LIFE SCIENCES

Life at the molecular level is driven by the interplay of many biomolecules. Much like man-made machines in everyday life, they need to move, rotate, react to signals or use and provide resources. Unlike man-made machines, however, they function at the atomic level so directly observing their workings is impossible as they are invisible both to the naked eye and regular optic microscopes. Specific highly specialised equipment can provide insight into the inner working of these atomic-sized machines, but such equipment is very expensive and the required wet-lab setups can be highly involved. A powerful complement to such involved wet-lab setups is using molecular simulations on supercomputers to provide insight about theses atomic-sized machines. In such simulations, the investigated biomolecules are represented via sophisticated mathematical descriptions. Running such simulations is computationally very expensive as these descriptions need to be run on very short timescales. Akin to a movie, which requires 24 frames per second to appear smoothly moving, such so-called Molecular Dynamics simulations need to re-calculate a new snapshot of the molecular movie every 2 femtoseconds. This means that even simulations describing 1/1000 of a second need to run 1 Million times 1 Million different snapshots for all involved atoms and their interactions- an impressive feat of engineering! But once the technical difficulties are solved, such simulations allow to see a “virtual movie” of the investigated biomolecules akin a virtual microscopic and observe their dynamics and function with atomic detail at the investigated timescale.

In this project, the focus is running Molecule Dynamics simulations on a specific class of biomolecules, so-called Histidine Kinases. These Kinases are part of specific signalling cascades in living systems and enable for examples bacteria to sense environmental conditions and react to them. Molecular Dynamics Simulations on the supercomputer Juwels looked with atomic detail at the molecular motions and interactions with other biomolecules to better decipher their inner working. A specific focus was adjusting the resolution of the simulation for the simulations, as highly detailed simulations, while providing more detailed insight, are also computationally considerably more demanding. Accordingly, specific techniques to adjust the granularity of the simulations (“coarse-graining”) were used to fine-tune the required level of detail to optimise the use of computational resources. In the simulations, the scientists around Alexander Schug and Fathia Idiris could highlight in detail how Histidine Kinases reacts with the environment and how a chemical reaction with ATP, a molecular energy storage, occurs. Understanding the details of these Kinases is crucial both to better understand life at the molecular level but at the same time has important pharmaceutical and medical implication, as this understanding helps targeting bacteria which rely on such molecular sensors more directly.

In recent decades, simulation methods for materials science, chemistry and soft matter applications have undergone massive developments. Although these are efficient and accurate for the specific field of application, they are usually limited to a specific time and length scale. This is problematic for systems where multiple time and length scales are relevant for understanding the molecular mechanisms. The displayed biomolecule Histidine Kinase is involved in cellular signal processing. The molecular mechanisms of these tasks are investigated in simulations on various scales. These simulations range from a simplified view with a reduced number of atoms (left) to the inclusion of all atoms including the water environment (center and right), using classical force fields or complex quantum chemical calculations.