Researcher: Dominik Legut
Project: Multiscale design of novel Rare Earth free permanent magnets
Allocation: 5 261 000 core hours
Abstract: Permanent magnets are a key technology for modern society with applications in air conditioning, mobility, or power generation. In state-of-the-art permanent magnets the atomic scale defects, like for instance in the grain boundary phase, have the most significant influence on the macroscopic properties (e.g. coercivity), but these effects are the least understood. In this project, we develop a quantitative theory of coercivity, taking into account the local atomic structure, the spatial variation of the intrinsic magnetic properties, and the physical microstructure of the magnet. To achieve this goal, we bridge the length scales between ab-initio, atomistic spin dynamics, and continuum micromagnetic simulations. Atomic defects at interfaces and grain boundaries will be considered already at the smallest possible length scale and former assumptions based on bulk material properties can thus be avoided. A quantitative description of the effect of defects at interfaces and grain boundaries, inaccessible experimentally, becomes accessible via a validated multiscale coercivity model.

 

Researcher: Andrzej Kadzielawa
Project: Determination of the k-dependent hybridization between 4f states and the conduction bands in CeCoIn5
Allocation: 1 147 000 core hours
Abstract: We model the band structure of CeCoIn5, a heavy-fermion superconductor with the Ce-In termination, to model a behavior of the electrons near the Fermi surface, as studied by angle-resolved photoemission spectroscopy (ARPES) at temperature of T=6 K. As the standard DFT methods of the Local Density Approximation (LDA) and Generalized Grandient Approximations (GGA) both with and without arbitrary onsite repulsion U (of Hubbard type) are failing to grasp the physics of itenerant and localized electrons near Fermi surface, we employ the Strongly Constrained and Appropriately Normed semilocal density functional (SCAN) metaGGA. This should give us inside into the physics behind the hybridization between localized f-electrons with the conduction band.

 

Researcher: Stepan Sklenak
Project: Periodic DFT studies of zeolite-based catalysts
Allocation: 962 000 core hours
Abstract: Zeolite based catalysts are the most important industrial catalysts used mainly in petrochemistry (production of fuels, various hydrocarbons), and furthermore, for N2O decomposition and N2O selective catalytic reduction (SCR), for direct NO decomposition as well as for SCR of NOx using various reducing agents. In addition, zeolites are also employed as catalysts in synthesis of fine chemicals and as well as in the processing of biomass. Zeolites are also used as ion exchange agents (e.g. they replaced harmful phosphates in washing powder detergents). Zeolites are crystalline microporous aluminosilicates with a unique microporous nature, where the shape and size of a particular pore system exerts a steric influence on the reaction, controlling the access of reactants and products. Thus, zeolites are often said to act as shape-selective catalysts. Increasingly, attention has focused on fine-tuning the properties of zeolite catalysts in order to carry out very specific syntheses of high-value chemicals. Periodic DFT methods permit investigations of properties of zeolite-based catalysts which are needed for their fine-tuning. DFT calculations are complementary to experimental examinations and together they can provide more complex knowledge of the properties of the studied catalysts and the reactions they catalyze.

 

Researcher: Karel Vyborny
Project: Screening of transport properties in selected magnetically ordered materials
Allocation: 1 178 000 core hours
Abstract: Transport of charge (electric conductivity) and spin will be studied by ab initio methods in materials relevant for antiferromagnetic spintronics. This field aims at designing useful electronic devices (such as memories) where spin apart from the electron charge plays a role and in particular, materials with magnetic order resulting in zero-net-magnetisation will be considered. Various deviations from  a perfect crystal (disorder, domain walls) will be taken into account and materials with most favourable characteristics will then be passed on to experimentalists. The project can be divided into two sections: a more targeted study of NaMnAs - a material already under experimental scrutiny - and a less detailed screening of about 60 magnetically ordered materials (mostly antiferromagnetic semiconductors) focusing mainly on spin currents and magneto-optical spectra. In the latter part, promising materials will be identified for further experimental research and in both parts, guidance for material optimisation (composition, domain structure) will be provided.

 

Researcher: Miroslav Rubes
Project: Understating the reactivity of the Brønsted acid sites in zeolites
Allocation: 1 345 000 core hours
Abstract: Protonic zeolites with Brønsted acid sites play crucial role in petrochemistry, and fine chemical production. Despite the intensive use of these materials as acid catalysts since the 1960s, the significant diversity of their catalytic activities is not fully understood yet. To improve the catalytic properties of existing catalysts as well as to perform the efficient and targeted development of new materials designed for specific catalytic applications, it is necessary to understand key factors controlling catalytic performances of these materials. The main purpose of this project is to model various Brønsted acid sites in different zeolites (i.e. environments) allowing the assessment of the material catalytic properties in the particular reaction from its framework structure and composition. 

 

Researcher: Azin Shahsavar
Project: Tunable Magnetic Interactions on Graphene
Allocation: 514 000 core hours
Abstract: In this project, we aim to provide a theoretical understanding of the electrical tuning of magnetic coupling between magnetic centers within metal-organic networks synthesized on graphene substrates. The metal-organic networks allow precise positioning of magnetic atoms into long-range-ordered lattices on the graphene substrate. We will exploit the possibility of an external tuning of the graphene Fermi level through a gate voltage and explore the possibility of modulation of the magnetic coupling between the magnetic centers. We aim at a theoretical description and understanding of the tunable magnetic interactions with the help of ab-initio DFT calculations. To provide a correct description, we will utilize a range of experimental data that will provide a solid background for theoretical calculations. The project is of great fundamental interest, as many conflicting predictions regarding strength and distance dependence of such indirect magnetic couplings exist in the literature, which are not easily discerned without the experimental data.

 

Researcher: Petr Strakos
Project: TACR Train Simulator
Allocation: 405 000 core hours
Abstract: Based on our background in computer visualizations and artificial intelligence we cooperate with Ixperta company in a research project that has been supported by the Technology Agency of the Czech Republic (TACR).
The project focuses on research and development of a functional sample of a railway vehicle with the ability to collect data and a software simulator with the ability to generate data for obstacle detection training in simulated conditions.
We specifically participate in the development of a software simulator to provide unique training data for the obstacle detection algorithm on a railway track.

 

Researcher: Georgiy Zadvitskiy
Project: Reflectometry synthetic diagnostic for tokamak experimental data interpretation and diagnostics design
Allocation: 935 000 core hours
Abstract: Nuclear fusion is a very promising source of clean and sustainable energy. Magnetic confined fusion on the base of the tokamak device has shown the most advanced results. In order to get strong power output of a fusion power plant one should minimize plasma power losses. One of the dominant mechanism which enhances energy and particle transport is turbulence. Based on radar technique, microwave reflectometry diagnostics are powerful tools that can provide an information about plasma turbulence. An interpretation of signals of these diagnostics is usually very tricky. Reflectometry synthetic diagnostic code will be used for experimental data analysis. This data will be then used for turbulence models validation, study of the transport and transition to advanced confinement regime.
COMPASS is a Czech tokamak that was operating from 2008 in Prague. Recently, fast swept reflectometry (FSR) and Doppler Back-Scattering(DBS)systems were used at experimental compagnes at Compass tokamak. Reflectometry synthetic diagnostics will be used to identify instrumental functions of these diagnostics and to help in data interpretation to retrieve turbulence properties.
Reflectometry synthetic diagnostics will be also used for COMPASS Upgrade tokamak reflectometer design. The beginning of COMPASS Upgrade (Compass-U) operation is planned in 2022. 

 

Researcher: Libor Veis
Project: New efficient computational methods for extended molecules with strongly correlated electrons
Allocation: 1 929 000 core hours
Abstract: There is no doubt that quantum chemical computations nowadays play a fully-fledged role in chemical research. Despite the huge progress in development of novel computational methods, there are, however, several classes of chemical problems which are still awaiting a more reliable theoretical description. One of those is the electronic structure of chemical systems suffering from quasidegeneracy, so called strongly correlated systems. Many important chemical problems fall into this category, the most challenging probably being the electronic structure of transition metal complexes with multiple transition metal atoms, such as iron-sulfur [Fe-S] clusters, important cofactors of different classes of metaloenzymes.
In this project, we would like to develop the new computational tools based on our recently developed massively parallel implementation of the density matrix renormalization group (DMRG) method, namely the dynamical correlation extensions by means of the adiabatic connection and DFT embedding and apply them on challenging real-world chemical problems such as the electronic structure of biologically relevant transition metal complexes mentioned above or polycyclic aromatic hydrocarbons possessing open shell electronic structure and featuring magnetic properties.

 

Researcher: Martin Kolisko
Project: Genomic investigation of organellar evolution and environmental adaptation in apicomplexan parasites
Allocation: 327 000 core hours
Abstract: Although much is known about the genetic makeup and organellar composition of core apicomplexan parasites, gregarines have been largely ignored despite being projected as one of the most species rich groups of eukaryotes. The aim of this project is to explore the biodiversity of this poorly studied group and genetically characterize its members. To accomplish this, we will sequence transcriptomes and genomes of understudied lineages including Neogregarinorida and Eugregarinorida from terrestrial and freshwater hosts. In combination with the recently published Archigregarinorida and Eugregarinorida data from marine hosts, we will produce a robust phylogeny of Apicomplexa and resolve two recently published conflicting results that have profound impacts on our understanding of evolutionary events within this group. With this evolutionary framework, we will explore the reductive evolution of organelles across Apicomplexa and identify genetic signatures for the adaptations to diverse host environments and the intracellular lifestyle observed in other apicomplexans.

 

Researcher: Marek Ingr
Project: Substituted hyaluronan molecules in aqueous and mixed solvents
Allocation: 486 000 core hours
Abstract: Hyaluronan (HA) is a key component of the extracellular matrix of skin and connective tissues. It finds wide applications in cosmetics and pharmacology as a compound supporting the tissue regeneration and wound healing. As a biocompatible material it is also used as a base of the drug-delivery systems, often used in various chemically modified forms. Although HA is highly hydrophilic, its technological applications, especially the formulations of medical and cosmetics products, often require its use in non-aqueous environment. In addition, the chemical modifications are often carried out in non-aqueous solvents, the composition of which may influence the reactivity of the functional groups of the HA molecule. In this project we propose to simulate the HA oligosaccharides substituted by aliphatic chains in water and solvents composed of water and organic liquids miscible with it by means of molecular dynamics (MD). Different numbers and configurations of the substituents will be used and their influence on the chain structure will be studied, at selected systems also interactions of two substituted HA molecules. Solvent-dependent differences will be investigated in the point of view of the chemical reactivity variations, but also the formation of secondary structures. The expected results should contribute to optimization of the syntheses of modified HA oligosaccharides and the design of different nanoscale systems with a remarkable application potential.

 

Researcher: Martin Friak
Project: Shape-memory Ni-Mn-Sn alloys as candidates for future smart materials
Allocation: 3 409 000 core hours
Abstract: Many current and future societal challenges are planned to be solved by applying so-called smart materials that are multifunctional, i.e. combine different properties so as to serve to multiple purposes. Our project focuses on Ni-Mn-Sn alloys as promising candidates for future multifunctional applications. The Ni-Mn-Sn alloys are magnetic Heusler-structure materials with shape-memory functionality and can possibly serve also as a source of spin-polarized current for spintronic devices. Their multi-functionality stems partly from their very complex nature that is so far not well understood. Our project will shed a new light on these complicated materials in order to enable and fine-tune their future use. 

 

Researcher: Rene Kalus
Project: Modeling of Cold-Plasma Transport Properties
Allocation: 880 000 core hours
Abstract: Nowadays, applications of low-temperature plasma are a topic of interest in areas like surface treatment [1], food industry [2], and medicine [3,4,5,6] with the last one motivating our current research efforts. This project will be a direct continuation of our previous research efforts, following up Van de Steen’s work [7,8,9,10] and several OPEN projects. As our current understanding of processes occurring in medicinal applications is not sufficient to tune the plasma for specific purposes[11], we aim to describe the processes in detail from the creation phase to the moment of application in the future. As the ionization and collisions of atomic and molecular ions have been covered already, we aim to continue with a) calculations of transport properties of collision complexes created by interactions of carrier-gas ions with the air and b) modeling of molecular ions formation. Both of these goals are tightly entangled with two double-degree Ph.D. theses co-directed by VSB-TUO and Université Toulouse III - Paul Sabatier, France, following the existing long-time collaboration of our team with LAPLACE, LCPQ, and LCAR research laboratories of the French university.

 

Researcher: Vladimir Ulman
Project: Fiji Bioimage Informatics on HPC - Path to Exascale
Allocation: 179 000 core hours
Abstract: “Bioimage Informatics on HPC” allows IT4Innovations to be involved in research on a completely new topical area of big biological image data processing on HPC. This specific research is focused on parallelization of key steps in lightsheet microscopy data processing as well as analysis of big data generated from other microscopic modalities. Particularly, multi-dimensional microscopy acquisitions present one of the main primary data sources in modern biological sciences and deployment of HPC in these areas is vastly unexplored approach to obtain biologically meaningful conclusions in a reasonable time.
This project is a continuation of the previous OPEN-19-3 call, solving particular VP3 subtasks of the Path to Exascale project. In the previous period, SPIM Workflow Manager was developed and published under Apache License. The project aims at further development and dissemination of HPC-aware plugins for the Fiji community. In this call we would like to specifically focus on utilization of a newly developed library for seamless parallel execution of SciJava plugins – SciJava Parallel and its connection to a new structured image data storage. As a new research topic we are working on a distributed version of a simulator of artificial time-lapse images from developmental biology – EmbryoGen.

 

Researcher: Valeria Butera
Project: Discovery of Novel Efficient Catalysts for CO2 Capture and in Situ Utilization: DFT investigation
Allocation: 179 000 core hours
Abstract: CO2 is considered the main culprit for global warming. Over half the CO2 emissions are from large, industrial point sources, while the remainder of these emissions are from small, mobile sources.  Reducing its emission below critical levels requires not only political commitment, but also novel scientific approaches to capture CO2 and to enable its conversion from a waste product into value-added products. The key measure to achieve this will be efficient and practicable processes for ambient air and large-scale CO2 sequestration and utilization. The main focus of this proposal is the development of innovative technologies aimed at slowing or stemming anthropogenic carbon emissions. Specifically, our challenge is the design of new efficient, selective and “green” homogeneous and heterogeneous catalysts for CO2 capture and CO2 conversion. We will first focus on materials that are suitable for CO2 sequestration and conversion directly from ambient air (direct air capture, DAC). The results of this investigation will pave the way for developing suitable technologies to extract CO2 from large industrial sources.

 

Researcher: David Wagenknecht
Project: Electrical transport and thermopower of spintronic materials at finite temperatures
Allocation: 965 000 core hours
Abstract: Everything around us is influenced by temperature – scientifically speaking, by finite temperature. For solids it means, that atomic nuclei are fluctuating around their equilibrium positions, which may have a huge impact on material behavior. For example, electrical resistivity of transition metals may increase by several orders of magnitudes when going from zero temperature (T → 0 K) to room conditions (approx. 300 K). This is well-know behavior, it can be easily measured for usual metals, but it is extremely difficult to calculate it, especially because of numerical expenses and because most of usually used codes are not prepared for such task. Therefore, how should one predict finite-temperature behavior of materials, which have not been even experimentally prepared yet?
Recently, we have implemented extremely efficient alloy analogy model (AAM) within our fully relativistic ab initio numerical codes. Most of the calculations of solid materials describe only zero temperatures, but our approach with the AAM reliably deals with finite temperatures.
We will use our codes to describe materials for spintronic devices, i.e., applications where the spin degree of freedom is considered. We will focus especially on electrical transport and thermoelectric power (Seebeck coefficient), which many codes (such as VASP) calculate only on very approximate level, but our efficient approach is much more suitable to investigate such effects at finite temperatures.

 

Researcher: Barnabas Barna
Project: Expansion of supernova driven shells near the galactic center
Allocation: 631 000 core hours
Abstract: Supernova (SN) explosions are some of the most energetic events in the Universe. These cataclysmic stellar deaths release high-temperature, high-velocity chemically enriched material which starts rapid expansion. The resulted cosmic bubble sweeps up the interstellar material (ISM). The formed shell containing gas of several hundreds of solar masses plays an important role in stellar feedback by triggering the formation of the next generation of stars. However, near the galactic center, the swept-up ISM may be delivered to the close vicinity of the central supermassive black hole (SMBH), triggering its activity. Whether or not this scenario is responsible for the past activity of the Milky Way's SMBH cannot be investigated directly due to the observational limitations and the rareness of these events. Thus, numerical methods are required to simulate the possible fate of a SN remnant exploring the effects of different initial setups. We are planning multiple hydrodynamic simulations with FLASH in 3D to investigate the impact of the interstellar environment on the expanding shell. The results will be matched with current observations, e.g. the current distribution of SN remnants and powerful X-ray flashes in the past.

 

Researcher: Josef Jon
Project: Massively multilingual and self-supervised neural machine translation for low-resource languages
Allocation: 1 193 000 core hours
Abstract: Machine translation of human language made a tremendous progress due to recent improvements in machine learning. Neural machine translation (NMT), based on sequence-to-sequence neural networks, is nowadays a largely predominant approach to this task. Recently, a number of evaluation studies demonstrated that translations generated by current systems are, under specific constraints, even comparable to a human translation. One of these constraints is that the language pair which is being used is high-resource, i.e. a large of parallel, human-translated texts are available for the given pair. This is true only for a very limited number of language pairs.
One of the possible approaches to improve  NMT performance on low-resource languages is to train a multilingual translation model which translates from and into multiple languages and transfers knowledge on how to process the text between them. Sharing of the model between multiple languages is beneficial for the translation quality of the low-resource pairs.
Another approach is inspired by the success of self-supervised language models. No manually translated data are needed, the model is trained only to predict next word in a corpus of large amounts of text from the Web. Recent work shows that these models perform surprisingly well for a number of various tasks, including machine translation. We aim to combine these approaches and study the effect of model architecture and pretraining objective choices on the translation performance.

 

Researcher: Lubomir Riha
Project: POP2
Allocation: 316 000 core hours
Abstract: The growing complexity of parallel computers is leading to a situation where code owners and users are not aware of the detailed issues affecting the performance of their applications. The result is often an inefficient use of the infrastructures. Even in the cases where the need to get further performance and efficiency is perceived, the code developers may not have insight enough on its detailed causes so as to properly address the problem. This may lead to blind attempts to restructure codes in a way that might not be the most productive ones.
POP2 extends and expands the activities successfully carried out by the POP Centre of Excellence since October 2015. The effort in the POP project resulted in more than 120 assessment services provided to customers in academia, research and industry helping them to better understand the behaviour and improve the performance of their applications. The external view, advice, and help provided by POP have been extremely useful for many of these customers to significantly improve the performance of their codes by factors of 20% in some cases but up to 2x or even 10x in others. The POP experience was also extremely valuable to identify issues in methodologies and tools that if improved will reduce the assessment cycle time. The objective of POP2 is to continue and improve the POP project operating a Centre of Excellence in Computing Applications in the area of Performance Optimisation and Productivity with a special focus on very large scale towards exascale. POP2 will continue the service-oriented activity, giving code developers, users and infrastructure operators an impartial external view that will help them improve their codes. We will still target the same categories of users and focus on identifying performance issues and proposing techniques that can help applications in the direction of exascale. Our objective is to perform 180 services over a three-year period.

 

Researcher: Miroslav Šoóš
Project: A molecular dynamics investigation of the desolvation of Ibrutinib solvates
Allocation: 1 178 000 core hours
Abstract: In drug manufacturing, active pharmaceutical ingredients (APIs) are commonly prepared by solvent-based methods, such as crystallization from a suitable solvent. Often, the solvent of crystallization becomes entrapped in the API crystal structure, forming a new solid compound called solvate. During the formulation of an API in the final drug product, an API can undergo heating (e.g. during drying, tableting, compaction, milling) resulting in the desolvation of the solvent molecules from the crystal structure. In such process, crystal structure transformation can occur resulting in the change of physicochemical properties of an API, such as its stability and solubility. In our research, we have been focusing on investigating the desolvation of solvates of Ibrutinib, an anticancer drug which is used to treat B cell cancers like mantle cell lymphoma, chronic lymphocytic leukemia, and Waldenström's macroglobulinemia. Recently, we have developed a methodology to calculate the kinetics of desolvation based on Molecular Dynamics (MD) simulations. We have used this method to research the kinetics of the desolvation of fluorobenzene (PhF) solvate of Ibrutinib and our results were in excellent agreement with the corresponding experimental kinetic data collected using classical TGA analysis but also our novel in-situ XRPD2. Therefore, in this work, we aim to apply our methodology to investigate the desolvation of several Ibrutinib solvates, such as m-xylene, anisole and methanol solvates, for which we have collected the experimental data by the above-mentioned techniques. For systems, which desolvation possess very high activation energy barriers we will utilize enhanced sampling techniques, such as metadynamics, and thus generalize the applicability of our methodology.

 

Researcher: Martin Matys
Project: Plasma shutter in laser-driven ion acceleration using targets with surface impurities
Allocation: 585 000 core hours
Abstract: Usage of plasma shutter in a form of a thin solid membrane provides a technique for improving laser pulse contrast by mitigation of prepulses accompanying the main laser pulse. The prepulses are filtered out as they need to burn through the plasma shutter before the interaction with a main foil. Moreover, a part of the main laser pulse is also spent during the process, resulting into a generation of a steep-front laser pulse. In our ongoing research we are investigating effects of this pulse shape modification on the laser-driven ion acceleration. We had demonstrated the increase of the maximal energy when the plasma-shutter is applied in our 3D simulations with single layer heavy ion target. In this project we will investigate a more realistic situation with impurities attached to the surface of the target. In experiments, the heavy ions are usually hard to accelerate to reasonable energies, as a large portion of laser pulse energy is usually absorbed into impurities, which are much lighter. The situation should be improved using the plasma shutter, as the steep-front is beneficial for the acceleration via radiation pressure, which can accelerate heavy ions in the bulk target more efficiently than other mechanisms.   

 

Researcher: Jan Psikal
Project: Studies of multidimensional effects in laser-driven ion acceleration
Allocation: 403 000 core hours
Abstract: Since the plasma provides much stronger electric fields compared with conventional particle accelerators (more than four orders of magnitude), acceleration of charged particles from plasmas produced by intense laser pulses is intensively studied in last two decades. With the continuous development of laser technology, laser facilities delivering femtosecond pulses (with pulse duration in tens of femtoseconds) of ultra-high peak power of several petawatts (PWs) have been constructed. These laser installations should enable to accelerate ions to energies about several hundreds of megaelectronvolt (MeV) per nucleon, suitable for many applications. The main laser pulse interaction with almost instantaneously ionized targets occurs during several tens of femtoseconds and experimental measurements cannot well resolve all processes during such short time interval.  Thus, numerical simulations are indispensable tool for this scientific research. These simulations are usually performed in simplified 2D geometry since real 3D geometry is very demanding on computational resources. However, such simulations in real 3D geometry enable to assess important multidimensional effects whose importance varies depending on target densities. Therefore, we plan to compare simulation results obtained in 2D and 3D geometries in a few representative cases in the frame of this project. 

 

Researcher: Lubomir Rulisek
Project: Fundamental Principles of Protein Folding and Protein-Ligand Interactions Revealed by High-Level Quantum Chemical Calculations
Allocation: 1 966 000 core hours
Abstract: To What Extent Conformational Strain in Proteins Determine Their Three-dimensional Structure? Large-scale quantum chemical calculations coupled with modern solvation methods represent unique set of tools to elucidate key determinants of the biomolecular structure ab initio. Understanding conformational strain in proteins and in their ligands may represent a new and computationally tractable way how to significantly deepen our understanding of protein folding and of protein-ligand interactions. The computed data will be correlated with conservation of amino acid residues to see what physico-chemical properties are retained by evolution. In addition to computed data, an experimentally verifiable set of tests is suggested to provide evidence for the proposed hypotheses.

 

Researcher: Ales Vitek
Project: Big water clusters under pressure
Allocation: 345 000 core hours
Abstract: In long-term collaboration of Moldyn group from IT4Innovations with Dr. Rita Prosmiti from PAMS-IFF-CSIC, Madrid, we developed methodology enable to compute full phase diagram and properties of finite sub-nano systems under various thermodynamic conditions via classical Monte Carlo methods. In our recent research [1], we compared different approaches how to simulate small molecular systems under high pressure and non-zero temperature. We have found significant differences between different volume (or pressure) approaches implemented in the Monte Carlo simulations. But computations have been performed for very small system, particularly for (H2O)12 water cluster. All different volume (or pressure) models differ from the each other’s. In the bulk, macroscopic limit, all this methods should exhibit the same results. In submitted project, we would like to explore the convergence of different volume (pressure) models with the increase of the size of simulated systems.

 

Researcher: Jana Pavlikova Precechtelova
Project: Towards the prediction of chemical shifts in intrinsically disordered proteins phosphorylated at tyrosine
Allocation: 285 000 core hours
Abstract: The objective of the present project is to develop a database of chemical shift (CS) patterns for tripeptide models of intrinsically disordered proteins (IDPs) phosphorylated at tyrosine residues. The database presents an essential building block in the development of a software (SW) tool for the prediction of NMR CSs. Prediction of CSs is a mandatory step in the structure characterization of IDPs by  NMR spectroscopy. The project designs SW that predicts CSs for phosphorylated residues based on CSs computed by quantum mechanical methods. Phosphorylated IDPs can influence regulation of neurodegenerative processes  that cause e.g. Alzheimer and Parkinson's diseases. This project helps to accelerate the research on neurodegenerative diseases and contributes to reducing the related research costs and thus it can improve the available medical care.

 

Researcher: Ondrej Chrenko
Project: Formation of dust rings and gaps in protoplanetary disks
Allocation: 355 000 core hours
Abstract: Recent high resolution imaging of protoplanetary disks has brought evidence that these disks often contain concentric rings and gaps in the dust distribution. Understanding the physical origin of dust rings is important because spontaneous concentrations of dust can trigger formation of larger objects such as planetesimals and, ultimately, planets. Here we aim to study if the ringed structure of protoplanetary disks can be explained by the dust-driven viscous ring instability. Such an instability is facilitated by a feedback loop between the dust-to-gas ratio, disk conductivity and turbulence-driven viscosity. The viscous ring instability has been investigated only mathematically and we therefore propose to study it numerically, using 2D hydrodynamic simulations of a two-fluid interacting mixture of gas and dust.

 

Researcher: Luigi Cigarini
Project: Interplay between electronic, conformational and phononic structure in the alpha form of the nitrogen-phosphorus (NP) two-dimensional material
Allocation: 123 000 core hours
Abstract: An important part of the modern Material Science research in atomistic simulations is based on the Born–Oppenheimer approximation, which allows to treat separately the electronic charge distribution in space with respect to the slower vibrations which involve the atomic nuclei (phonons). With this comes that the effect of the vibrational movements on the electronic distribution are generally neglected. Here, to overcome this, we want to study the temperature dependence of the electronic switching behaviour of a semiconductor material. 

 

Researcher: Martin Beseda
Project: Computations of [N2/He]+ potential energy surfaces 3
Allocation: 134 000 core hours
Abstract: This project is a supplement to an already accepted project DD-20-23. Due to the more demanding ab initio computations together with the sensitivity of LEVEL16 software to the grid density, we ask for another 350 000 core-hours, as the previous amount was already depleted. We aim to finish two last computations sets describing the N2+ PESs for the largest active space and AV5Z, AV6Z basis sets. The rest of the project form remains the same as the latest one, as the project is literally the same.
This project is a direct continuation to the  OPEN-17-15 project, especially to its part focused on computations of properties of the collision complex [N2/He]+.  The previous computations were not finished mostly due to a communication delays with our colleagues from Université Paul Sabatier, so this project should cover the next few months to finish the results now, before the start of new OPEN -20 projects. The results will be also a part both investigators’ doctoral theses.
Considering the supplementary character of the project we cite the relevant parts from OPEN-17-15 in the following parts.
Considering the former one , it was shown experimentally, that rare-gas plasmas are well-working in medical applications. To understand the healing properties of cold rare-gas plasmas, however, detailed knowledge of processes is of crucial importance. We are mainly focused on the interaction of He with N2, i.e. this part is a direct continuation of the OPEN-14-25 project. Considering the computational demands of [N2/He]+ complex, the project will contain a thorough convergence analysis, taking different active spaces, basis sets and point groups available for a  given geometry, into consideration.

 

Researcher: Jiri Klimes
Project: Accuracy and precision for extended systems VI
Allocation: 2 058 000 core hours
Abstract: Computer simulations have become indispensable for understanding the majority of experimental observations and thus understanding our world. This is especially true at the atomic and molecular level where experimental observations can become very difficult to interpret. However, if simulations are to be useful, they need to model the system reliably. For example, if there are different pathways for a chemical reaction, the simulation needs to be able to correctly predict their relative importance so as to increase our understanding of the system. The world at the atomic level is governed by the complex laws of quantum mechanics. In practice, this means that increasing the reliability of simulations, i.e., using less approximations, increases the computational cost substantially. While we understand well the impact of some of the approximations, there are some for which our knowledge is limited. In some specific cases, these can lead to unexpectedly large losses of the reliability of calculations. In our project we will analyse some of the problematic cases occurring when calculating interactions between molecules, either in small clusters, or in molecular crystals.

 

Researcher: Pavel Balaz
Project: Phase transitions in two-dimensional topological structures
Allocation: 346 000 core hours
Abstract: Phase transitions are best known to occur when solid material is melting and transforming into liquid, or a liquid changing into gas and vice versa. These phase transitions are usually described in three-dimensional (3D) space, where solid materials are formed by 3D lattices of atoms. The term phase transition, however, has more general meaning and can describe also phenomena occurring in two-dimensional (2D) lattices of magnetic moments, like the well known transition between non-magnetic and magnetic state of 2D Ising model. In more complex magnetic systems, featuring asymmetric exchange Dzyaloshinskii-Moriya interaction, formation of 2D topological vortex-like structures – so called skyrmions – has been predicted and observed experimentally. Under specific conditions (given by applied magnetic field and temperature) skyrmions can form a regular lattice resembling the one of solids but in two dimensions. Varying external conditions one can stabilize the skyrmion lattice or melt it into disordered liquid-like phase. Interestingly, 2D system can be found in phases and undergo phase transitions that do not appear in three dimensions. These are, for example, topological phase transitions described by Kosterlitz-Thouless theory awarded by Nobel Prize in 2016.
In this project, we shall inspect topological phase transitions in 2D systems of Neel-type skyrmions using precise atomistic spin models.

 

Researcher: Vojtech Mrazek
Project: Scalable neuro-evolutionary algorithms
Allocation: 140 000 core hours
Abstract: Due to their state-of-the-art performance in many applications, deep neural networks (DNNs) have become an essential part of resource-constrained embedded devices. These devices are equipped with specialized hardware accelerators to effectively perform the underlying operations of  DNNs, where the inference phase is of the highest importance. The accelerators have many architectural parameters that, when selected carefully, can lead to optimal performance and energy-/power-efficiency. Similarly, at the software-level, many different types of DNN architectures are available. The proper selection of the DNN structure and hyperparameters can lead to optimal accuracy and performance. This work aims at designing high-quality architectures of deep neural networks that take into account the parameters of the target hardware accelerator. We will propose a methodology for the automated design of DNN architectures that show excellent tradeoffs between accuracy and energy efficiency on a given accelerator. This work will focus on two main aspects of energy-efficient DNN accelerator design: computational data path optimization and memory subsystem organization. As the proposed design method is based on computationally expensive neuro-evolution, the use of a supercomputer is essential for its success. 

 

Researcher: Jakub Oprsteny
Project: CPU and GPU scaling of DFT calculations
Allocation: 135 000 core hours
Abstract: Actinide carbides are prospective nuclear fuels for the so-called generation VI nuclear reactors, because of their high thermal conductivity and more efficient heat transfer. Actinides contain 5f electrons, which are not completely localized nor completely itinerant. Such complex behavior is difficult to model computationally. In this study, we aim to compare the CPU and GPU-based electronic structure static calculations of  recently studied Th2C3 and U2C3, both for their unit cells and supercells with broken symmetry, which are commonly employed to study lattice dynamics. We will study not only the influence of different parallelization schemes, but also the influence of magnetism, spin-orbit coupling and other approaches in order to analyze their influence upon the speed of calculations, memory requirements and other. We also plan to study the possibilities of utilizing the linear response method in the calculation of the materials elastic properties.

 

Researcher: Karel Carva
Project: Unconventional Superconductivity and Magnetism at Interfaces Between Fe-Chalcogenide Thin Films and Insulators
Allocation: 228 000 core hours
Abstract: The recently discovered  superconducting phase at high temperatures (~100K) in tensile strained FeSe on SrTiO3, one order of magnitude higher than in bulk FeSe, is yet unexplained. At present this effect is attributed to the presence of interfaces of FeSe to non-conductive perovskite-type oxide substrates, however the physical mechanisms are under debate, and only few other substrates have been studied for comparison.
This project concentrates on calculating the influence of selected interface materials on the electronic properties of quasi-2D FeCh (Ch = Se, Te) systems. We will study magnetism, and electronic properties of FeCh sample systems by first principle calculation methods.

 

Researcher: Pavel Hobza
Project: IN SILICO DRUG DESIGN
Allocation: 5 331 000 core hours
Abstract: The goal of in silico drug design is to recognize ligands of a pharmaceutically relevant target and reliably describe their interactions. The accuracy of the predictions depends on the description of noncovalent interactions. An accurate description is demanding because it requires the use of quantum mechanics (QM). Semiempirical QM (SQM) methods make QM description of extensive systems feasible. However, a large number of approximations resulted in rather poor description of noncovalent interactions in original SQM methods such as PM6. We have developed1-4 the corrected SQM methods (e.g. PM6-D3H4X) with improved description of noncovalent interactions. We have successfully applied our scoring function (SF) based on the corrected SQM methods to ranking of inhibitors of various kinases,5-7 proteases,8-10 aldo-keto reductases11-13 and others.14-16  The full version of our SQM-based SF is well suited for activity ranking in small series. For virtual screening of thousands of drug candidates, we have developed an accelerated SQM-based SF called SQM/COSMO.17 We have  demonstrated that SQM/COSMO outperforms classical SFs in native pose identification,17-19 ligand ranking,20 and virtual screening.21,22 Currently, we examine the performance of  SQM/COSMO on an extended series of targets, and also its ability to identify hotspots in protein-protein interactions for peptidomimetic compound design (e.g. the insulin receptor-insulin complex).

 

Researcher: Jiri Jaros
Project: Photoacoustic tomography of the breast IV
Allocation: 240 000 core hours
Abstract: Photoacoustic tomography (PAT) is a biomedical imaging modality based on the photoacoustic effect. In PAT, non-ionizing laser pulses are delivered into biological tissues. Some of the delivered energy will be absorbed and converted into heat, leading to transient thermoelastic expansion and thus wideband ultrasonic emission in the low MHz range. The generated ultrasonic waves are detected by ultrasonic transducers and then analyzed to produce images. Since the optical absorption is closely associated with physiological properties, such as hemoglobin concentration and oxygen saturation, the PAT is used to visualize vasculature inside tumors with a very high resolution.
The purpose of this allocation is to reconstruct photoacoustic images of the breast of 20 patients scanned by the Pammoth imager.  The resolution, accuracy, sharpness, motion and noise artifact, and the depth of penetration will be investigated and optimized. This study is the fourth phase of the model validation and moves us towards the deployment in a real PAT system for breast mammography.

 

Researcher: Michael Komm
Project: Particle-in-cell simulations of the secondary emission and electron back-scattering in ITER
Allocation: 403 000 core hours
Abstract: The assembly of the ITER tokamak has recently started, with first plasmas expected in 2025. ITER will be the first tokamak to have larger power output than input, which will result in extreme loading of the plasma-facing components by impacting particle and heat fluxes. The temperature of plasma particles impacting the divertor of ITER can exceed 100 eV during crashes of edge-localised modes, which can result in significant secondary electron emission and electron back-scattering from the exposed surfaces. In this project, we aim to simulate such processes using particle-in-cell method, to improve understanding of the properties of plasma sheath under such conditions.

 

Researcher: Václav Čuba
Project: Molecular modelling and dynamics study of materials for use in controlled release fertilizers
Allocation: 123 000 core hours
Abstract: Introduction of "NPK" and other composite fertilizers made taking care of soil and plants much easier for farmers. These fertilizers increased yield but caused new, mainly, ecological problems. To solve ecological stresses, connected mainly to excess of Nitrogen in soil and underground water, new slow and controlled release fertilizers were developed. Slow release fertilizers copy the curve of plant nutrient uptake with their curve of nutrient release. Because of their ecological and natural release curves they are promoted by Japanese Ministry of agriculture, forestry and fishery to be adopted nationwide [1]. An issue with slow release fertilizers is that they usually carry only one nutrient and ideally should more. Expensive materials and complex technology are required to manufacture slow release fertilizers containing more nutrients.
This project is focused on finding and understanding release properties of materials that are usable as a base for slow release fertilizer. Release and binding of nutrients from selected materials will be studied by means of molecular modeling and molecular dynamics. This study will assist in preparing manufacturing methods for new advanced forms of slow release fertilizer. Additionally, release process to simulate behavior of such fertilizers will be studied and compared with already available ones.

 

Researcher: Marek Lampart
Project: Dynamics of a free-body colliding mechanical system with a friction
Allocation: 70 000 core hours
Abstract: The main aim of the project is to analyse the dynamic properties of a mechanical system described by a mathematical model. The investigated system consists of a cylinder freely joined on a horizontal string fixed on the wall and of a moving belt. Such a system with impacts and dry friction is an image of many industrial applications, like stones falling on a conveyor moving belt. The mathematical model of the system has two degrees of freedom from which one corresponds to the position of the cylinder centre and the second one to its angular rotation. The studied system is excited by a slider moving in the vertical direction and by impacts between the cylinder and the belt. As a main result, it will be observed that the cylinder exhibits movement with both regular (periodic) and irregular (chaotic) patterns depending on the excitation amplitude and frequency. The goal of the research is to qualify and quantify the movement character. For this purpose, the 0-1 test for chaos together with approximate entropy will be utilized to find the regions of parameters for which chaos or regularity will be observed.

 

Researcher: Martin Bouda
Project: Kruger Park Soil-Plant Hydrodynamics
Allocation: 35 000 core hours
Abstract: Kruger National Park in South Africa is home to iconic wildlife as well as a natural gradient in rainfall accompanied by a transition from forest to open savanna. Available soil moisture shapes the vegetation, but the reverse is also true: vegetation mediates land-atmosphere feedbacks in the flows of energy, carbon, and water. These feedbacks are not as yet well understood or implemented in global and regional climate models, in large part because small-scale (milimeter-meter) root water uptake processes can limit large-scale (kilometer) flows, making the problem computationally intractable. In this research, we will bridge the gap in scales using a new analytical scaling method, which will allow us to characterize root water uptake dynamics in the Kruger Park at vegetation scale. Our modelling effort will be based on Kruger Park field data from international collaborators. After establishing model parameters for long-term experimental sites, we will use the calibrated model to show how root-zone hydrodynamics affect large-scale flows of matter and energy throughout the Park. This will represent a significant advance for process-based land surface modelling, where soil-plant hydrodynamics is presently a chief source of error in predictions of land-atmosphere flows of heat, water, and carbon. The outputs of this research will contribute to international Earth system model development efforts and ultimately to the IPCC’s 7th Assessment Report on global climate change.

 

Researcher: Milan Lazecky
Project: Development and application of interferometric approaches exploiting Sentinel-1 data over Czechia
Allocation: 20 000 core hours
Abstract: Sentinel-1 satellite system continuously observes European countries in a relatively high revisit frequency of 6 days per orbital track. Given the Sentinel-1 configuration, most areas in Czechia are observed every 1–2 days by different tracks in a moderate resolution. This is attractive for various types of remote sensing analyses. A specific system has been already developed for these purposes at IT4Innovations, IT4S1. The starting point for processing is the original data provided by ESA in their Single Look Complex (SLC) level. These are specifically augmented to a new level of data, SLC-C that are processed over all Czech data and stored within a Czech nation-wide Sentinel-1 data storage, run at CESNET. The special data type is ready for fast interferometric and polarimetric analyses. So far the data were used directly into several open-source multitemporal interferometry techniques that helped generating full Czech map of terrain deformations and also providing fast on-demand analyses to identify dangerous displacements over Czech infrastructure, including mine-induced subsidence or landslides. Other research topics include detection of forest loss due to hurricane, delineation of flooded areas or an urban growth. This project brings perspectives for continuation of the system development and applications towards new experimental products and techniques and preparation for future satellite radar missions.

 

Researcher: Michal Kolar
Project: Computational protocol to study protein behavior in a confined space
Allocation: 75 000 core hours
Abstract: Proteins are key biomolecules that play a role in almost all processes in living organisms. For instance, proteins catalyze chemical reactions, transfer cellular signals, or form the cytoskeleton. During their lives, proteins occupy various environments. Most often, they are dissolved in water or lipid membranes, but they also interact with other proteins in a complex fashion. It is not rare that proteins exist in confined spaces such as channels, tunnels or cavities. Here we design, optimize, and test a computational protocol based on all-atom molecular dynamics simulations to study the entrance of proteins to the confined space.

 

Researcher: Jan Martinovic
Project: Weather and Climate Pilot (H2020 LEXIS Project)
Allocation: 58 000 core hours
Abstract: The increasing quantities of data generated by modern industrial and business processes pose enormous challenges for organizations seeking to glean knowledge and understanding from the data. Combinations of HPC, Cloud and Big Data technologies are key to meeting the increasingly diverse needs of large and small organizations alike. Critically, access to powerful compute platforms for SMEs - which has been difficult due to both technical and financial reasons - may now be possible.
The LEXIS (Large-scale EXecution for Industry & Society) project is building an advanced engineering platform at the confluence of HPC, Cloud and Big Data, which leverages large-scale geographically-distributed resources from existing HPC infrastructure, employs Big Data analytics solutions, and augments them with Cloud services. Driven by the requirements of several pilot testcases, the LEXIS platform relies on best-in-class data management solutions (EUDAT) and advanced, distributed orchestration solutions (TOSCA), augmenting them with new, efficient hardware and platform capabilities (e.g. in the form of Data Nodes and federation, usage monitoring and accounting/billing support). Thus, LEXIS realises an innovative solution aimed at stimulating the interest of European industry and at creating an ecosystem of organisations that benefit from the LEXIS platform and its well-integrated HPC, HPDA and Data Management solutions.
The consortium is thus developing a demonstrator with a significant Open Source dimension, including validation, test and documentation. It will be validated in large-scale pilots – in the industrial and scientific sectors (Aeronautics, Earthquake and Tsunami, Weather and Climate), where significant improvements in KPIs including job execution time and solution accuracy are anticipated. LEXIS will organise a specific call which will stimulate adoption of the project framework, increasing stakeholders’ engagement with the targeted pilots.
This proposal is specifically focused on computing resources about the Weather and Climate pilot. This pilot is aiming at facilitating a complex stack of weather-related computational models to improve the prediction of water-food-energy nexus phenomena and their associated socio-economic impacts. 

 

Researcher: Lukas Halagacka
Project: Training on optical modeling with CST microwave studio on IT4Innovations supercomputing facility
Allocation: 11 000 core hours
Abstract: CST Microwave studio is a complex numerical tool available on IT4I supercomputing clusters. With different solvers available, e.g. time-domain, frequency-domain, integral, and asymptotic, it enables to perform electromagnetic field simulations of devices with footprint size from nano-scale to macro-scale dimensions. Of course, it is impossible to perform simulation of large-scale object with nano-meter scale meshing. Therefore, approach multilevel simulation, where a results of simulated nanoobject in dense mesh is used as source for large scale simulation. From perspectives of our laboratory,  simulations of new antennas used for  generation and detection of THz electromagnetic field is the typical usage od the multilevel modelling. The dipole of the antenna has typical dimension of tens of microns while complete device with lenses, housing and PCB is the device with footprint in centimeter scale. This project is focused on training of end-users to be able to design  complex structures and to perform optical simulations on the IT4Innovations Supercomputing clusters.

 

Researcher: Tereza Ševčíková
Project: Identification of druggable gene targets in aggressive stage of multiple myeloma blood cancer
Allocation: 14 000 core hours
Abstract: Cancer is the second most important cause of death and morbidity in Europe and is generally developed by accumulation of genetic changes in different cell types. Treatment is designed to overcome negative effects of these aberrations. Thus, complex understanding of genetic background in cancer cells by modern laboratory and bioinformatics methods already contributed to improve treatment of many types of cancer. Multiple myeloma is a second most common form of blood cancer, which is currently incurable despite the development of novel drugs in the past years. Multiple myeloma has several stages with the extramedullary disease (EMD) being the advanced aggressive stage which now occurs more often, probably in association with application of modern treatment and extended survival of patients. Nevertheless, the current treatment strategies are not effective at this disease stage and better understanding is needed in order to develop more potent therapies. Genomics and transcriptomics could definitely contribute to novel important finding on this disease.

 

Researcher: Thibault Derrien
Project: FLAMENCO (First-principle investigation of transient excitation and optical properties of laser-irradiated materials)
Allocation: 2 332 000 core hours
Abstract: This proposal comes to computationally support the Horizon 2020 RISE “ATLANTIC” project No. 823897, that aims at joining several theoretical formalisms in view of improving the predictions capabilities in the field of laser processing of solids, with a consortium of scientists who have
pioneered these techniques. In this 4-years EU project (2019-2023), the HiLASE Centre is driving the effort for describing the excitation of the electrons in various laser-irradiated materials, the resulting change of optical properties of these materials, and the absorbed laser energy. By
transferring the insights gained from available first-principles microscopic descriptions [particularly from the time-dependent density functional theory (TDDFT)] to large-scale approaches, the project will guide the invention of novel usages of intense laser light for modifying and customizing materials. By shifting from an intuition-based try-and-error approach to a computer-guided investigation, this project will enable applications in photonics and plasmonics, but also will reveal unforeseen applications based on the laser processing of materials. Thanks to ATLANTIC, it will help to train other researchers in using advanced theoretical techniques adapted to the problems met in the growing engineering field of laser processing.

 

Researcher: Tomas Martinovic
Project: LEXIS/WP6
Allocation: 6 000 core hours
Abstract: Tsunamis can result in disasters with huge numbers of victims, including areas that have felt nothing of the earthquake generating the tsunami. The emergency response depends on a timely and accurate warning. To handle that, the LEXIS project (H2020 – EU) works on integrating real-time tsunami simulations with inundation into the emergency process. With the study performed within the 16th open call, the tsunami simulation with TsunAWI could be tuned to run in just a few minutes for regional setups.
TsunAWI is a tsunami simulation code that employs unstructured triangular meshes, which allow to increase the resolution at the coast, while it requires only a coarse resolution to resolve the long tsunami wave in the deep ocean. The code was employed to compute a database of scenarios for the Indonesia Tsunami Early Warning System. The research in LEXIS allows us to take the step to real-time computation.
Within the present 20th call, the challenge is to provide warning information on a larger scale. Further optimization is needed to tackle larger regions, e.g. the whole coast of Chile, compared to a single city as it was the case in the 16th call, and a number of possible optimizations have to be explored and evaluated on both the source code and the datasets themselves. Furthermore, the extent of inundation events depends strongly on the tsunami source, however, in the first minutes after an earthquake, only little is known about the exact nature of the source. We have to set up a probabilistic overview of the possible tsunami inundation by calculating and combining about 10 to 100 possible scenarios for an event.

 

Researcher: Dominika Maslarova
Project: Optical injection improvement in laser wakefield accelerators
Allocation: 755 000 core hours
Abstract: Laser wakefield acceleration (LWFA) is currently considered as one of the most promising mechanisms to potentially reduce the size and cost of future electron accelerators. In this technique, plasma electrons are injected into a plasma wave (wakefield), generated and dragged by an ultra-short, ultra-intense laser pulse in optically transparent plasma. Such electrons gain relativistic energy within a few millimeters, which is three orders of magnitude lower than the contemporary technology. In order to achieve a full introduction for practical applications, which requires very precise reproducibility of electron beam properties (e. g. specific values of charge and energy spectra), these accelerators need further improvement. The aim of this project is to study the generation of wakefield by two separate laser pulses. The research will be carried out by numerical particle-in-cell simulations. The aim is to design and optimize experiments which are currently carried out in Extreme Light Laboratory in The University of Nebraska–Lincoln in the USA. The simulations will also help to analyze the details of experimental results, which cannot be directly observed and analyzed by temporary experimental diagnostics.    

 

Researcher: Marek Lampart
Project: Hidden attractors of the Cournot Oligopoly model
Allocation: 21 000 core hours
Abstract: The main aim of the project is to analyse the dynamic properties of the Cournot oligopoly model introduced by Baiardi-Naimzada. This heterogeneous discrete two-dimensional model describes the decision mechanism of players. More precisely, there are constructed two types of quantity settings characterized by different decision mechanisms that coexist and operate simultaneously. The goal of the research is to qualify and quantify the movement character. For this purpose, the 0-1 test for chaos together with maximal Lyapunov exponent will be utilized to find the regions of parameters for which chaos or regularity will be observed. Moreover, the goal will be focused on hidden-attractors finding and also multi-stability detection

 

Researcher: Radim Blaheta
Project: Development and testing of parallel solver for hydro-mechanical problems in fractured porous media
Allocation: 35 000 core hours
Abstract: The project is motivated by a strong interest in the numerical analysis of flow in fractured porous media, e.g. rocks in geo-engineering applications. The main application is the modelling of the flow of water in the area of the nuclear waste repository in deep geological formations. Detailed knowledge of the flow around the repository will allow modelling of a possible spreading of contamination through the rock massif. Such results can be used as an input to the overall safety assessment of the repository. Development of methods that allow such complex modelling is one of the goals of the H2020 EURAD project as well as the TACR ENDORSE project.
The numerical models need sufficiently fine discretization to provide precise results and the modelling of the repository itself needs to capture a relatively large area with fractures and failure zones that together demand high computational power. Moreover, inverse and uncertainty quantification problems that are governed by the repository model typically require many evaluations of the model. Thus, parallel computation and efficient computational methods are needed for the proposed numerical modelling.

 

Researcher: Judita Nagyova
Project: Movement characteristics of a non-smooth model with a closed curve equilibrium
Allocation: 22 000 core hours
Abstract: The main aim of this study is to analyze the dynamical properties of a model with a closed curve equilibrium. The corresponding three-variable model is given as a set of nonlinear ordinary differential equations containing non-smooth functions. The dynamics of the model are studied depending on three parameters. For this purpose, new methods, as the 0-1 test for chaos and approximate entropy, are applied. Using these tools, the dynamics are quantified and qualified. It will be shown that depending on the system's parameters, the system exhibits both irregular (chaotic) and regular (periodic) character.

 

Researcher: Radek Vitasek
Project: Changes in aortic haemodynamics leading to aneurysm formation and rupture
Allocation: 19 000 core hours
Abstract: The vascular tree is used for the distribution of oxygenated blood that comes out of the heart during heart cycle. Due to the elasticity of the walls, the arteries can reduce the pressure to such an extent that a constant blood flow occurs at the periphery of the vascular tree. However, over time, arterial walls become stiff, resulting in a change in the shape and amplitude of the pulse wave, which is directly related to blood pressure which has been strongly connected to various vascular diseases. Hypertension has been shown to cause various types of diseases such as coronary heart disease, heart failure, peripheral arterial disease, stroke, renal disease or abdominal aortic aneurysm (AAA).
AAA is a permanent diameter enlargement of abdominal aorta. Its rupture is a life-threatening event. On the other hand the non-ruptured AAA does not bring any complication to patient in most cases, therefore operations should be done only in cases with AAA close to rupture. Mostly used criterion for operation is maximum diameter. However it is not very accurate since not all large AAAs rupture while some small does. Therefore new criterions have been searched by scientific teams. One of the criterions is based on a wall stress analysis on the AAA. This project aims at analyzing of blood flow and pressure to inspect AAA creation and to analyze different mean arterial pressures and wall thickness affecting AAA.

 

Researcher: Miroslav Jícha
Project: Influence of infants and children lungs development on flow field and aerosol deposition–computational modelling using lattice Boltzmann technique
Allocation: 21 000 core hours
Abstract: The project will focus on elucidation of mechanisms for penetration of ambient and pharmaceutical aerosols in the infants´ and children´s respiratory tract aged several weeks, months and years till 5 years. Upper and lower airways (mouth, nose, trachea and tracheobronchial tree) that develop during maturation and influence heavily transport and deposition of aerosols will be considered. Image processing will be developed to transfer CT scans to digital and physical models. Lattice Boltzmann technique with LES and DNS approach will be further developed for steady inhalation and RANS models for turbulent flow at full inhalation profile with anisotropy corrections. Concerning particles tracking, Euler-Euler approach will be developed and compared with Euler-Lagrange. Lattice Boltzmann efficacy will be compared with CFD. Computational models will be validated by experiments on 3D printed models in a unique simulator using LDV/PDA and for deposition fluorescence will be used for full inhalation profile and PET for steady inhalation.

 

Researcher: Jakub Beranek
Project: Benchmarking distributed task workflow frameworks
Allocation: 50 000 core hours
Abstract: Our main objective is to benchmark popular distributed task workflow frameworks, such as Dask, Ray, or Spark on a range of tasks, primarily focusing on tabular SQL-like operations on dataframes. Our aim is to find what are the performance bottlenecks of these frameworks on HPC infrastructure. We want to leverage this information in our work and publications on improving the performance of distributed task frameworks and to select the most appropriate task frameworks for future IT4Innovations projects that will require distributed task processing.

 

Researcher: Pablo Nieves
Project: Magnetostriction beyond cubic and hexagonal systems
Allocation: 3 346 000 core hours
Abstract: Magnetostriction is a physical phenomenon in which the process of magnetization induces a change in shape or dimension of a magnetic material. Nowadays, in many technical applications such as for electric transformers, motor shielding, and magnetic recording, magnetic materials with extremely low magnetostriction are required. By contrast, materials with large magnetostriction are needed for many applications in electromagnetic microdevices as actuators and sensors. Until recently, the crystalline magnetostrictive materials of scientific and practical interest have come mainly from the cubic and hexagonal systems, e.g. cubic RFe2 and RCo2 Laves phases (R=Rare-Earth) or hexagonal RCo5 and R2Co7. However, compounds beyond cubic and hexagonal systems with high Curie temperatures and large magnetic anisotropy could be very promising magnetoelastic materials too. In this proposal, we aim to explore the magnetostriction of some trigonal and tetragonal systems by using our new developed software MAELAS.

 

Researcher: Matus Dubecky
Project: Many-body physics of 2D materials
Allocation: 3 283 000 core hours
Abstract: This projects focuses on development and application of stochastic computational protocols for benchmark quantum electronic structure computations of key  physical properties (thermodynamic stability, optical and fundamental gap, adsorption properties, electron correlation effects) of promising 2D materials, like, e.g., graphene derivatives (fluorographene), MXenes, and beyond.

 

Researcher: Stanislav Zalis
Project: QM/MM molecular dynamics simulations of photochemical activation of redox centers in proteins containing organometallic photosenzitizers.
Allocation: 917 000 core hours
Abstract: Redox reactions (electron transfer/ET) play a key role in bioenergetic processes (respiration, photosynthesis) and in large number of enzymatic reactions in living organisms e.g. synthesis of RNA, oxidation of toxic agents etc.[ Kretchmer]. The number and sequence of redox active amino groups strongly influence ET in protein chains.  In the frame of MSMT project Inter-Excelence  LTAUSA18026 “Photoactivation of protein redox sites“ ET rates have been studied experimentally by time-resolved techniques (FTIR, resonance Raman) at fs-ns scale. In order to understand a mechanism of these processes, experimental study needs strong support of theoretical computational research using molecular dynamics (MD). MD calculations of systems containing several redox active amino acids require simultaneous treatment of several excited states. Therefore, the use of quantum chemical – molecular mechanical methods (QM/MM) with QM part described by time-dependent DFT (TD-DFT) technique is necessary.
With our recently utilized QM/MM/MD approach [Takematsu ] we will investigate the electron transfer properties for different redox pathways of experimentally studied blue copper protein mutants with attached Re photosensitizer.  Relaxation from higher excited states to the lowest one (<500fs after excitation) will be newly described by MD, including surface hopping in the adiabatic representation [Mai]. Special attention will be paid to explain how environment (solvent, protein matrix) influence the electron transfer properties.

 

Researcher: Sergiu Arapan
Project: Exploring the magneto-crystalline anisotropy of the Laves phase based (Fe,B)_{2+x}A_{1-x} ternary alloys
Allocation: 1 717 000 core hours
Abstract: Permanent magnets are an indispensable part of the modern technology. As data storage components, they are present in smartphones, laptops, audio and video devices. Recently, however, they are mainly used in industrial applications, such as electrical motors and wind turbines, for energy conversion. A good PM must have a large coercivity H_{c} to resist demagnetization. It is determined by the magneto-crystalline anisotropy energy (MAE), the energy to rotate the magnetization from the easy to the hard direction. In a recent work we demonstrated a computational approach to predict hard magnetic phases in the Fe-Ta binary system. Interestingly, some predicted Fe_{1-x}Ta_{x} structures resemble the Fe_{2}Ta Laves phase with some partial occupancy of the Fe sites with Ta. This fact shows some similarities with the effect of enhancing MAE in Fe_{2}Hf, by substituting some sites of Fe with Sb. In this project, we propose a systematic investigation of the MAE of Fe_{2}A Laves phases alloyed with a substitutional element B, where A = Ta, Hf, Nb, Mo, Cr, Zr and B = Sb, Sn,  by using electronic structure calculations based on the density functional theory.