With increasing capacities of computers and global digitalization, new methods and algorithms grow with increasing speed throughout almost every research area. Biotechnology in particular often uses computationally expensive simulations and algorithms and benefits from this trend in consequence. We dived into this subject by experimenting with state-of-the-art algorithms for RNA and DNA folding and structure prediction on the one hand and Molecular Dynamics (MD) simulations on the other.
We aimed to support the laboratory work with computational methods from the bioinformatics community. In particular, our biosafety project was in need of a hokD kill-switch to disable bacteria, which lead to the idea of a computationally aided riboswitch design. Additionally, we wanted to select a promotor for the XylE-transporter based on a MD simulation.
We managed to develop a machine learning based system for general RNA secondary structure prediction and simulated membrane integrity for the XylE-transporter. Furthermore, based on the insights we gained from researching computational structure prediction techniques, we additionally developed a genetic algorithm for the automatic design of RNA riboswitches.