Development of bioinformatics tools to understand mechanisms of non-coding small RNA interactions

点击翻译按钮获取中文摘要

RNA elements are transcribed (expressed) DNA fragments that are not translated into proteins. They play key roles in translation, splicing, and gene regulation. In Diplonema, mitochondrial genes are fragmented into modules, which are transcribed separately and then assembled by an unknown mechanism. It has been hypothesized that this process be mediated by RNA elements. Likewise in animal cells, messenger RNAs and RNA elements are packaged together in a non-random fashion into microvisicles called exosomes. These vesicles play important roles in regular and dysfunctional physiological processes. What is common to the above examples is that the biological function of these elements critically depends on intra- and inter- RNA interactions. Computational tools to assist the identification and annotation of RNA interaction motifs are seriously lacking. Our group has solid expertise in solving RNA structure puzzles. In recent years, we developed tools for the simultaneous alignment and structure prediction of RNA sequences (eXtended Dynalign and pDynalign). We used suffix arrays for the discovery of RNA secondary structure motifs (Seed). More recently, we adapted frequent subgraph mining for the discovery of RNA interaction motifs (RiboFSM). With our life science collaborators, we studied the IRES motifs in the UTRs of mammalian genes, and developed computational approaches to investigate the possible role of RNA elements in cis- and trans- splicing activities in the mitochondrial genome of Diplonema. High-throughput (HT) methods, such as expression microarrays, genome-wide physical and genomic interaction screens, while allowing to monitor the behavior of the cell as a whole, are generating wealth of information that needs to be studied and interpreted. To help experts interpret experimental data, we recently proposed a novel logic-based Annotation Concept Synthesis and Enrichment Analysis (ACSEA) approach. ACSEA merges inductive logic reasoning with statistical inference to discover complex concepts from experiments. Finally, we also introduced a logic based approach, called Module Inducer, to study the architecture of DNA landmarks. These approaches will be further extended in the proposed research. This proposal has two themes and five objectives: 1. Identification of RNA interaction motifs 1.1 Adapt frequent subgraph mining to determine RNA interaction motifs 1.2 Generalize the suffix array based approaches, namely Seed, to determine RNA interaction motifs 1.3 Apply and compare the newly developed tools to analyze the mitochondrial genome of Diplonema, as well as exosome shuttle RNAs 2. Integration and mining of heterogeneous data 2.1 Research the use of Description Logic, as an alternative to first-order logic, for Annotation Concept Synthesis and Enrichment Analysis 2.2 Extend Module Inducer to use ACSEA for the inference of structural rules describing the architecture of transcription factor binding sites The long term objective of this research is to combine the motif discovery methods developed in theme 1 with the data integration techniques established in theme 2 so as to create performant ncRNA gene detection methods combining structural and contextual information. These tools will help life science experts validate or refute hypotheses about the roles and mechanisms of RNA elements. RNA elements play critical roles in the cell, the dysregulation of the associated processes is often associated with disease state, consequently progress in understanding them will therefore have a direct impact on human health, agriculture, and on our understanding of cellular biology in general.