Long noncoding RNAs in eukaryotic gemomes

Noncoding RNAs (ncRNAs) are transcripts that are not translated to proteins but act on the level of the RNA. During the past few years, it has become evident that such "RNA genes" are more abundant than previously thought. Most prominently, the discovery of micro RNAs (miRNAs) led to a paradigm shift in our understanding of gene regulation. However, there is mounting evidence that there are thousands of ncRNAs that do not belong to the class of miRNAs. Sensitive experimental methods to map transcripts on a genome-wide level, revealed a class of mRNA-like ncRNAs that are spliced, polyadenylated and several kilobases in length. Unlike miRNAs, their role is poorly understood and only for a handful of examples their function is known. However, the few described examples show that these ncRNAs are involved in a surprisingly diverse set of processes, such as (post- )transcriptional regulation, chromatin modification/epigenetics, cell-proliferation and development. The goal of this research project is to better understand the role of this enigmatic class of ncRNAs by a combination of computational and experimental methods. In a first step, we aim at a better annotation of ncRNAs in eukaryotic and in particular mammalian genomes. A high confidence parts-list is a prerequisite for any further analysis. We will improve existing methods to predict conserved RNA secondary structures which is currently the most promising approach for de novo annotation. More specifically, we will extend our software RNAz to deal with large comparative datasets and lineage specific structures. As a complementary approach to structure based RNA prediction, we will explore novel evolutionary signatures associated with ncRNAs including `ultraconservation`-effects, mutation-rate asymmetries and substitution rate heterogeneity. In addition to these purely computational approaches, we will use high-throughput experimental data to support ncRNA annotation. In contrast to previous research in this field that had a focus on transcriptomics data, the focus of this project will be on the conceptually new idea of using histone-modification patterns to identify transcribed regions. Using these different methods, sets of candidate RNAs will be generated that will undergo computational analysis and (in collaboration with experimental groups) experimental validation. We will classify the candidates based on local sequence/structure motifs, upstream regulatory regions on the DNA level, and potential RNA/RNA and RNA/DNA interaction partners. Moreover, we will analyze transcript structure, coding potential and the relationship to neighbouring or overlapping protein genes. On the experimental side, detailed transcript structures will be determined using established PCR based methods and new techniques based on next-generation sequencing. Towards functional characterization of novel ncRNA candidates, custom micro-arrays will be used to study tissue and environmental specific expression patterns as well as the co-expression and co-regulation with known protein coding genes. For the most promising candidates individual functional studies including siRNA knock-down will be carried out. This integrative approach of well established experimental and computational methodology combined with conceptually new approaches and techniques will enhance our knowledge of ncRNAs in eukaryotic genomes. It will contribute to a better understanding of genomic organization, evolution and ultimately function of this still largely ignored type of genes.