecules, or a combination of both of these mechanisms. miRNAs are implicated in a large number of developmental processes across the animal kingdom, underscoring their importance to biological research, and in particular, research relating to diseases states as a result of aberrant development. Investigation of the precise role of individual miRNAs or groups of miRNAs within cells requires the accurate identification of miRNA targets. Limitations in experimental methods for the identification of miRNA targets, necessitates the use of computational algorithms for this purpose. However, accurate computational identification of miRNA targets can be difficult, due to the short six or seven nucleotide seed sequence of the miRNA which is used for the recognition of targets, leading to a large number of false positive predictions being made. In this thesis, I demonstrate how data from transcriptome-wide bulk RNA sequencing experiments can be used to increase the accuracy of miRNA target prediction workflows. Firstly, I show how data of this type can be used to generate 3’UTR annotations specific to the biological context in which sequencing occurred, and secondly, how it can be used to remove lowly expressed mRNA transcripts from the target prediction process. Implementation of both of these steps in miRNA predictions workflows is shown in this thesis to increase prediction accuracy. In addition, I explore how data from bulk RNA Sequencing can be used in combination with data generated from small RNA sequencing experiments in order to infer the regulatory activity of individual miRNAs during given developmental processes