Bioinformatics of RNA structure and transcriptome regulation
Overview and introduction
RNA transcripts are the primary products of activated genes. They yield as products proteins and functional RNAs which constitute key players in all living orgamisms. Yet, how the transcriptome is regulated on RNA level to yield these products remains surprisingly underexplored.
We are particularly interested in mechanisms of gene regulation that are mediated by RNA structure features and by trans RNA-RNA interactions. Both are difficult to probe on a transcriptome-wide scale and in vivo using experimental methods, but exciting progress has been made recently (SHAPE, PARIS, LIGR-seq protocols, all 2016). We have contributed a range of often unique computational methods and analysis pipelines that have allowed us detect a variety of functional RNA features in silico due to their evolutionary conservation and based only on sequence information (e.g. RNA-seq data). On our computational sides, this typically requires dedicated methods that employ sophisticated, probabilistic methods for detecting RNA structure features, trans RNA-RNA interactions and other evolutionarily conserved signals and for testing detailed hypotheses about the underlying molecular mechanisms.
Ongoing projects and collaborations
Since starting here at the MDC in 2016, we have embarked on the following projects and started several exciting collaborations with experimental group on- and off-campus:
- Together with the experimental fruit-fly group of here at the BIMSB-MDC, we have started investigating the role of specific trans RNA-RNA interactions during early fly neurogenesis. This extends our earlier where we show that A-to-I editing in the fruit fly induced changes of local RNA structure features around splice sites that yield splice variants that are specific to cells of the central nervous systems.
- Together with the experimental group of here at the MDC, we have started investigating novel biological classes of functionally important trans RNA-RNA interactions in human embryonic stem cells. This project is particularly exciting as we will look at transcriptome-wide data at sub-cellular resolution.
- Detecting truly novel classes of trans RNA-RNA interactions based on sequence information alone remains computationally and conceptually extremely challenging, see the following recent papers and for more information. We are developing new probabilistic methods for predicting trans RNA-RNA interactions de novo that overcome these challenges and that can also be applied on a transcriptome-wide scale and in a eukaryotic setting.
- A number of exciting, experimental methods for probing the RNA structurome and RNA-RNA interactome in vivo and on a transcriptome-wide scale have recently emerged. As our explains, these require sophisticated computational pipelines to assemble, map and interpret the raw experimental data. We continue to work on new computational methods that aim to combine the best in terms of experimental in-vivo probing with the state-of-art in computational methods in order to get a transcriptome-wide view of RNA structures and trans RNA-RNA interactions in vivo.
- Viruses are wizards at readily combining many, sometimes overlapping signals into their compact genomes. We continue our past interest in understanding virus regulation, see our earlier work on dedicated computational methods ( and ) and their application, e.g. on the RNA structure annotation of the entire . Together with the proteomics group of at the MDC and the virology group of a the Robert-Koch-Institute here in Berlin, we have started to analyse mechanisms of transcriptome regulation that may play a decisive role in determining the species-specificity of different influenza strains.