PhD Defence: Alexander Junge

2017-06-23: Towards a systems-level understanding of RNA secondary structure and interactions. The defence will take place on June 23rd, 2017, 14.15-17.15 in Auditorium A2-70.03, Thorvaldsensvej 40, 1870 Frederiksberg C.

Everbody is welcome.

Reception starting approx 17.30 at Grønnegårdsvej 7, library on 1st floor, 1870 Frederiksberg C.

Principal supervisor:
Professor Jan Gorodkin
Center for non-coding RNA in Technology and Health, IVH, KU


  • Senior Research Scientist Jack Bernard Cowland, Section of Haematology-Oncology, Department of Clinical Genetics, National University Hospital, KU
  • Associate Professor Jakob Hull Havgaard, Center for non-coding RNA in Technology and Health, IVH, KU

Assessment committee:

  • (Chairperson) Professor Søren Brunak, Novo Nordisk Foundation Center for Protein Research, University of Copenhagen, DK
  • Professor Ivo Hofacker, Institute for Theoretical Chemistry, University of Vienna, AT
  • Professor Hanah Margalit, Department of Microbiology and Molecular Genetics, Faculty of Medicine, The Hebrew University of Jerusalem, IL

Non-coding RNAs (ncRNAs) regulate gene expression, catalyze biochemical reactions, and mediate multiple other crucial cellular functions. The structure of an RNA molecule, its biochemical makeup including intramolecular base pairs between nucleotides, mediates the function of an RNA. Albeit, ncRNAs do not function in isolation but in interaction with other molecules to form cellular pathways, gene regulation networks or molecular complexes. In this thesis, three lines of work that address RNA bioinformatics from a systems-level perspective are introduced. It presents approaches to cluster evolutionarily conserved RNA structures and to catalog ncRNA interactions without focusing on a specific class of RNAs or interactions.

Firstly, RNAscClust is introduced, a tool for sequence-structure clustering of RNA sequences from an organism of interest aligned to related sequences found in other species. Contrary to most previous clustering approaches that cluster unaligned sequences, RNAscClust makes use of information about structure conservation apparent from compensatory mutations that preserve base pairs in evolution. Its clustering performance benefits from an increasing amount of covariation in our benchmarks. To illustrate its applicability to real world data sets, RNAscClust is applied to cluster conserved RNA structures identified by screening vertebrate genomes. This approach identifies known and predicts putative regulatory elements in untranslated regions of genes.

Secondly, RAIN is presented, a novel database cataloging ncRNA–RNA and ncRNA–protein interactions, encompassing direct (physical) and indirect (functional) associations. RAIN integrates interactions from a wide range of resources including experimentally supported and predicted interactions as well as associations extracted by automated mining of the biomedical literature. A confidence score is assigned to each interaction which offers a probabilistic interpretation of the evidence underlying the association. RAIN is tightly integrated with the STRING database for protein–protein interactions which enables a more complete understanding of biological interaction networks.

Thirdly, the role of competing endogenous RNAs (ceRNAs) in t(8;21) acute myeloid leukemia (AML) is investigated. T(8;21) AML is characterized by a translocation between chromosome 8 and 21 leading to the formation of a RUNX1-RUNX1T1 fusion gene. CeRNAs compete for microRNA binding with other transcripts leading to a mutual effect on gene regulation. Since RUNX1T1 is heavily upregulated in t(8;21) AML compared to healthy controls, we investigated its potential role as a ceRNA. We pro- pose that the ceRNA RUNX1T1 influences the expression levels of multiple genes and pathways that have previously been linked to leukemia development.