Open Access

Sebastian Lobentanzer

• Natural science is only just beginning to understand the complex processes surrounding transcription. Epitranscriptional regulation is in large parts conveyed by transcription factors (TFs) and two recently discovered small RNA (smRNA) species: microRNAs (miRNAs) and transfer RNA fragments (tRFs). As opposed to the fairly well-characterised function of TFs in shaping the phenotype of the cell, the effects and mechanism of action of smRNA species is less well understood. In particular, the multi-levelled combinatorial interaction (many-to-many) of smRNAs presents new challenges to molecular biology. This dissertation contributes to the study of smRNA dynamics in mammalian cells in several ways, which are presented in three main chapters. I) The exhaustive analysis of the many-to-many network of smRNA regulation is reliant on bioinformatic support. Here, I describe the development of an integrative database capable of fast and efficient computation of complex multi-levelled transcriptional interactions, named miRNeo. This infrastructure is then applied to two use cases. II) To elucidate smRNA dynamics of cholinergic systems and their relevance to psychiatric disease, an integrative transcriptomics analysis is performed on patient brain sample data, single-cell sequencing data, and two closely related in vitro human cholinergic cellular models reflecting male and female phenotypes. III) The dynamics between small and large RNA transcripts in the blood of stroke victims are analysed via a combination of sequencing, analysis of sorted blood cell populations, and bioinformatic methods based on the miRNeo infrastructure. Particularly, importance and practicality of smRNA:TF:gene feedforward loops are assessed. In both analytic scenarios, I identify the most pertinent regulators of disease-relevant processes and biological pathways implicated in either pathogenesis or responses to the disease. While the examples described in chapters three and four of this dissertation are disease-specific applications of miRNeo, the database and methods described have been developed to be applicable to the whole genome and all known smRNAs.