28 new Czech research projects utilizing LUMI chosen

IT4Innovations National Supercomputing Center in the Czech Republic has announced the results of the 32nd round of its Open Access Grant Competition, including the Czech national resource share for LUMI. A total of 28 new Czech research projects utilizing LUMI were chosen in this call.

The selected projects represent various disciplines. Five selected projects are described in more detail below – in these projects researchers are, e.g., studying how graphene materials interact with biomembranes to find ways to Parkinson’s disease treatment, advancing studies related to thermoelectricity which holds significant potential for improving energy efficiency in industries, vehicles, and household devices, and to better understand seismic tremors that can threaten key infrastructure.

Graphene-based materials in nanomedicine

Researcher: Markéta Paloncýová, CATRIN, Palacký University Olomouc
Research area: Biosciences

The Karolina and LUMI supercomputers are helping CATRIN scientists investigate how graphene materials interact with biomembranes using molecular dynamics simulations. Graphene derivatives have been studied in recent years for their prospects in nanomedicine due to their mechanical, electrical, and chemical properties. This project intends to test the biocompatibility and possible toxicity of reduced graphene oxide (rGO) when in contact with brain cells to use it as a stimulating electrode in treating Parkinson’s disease. The simulations will be used to study how rGO interacts with membrane models in different dimensions and compositions and the interplay between the behaviour of rGO and lipids in membranes.

The project aims to develop new approaches combining molecular dynamics simulations at the atomic and coarse-grained level that can be extended to other models of biomembranes or nanomaterials, thus enabling systematic investigation of interactions at the interface of bio- and nanomaterials. This research is part of the HE RIA MINIGRAPH project to develop an rGO-based electrode for treating Parkinson’s disease.

Learning Large Scale Object Manipulation Dynamics from the Video

Researcher: Georgij Ponimatkin, Czech Technical University in Prague
Research area: Informatics

Researchers from The Czech Institute of Informatics, Robotics and Cybernetics (CIIRC CTU) will use the allocated computing time on the LUMI supercomputer to develop a large-scale dynamics model that can predict action trajectories in 2D space based on textual descriptions of desired actions. For example, given a photo of a person holding a water bottle and a caption like “pour the water into the cup,” the model will forecast the movement path of the action. These predictions can then guide robotic systems in performing tasks, offering a step forward in robotic manipulation. This research is supported by the ERC Advanced Grant “FRONTIER” GA no. 101097822.

Strategies of chemical doping for high-efficiency thermoelectricity in transition metal nitrides

Researchers: Luigi Cigarini, Urszula D. Wdowik and Dominik Legut, IT4Innovations
Research area: Material sciences

Image: The image shows a comparison of the effects on electron transport caused by atomic vacancies and substitutional oxygen atoms replacing the central nitrogen atoms in an infinite nanowire structure composed of 5×5 crystal cells of scandium nitride.

Luigi Cigarini, Urszula D. Wdowik, and Dominik Legut from IT4Innovations will utilize the computational power of the Barbora, Karolina, and LUMI supercomputers to explore the thermoelectric properties of scandium nitride and chromium nitride — two promising materials for energy conversion technologies. The theoretical calculations will focus on the effects of defects and dopants in the crystal structures of these materials.

Thermoelectricity, the ability to convert temperature differences into electric voltage and vice versa, holds significant potential for improving energy efficiency in industries, vehicles, and household devices. The project aims to develop strategies to enhance the performance of cost-effective thermoelectric materials, paving the way for innovative energy solutions and advancing technological progress in energy efficiency. The research is supported by the Czech Science Foundation by the grant No. 23-07228S.

POL2PHASE

Researcher: William Shakespeare Morton, CEITEC
Research area: Biosciences

The image shows the internal organization of a two-proteins undergoing LLPS. One component is colored yellow, while the second component is separated into grey (folded part of the protein) and red (disordered part of the protein). By studying these interactions, we can learn which are crucial for healthy cellular function, and what effects mutations might have on said function.

Liquid-Liquid Phase Separation (LLPS) is key for cellular function, as it allows proteins to be stored in specific areas within cells. This process is (mostly) driven by intrinsically disordered regions of proteins, which enable cells to quickly assemble or disassemble these clusters as needed. It’s like how we might store fruit in a fruit bowl; semi-contained but easily accessible. Improper control of these clusters can lead to neurodegenerative diseases, cancer, and other illnesses.

William Shakespeare Morton from CEITEC Masaryk University will use the Karolina and LUMI supercomputers to develop methods for studying protein clustering during LLPS. They will also apply machine learning and experimentally validate their results. The research is supported by the MSCA Postdoctoral Fellowship “POL2PHASE.”

3D Modeling and Risk Assessment for Nuclear Facility Safety

Researcher: Ivo Opršal, Czech Academy of Sciences
Research area: Earth Sciences

Image: Simulated and observed spectral amplification (here called HVRS ratios) in the near-surface geological structure of the Mýtina maar in western Bohemia. Black: Observed data for two rarely perpendicular horizontal components (“150deg” represents amplified mode 1 oscillation at a frequency of 1.83 Hz, “60deg” mode 2 at a frequency of 1.91 Hz). In colour: Synthetic seismograms modelled by the finite difference method (using IT4I). The amplification of waves at the MYT001 station is about 10 times. Modified from Labuta et al., 2025, Fig. 6 a

IT4Innovations supercomputers (Karolina, LUMI and Barbora) will help scientists at the Czech Academy of Sciences better understand seismic tremors that can threaten key infrastructure such as nuclear power plants, dams, power lines, and pipelines. The research will focus on evaluating and numerically modelling the local amplification of seismic tremors caused by the geological structure of the underlying bedrock and near-surface geological structures in basin areas of the Czech Republic considered for potential nuclear power plants.

This seismic amplification is one of the key factors affecting seismic risk. The project aims to develop an optimal methodology to assess this amplification and to investigate all seismotectonic risks. The results of the research will be combined with geological data and the conclusions of previous studies into a regional model allowing more accurate prediction of seismic tremors and analysis of other potential hazards. Wave propagation numerical modelling using advanced computational tools (IT4I) will play a crucial role specifically in simulating seismic scenarios leading to the evaluation of the impacts of geological conditions on the amplification of earthquakes. This approach will allow the evaluation of the wider social and economic impacts of earthquakes, which can endanger human lives but may also cause damage to critical infrastructure. This will enable more detailed analysis and effective planning for its protection.

Reference:
Labuta, M., Oprsal, I., Landa, D.-A., Burjánek, J., 2025. Ambient Vibrations of a Deep Maar Resonator. Soil Dynamics and Earthquake Engineering, 188, Part B, 2025, 109072. https://doi.org/10.1016/j.soildyn.2024.109072

Image copyrights belong to the respective research groups.

Read also:

Learn how to apply for LUMI’s resources in the Get started section.