Advances and Challenges in Protein-RNA Recognition, Regulation and Prediction

Summary

DNA is often described as the blueprint of life, since it encodes all the information necessary for an organism to develop and maintain its biological functions. Single blueprints for specific functions are stored inside DNA regions called genes. The primary product produced (also termed expressed) from genes is RNA, which can either become biologically active itself (non-coding RNA or ncRNA) or is further translated into proteins (messenger RNA or mRNA), which then executes the gene functions. Given the astonishing complexity of biological functions, it is not surprising that the regulation of gene expression itself is a highly complex matter. Proteins, RNA, and DNA all can interact with each other, forming regulatory networks in order to control the expression of genes. To elucidate these networks, experimental scientists rely more and more on high-throughput methods, producing vast amounts of raw data. Computational methods to analyze these huge datasets are therefore of highest demand. The main focus of this seminar lies on RNA-protein and RNA-RNA interactions. In particular, transcriptome-wide binding patterns of RNA-binding proteins (RBPs), their computational predictability, and the biological effects of binding are discussed. Moreover, the seminar dealt with topics like combinatorial RBP binding prediction, RBP binding kinetics, RNA-RNA interaction prediction, subcellular RNA imaging, and RBP binding site classification. Regarding the computational methodology, several newly developed deep learning methods are presented, e.g. for RBP binding site prediction. Taken together, the aim of the seminar is to bring experimental and computational scientists together for the aforementioned topics and to engage them in fruitful discussions in order to:

• present the current experimental and computational methodologies,
• understand their implications, strengths, and limitations from first-hand experience,
• and spark ideas for developing new computational and experimental methods and improving on existing ones.