Transcriptional Regulation Group
The field of transcriptional regulation has gone through a rapid development over the last couple of years. This is due to the plethora of whole-genome sequence data and the functional genomics data on gene expression, DNA-binding proteins, and epigenetics, which have become available (e.g., the ENCODE data). The group works on exploiting this data for the purpose of gaining a better understanding of transcriptional regulation in eukaryotes. The main questions lie in the identification of the regulatory sequence motifs and the interplay between epigenetic marks and regulation. To this end, we develop methods and analyse particular data sets. The ultimate goal is to unravel biological networks and pinpoint possible transcriptional mechanisms behind the interactions.
Sequence-based gene regulation
[Morgane Thomas-Chollier, Matthias Heinig, Jose Muino, Meng Guofeng, Jonathan Göke, Annalisa Marsico]
For many years, our group has worked on developing methods to predict transcription factor binding sites from sequence motifs (the TRAP method, Roider at al., Bioinformatics 2007) and to find common motifs among co-expressed promoters (the PASTAA method, Roider et al., NAR 2009, together with Stefan Haas). This has also led to an interest in regulatory mutations, the recognition of which is the goal of the sTRAP method (Manke et al., Hum Mutat 2010). The software that has resulted from these efforts has been summarized in a Nature Protocol paper (Thomas-Chollier et al., Nat Prot 2012) and is available via a web server. A number of collaborative projects have profited from our expertise in this area, most notably the project with Norbert Hübner, Max Delbrück Center for Molecular Medicine (MDC), Berlin, on eQTLs (Heinig et al., Nature 2010) and within the department the projects with Sebastiaan Meijsing (see his report). Jose Muino has analysed regulatory circuitry in plants (e.g., Schiessl et al., PNAS 2014).
For many years, there had been the hope that the sequence-based prediction of regulation would allow us to understand a large part of the transcription factor–target relations. However, the increasing amount of tissue- and condition-specific functional data have made it apparent that cellular processes are much more dynamic and that sequence alone may (in part) be a prerequisite for regulation, but is by no means sufficient to explain gene expression. While this trivial realization has led to an increased push to understand epigenetic regulation, we have also used gene expression data in addition to sequence patterns for prediction of transcription factor–target relationships (Meng & Vingron, Bioinfomatics 2014). Likewise, in collaboration with Huck-Hui Ng, Genome Institute of Singapore (GIS), careful analysis of gene expression patterns together with transcription factor binding has led to the unravelling of a network of different ERK signalling pathways (Göke et al., Mol Cell 2013).
In the context of a DFG-funded collaboration (Transregio SFB) with Bernd Schmeck (initially at Charité and recently at Marburg University), we are studying miRNAs, their regulation, and their targets. In particular, in the case of miRNAs it is very difficult to tell where the promoter lies, because even in the sequencing data the 5’ end of the gene is usually not visible due to the processing of the miRNA. Again, other information beyond the usual promoter elements needs to be taken into account. Annalisa Marsico developed a promoter recognition method for miRNAs that utilizes CAGE data in conjunction with machine learning algorithms to extract the promoters of the miRNAs (the PROmiRNA method, Marsico et al., Genome Biol 2013). This was within collaboration with Ulf Ørom from the Otto Warburg Laboratory, whose group tested predictions experimentally. Annalisa Marsico has meanwhile become an Assistant Professor at the Freie Universität (FU) Berlin. Within a cooperation between MPIMG and FU, she now leads her own group at the MPIMG.