ZOZNAM ÚSPEŠNÝCH PROJEKTOV VO VÝZVE 1/23:
Development of predictive models of NMR shielding of ions in liquids
AUTOR: Andrej Antušek
Accurate ab initio NMR shielding constants allow to establish absolute shielding scales for NMR active nuclei and combined with experimental NMR resonance frequencies they lead to refinement of nuclear magnetic dipole moments and subsequently to the possibility of direct measurement of NMR shielding instead of chemical shift. Most of magnetic dipole moments for stable nuclei collected in a recent Stone’s (International Atomic Energy Agency) Nuclear Data Tables are based on our calculations. As a natural extension of our work we started collaboration with CERN ISOLDE laboratory on measurements of nuclear magnetic dipole moments of beta-decaying nuclei using beta-NMR experimental techniques. This collaboration led to the first measurement of nuclear magnetic dipole moment for short lived nucleus 26Na with ppm accuracy (four order of magnitudes higher accuracy than
previously). We plan to re-fine magnetic moments for other beta-NMR probe nuclei potassium and beryllium. To this end we will develop predictive models of NMR shielding of beta-NMR probe ions in ionic liquids – which are materials
used in beta-NMR experiments because of very low pressure of vapors – two ionic liquids are typically used EMIM-DCA and BMIM-HCOO. The development based on our previous work consists two parts: 1. structural part in which we will model the structure of beta-NMR probe ion solvation shells in ionic liquids using classical force field molecular dynamics. 2. NMR part in which we will
prepare tailored DFT functionals fitted to benchmark values of NMR shielding based on non-relativistic coupled cluster calculations and relativistic four-component DFT calculations.|basic_html|Accurate ab initio NMR shielding constants allow to establish absolute shielding scales for NMR active nuclei and combined with experimental NMR resonance frequencies they lead to refinement of nuclear magnetic dipole moments and subsequently to the possibility of direct measurement of NMR shielding instead of chemical shift. Most of magnetic dipole moments for stable nuclei collected in a recent Stone’s (International Atomic Energy Agency) Nuclear Data Tables are based on our calculations. As a natural extension of our work we started collaboration with CERN ISOLDE laboratory on measurements of nuclear magnetic dipole moments of beta-decaying nuclei using beta-NMR experimental techniques. This collaboration led to the first measurement of nuclear magnetic dipole moment for short lived nucleus 26Na with ppm accuracy (four order of magnitudes higher accuracy than
previously). We plan to re-fine magnetic moments for other beta-NMR probe nuclei potassium and beryllium. To this end we will develop predictive models of NMR shielding of beta-NMR probe ions in ionic liquids – which are materials
used in beta-NMR experiments because of very low pressure of vapors – two ionic liquids are typically used EMIM-DCA and BMIM-HCOO. The development based on our previous work consists two parts: 1. structural part in which we will model the structure of beta-NMR probe ion solvation shells in ionic liquids using classical force field molecular dynamics. 2. NMR part in which we will
prepare tailored DFT functionals fitted to benchmark values of NMR shielding based on non-relativistic coupled cluster calculations and relativistic four-component DFT calculations.
Structure prediction of Silicon-rich quasicrystal
AUTOR: Marek Mihalkovič
Recently discovered Silicon-rich quasicrystal with composition Si61Cu30Ca7Fe2 that apparently formed during atomic bomb test at Alamogordo, represents chemically new type of quasicrystal with surprising 61% content of Silicon. So far, no information about its detailed atomic structure is available, and presumably will not be available until the phase can be prepared in laboratory. Goal of our project is to predict the structure from the first principles, by educated-guess modelling, and by robust atomistic simulations.|basic_html|Recently discovered Silicon-rich quasicrystal with composition Si61Cu30Ca7Fe2 that apparently formed during atomic bomb test at Alamogordo, represents chemically new type of quasicrystal with surprising 61% content of Silicon. So far, no information about its detailed atomic structure is available, and presumably will not be available until the phase can be prepared in laboratory. Goal of our project is to predict the structure from the first principles, by educated-guess modelling, and by robust atomistic simulations.
hBN quantum Monte Carlo study of smallest benchmark for 2D vdW bonded materials
AUTOR: Ján Brndiar
Hexagonal Boron nitride (hBN) belongs to the van der Waals (vdW) bonded two-dimensional (2D) materials which have already revolutionized science due to their unique electronic properties ranging from metallic to insulating one. hBN has been proposed for use as an atomic flat insulating substrate or tunneling dielectric barrier based on a large and stable band gap (>7 eV for monolayer). As one of the simplest materials, similar to graphene, hBN can provide a wide range of benchmarking potential to better understand microscopic description of screening properties of those 2D vdW materials which is a key theoretical construct to explain and tune them on desire. We propose to deliver state of the art quantum Monte Carlo (QMC) results for distance dependent vdW bonding energetics in excited states of bilayer hBN using standard diffusion Monte Carlo (DMC) approach. As a consequence, this data will be used to explore the possibility to use machine learning approaches using FERMI net like artificial neural networks to speed up research in solving such many-body electronic problems.|basic_html|Hexagonal Boron nitride (hBN) belongs to the van der Waals (vdW) bonded two-dimensional (2D) materials which have already revolutionized science due to their unique electronic properties ranging from metallic to insulating one. hBN has been proposed for use as an atomic flat insulating substrate or tunneling dielectric barrier based on a large and stable band gap (>7 eV for monolayer). As one of the simplest materials, similar to graphene, hBN can provide a wide range of benchmarking potential to better understand microscopic description of screening properties of those 2D vdW materials which is a key theoretical construct to explain and tune them on desire. We propose to deliver state of the art quantum Monte Carlo (QMC) results for distance dependent vdW bonding energetics in excited states of bilayer hBN using standard diffusion Monte Carlo (DMC) approach. As a consequence, this data will be used to explore the possibility to use machine learning approaches using FERMI net like artificial neural networks to speed up research in solving such many-body electronic problems.
Quantum-Enhanced Drug Discovery: HPC-driven Innovations
AUTOR: Marek Štekláč
In the quest to address the evolving challenges posed by infectious diseases, cancer, and other diseases, the field of drug discovery has seen a significant shift towards harnessing the power of high-performance computing (HPC), allowing for implementation of quantum chemistry approaches. Utilizing HPC resources accelerates the evaluation of a vast space of chemical compound space, significantly increasing the efficiency of drug development while simultaneously reducing costs and promoting the discovery of novel life-saving medications.
The primary objective of the proposed project is to bridge the gap between theoretical quantum chemistry and practical drug development by harnessing the computational power offered by modern HPC systems. This novel approach utilizes quantum chemistry methods, such as density functional theory and molecular dynamics simulations, to provide a deeper understanding of molecular systems, protein-ligand interactions, and biological binding mechanisms.
Proposed project exploits state-of-the-art computational chemistry software that enables the comprehensive exploration of complex biological systems, paving the way for advanced simulations, the analysis of electronic properties, and the prediction of reaction pathways. Through virtual screening, molecular docking, and the calculation of electronic properties, the project seeks to identify promising drug candidates with optimal therapeutic efficacy and minimal side effects.|basic_html|In the quest to address the evolving challenges posed by infectious diseases, cancer, and other diseases, the field of drug discovery has seen a significant shift towards harnessing the power of high-performance computing (HPC), allowing for implementation of quantum chemistry approaches. Utilizing HPC resources accelerates the evaluation of a vast space of chemical compound space, significantly increasing the efficiency of drug development while simultaneously reducing costs and promoting the discovery of novel life-saving medications.
The primary objective of the proposed project is to bridge the gap between theoretical quantum chemistry and practical drug development by harnessing the computational power offered by modern HPC systems. This novel approach utilizes quantum chemistry methods, such as density functional theory and molecular dynamics simulations, to provide a deeper understanding of molecular systems, protein-ligand interactions, and biological binding mechanisms.
Proposed project exploits state-of-the-art computational chemistry software that enables the comprehensive exploration of complex biological systems, paving the way for advanced simulations, the analysis of electronic properties, and the prediction of reaction pathways. Through virtual screening, molecular docking, and the calculation of electronic properties, the project seeks to identify promising drug candidates with optimal therapeutic efficacy and minimal side effects.
Integrated geophysical modeling of geological structures
AUTOR: Ján Vozár
Due to the low number of existing deep boreholes, geophysical data, describing various physical properties of rocks, are key to understanding structures the earth’s crust or the lithosphere. Integrated modelling and interpretation of geoelectric, gravimetric, magnetic, and other data enables more accurate verification of hypotheses about tectonic and geological structure or information about raw material deposits and geothermal resources. We will calculate 2D and 3D distribution of parameters such as conductivity, density and magnetisation within the Earth independently for each sets of surface geophysical datasets (magnetotelluric, gravimetric and magnetic data). By combining geophysical data, we will focus on eliminating resolution inaccuracies in the models as well as the ambiguities of individual methods. With this approach, we achieve a more comprehensive picture of the geological structures, mainly in the areas of the flysch belts, the Neogene structures, crystalline structures, and faults in Slovakia and Europe. The latest modelling tools will be used in the selected locations, from simple comparison of models to fully automatic inversion of multiple data with coupled models of parameters.|basic_html|Due to the low number of existing deep boreholes, geophysical data, describing various physical properties of rocks, are key to understanding structures the earth’s crust or the lithosphere. Integrated modelling and interpretation of geoelectric, gravimetric, magnetic, and other data enables more accurate verification of hypotheses about tectonic and geological structure or information about raw material deposits and geothermal resources. We will calculate 2D and 3D distribution of parameters such as conductivity, density and magnetisation within the Earth independently for each sets of surface geophysical datasets (magnetotelluric, gravimetric and magnetic data). By combining geophysical data, we will focus on eliminating resolution inaccuracies in the models as well as the ambiguities of individual methods. With this approach, we achieve a more comprehensive picture of the geological structures, mainly in the areas of the flysch belts, the Neogene structures, crystalline structures, and faults in Slovakia and Europe. The latest modelling tools will be used in the selected locations, from simple comparison of models to fully automatic inversion of multiple data with coupled models of parameters.
Atmospheric and material chemistry modelling
AUTOR: Ivan Černušák
The goal of the project is to deepen our understanding of the role of mercury in molecular complexes of biological importance and their reactivity towards atmspheric oxidants, addressing a critical knowledge gap in environmental chemistry. Relatedly, all-carbon-atomic rings, cyclo[n]carbons, have recently attracted vivid attention of experimentalists and theoreticians. Among them, cyclo[18]carbon is the most studied system. In this project we plan to study the possibility to trap mercury in the carbon ring and explore the structure and dynamics of catenanes built of cyclo[n]carbons under external tension at finite temperature using DFT molecular dynamics within the NVT ensemble. Energy profiles of cyclo[n]carbons when exposed to external force, as well as the tensile properties like – specific strength, specific stiffness, and tensile stiffness will be studied.|basic_html|The goal of the project is to deepen our understanding of the role of mercury in molecular complexes of biological importance and their reactivity towards atmspheric oxidants, addressing a critical knowledge gap in environmental chemistry. Relatedly, all-carbon-atomic rings, cyclo[n]carbons, have recently attracted vivid attention of experimentalists and theoreticians. Among them, cyclo[18]carbon is the most studied system. In this project we plan to study the possibility to trap mercury in the carbon ring and explore the structure and dynamics of catenanes built of cyclo[n]carbons under external tension at finite temperature using DFT molecular dynamics within the NVT ensemble. Energy profiles of cyclo[n]carbons when exposed to external force, as well as the tensile properties like – specific strength, specific stiffness, and tensile stiffness will be studied.
Simulation of structural phase transitions in crystals by metadynamics
AUTOR: Roman Martoňák
The project aims at study of structural phase transitions in crystals by metadynamics. We plan to study mainly transitions induced by pressure (compression and decompression) where strong kinetic effects are often observed in experiments. The main focus will be on important elements silicon and carbon, including the study of long predicted but never observed post-diamond phases of carbon. Important part of the project is the study of nucleation, including homogeneous nucleation in ideal crystal and heterogeneous one in systems with defects, such as grain boundaries and dislocations. The project has also a strong methodological aspect consisting in identification and optimization of suitable collective variables for particular transitions. The results are likely to be useful in fundamental solid state physics and in materials science, and possibly also in Earth and planetary science.|basic_html|The project aims at study of structural phase transitions in crystals by metadynamics. We plan to study mainly transitions induced by pressure (compression and decompression) where strong kinetic effects are often observed in experiments. The main focus will be on important elements silicon and carbon, including the study of long predicted but never observed post-diamond phases of carbon. Important part of the project is the study of nucleation, including homogeneous nucleation in ideal crystal and heterogeneous one in systems with defects, such as grain boundaries and dislocations. The project has also a strong methodological aspect consisting in identification and optimization of suitable collective variables for particular transitions. The results are likely to be useful in fundamental solid state physics and in materials science, and possibly also in Earth and planetary science.
Finite temperature ab initio calculations of materials properties and catalytic processes
AUTOR: Tomáš Bučko
In this project, we will use a cutting-edge methodology based on first principles of physics and machine learning (ML) to efficiently predict the thermodynamic properties of technologically important materials and molecular systems. Problems to be addressed include investigating the catalytic transformation of isobutanol to butenes, determining the electronic structural properties and relative stability of chalcogenide perovskites, calculating the free energy and heath of adsorption of small molecules in zeolites or on the surfaces of 2D materials, and defining a CCSD(T)-quality finite-T benchmark set to test the accuracy of electronic structure methods.|basic_html|In this project, we will use a cutting-edge methodology based on first principles of physics and machine learning (ML) to efficiently predict the thermodynamic properties of technologically important materials and molecular systems. Problems to be addressed include investigating the catalytic transformation of isobutanol to butenes, determining the electronic structural properties and relative stability of chalcogenide perovskites, calculating the free energy and heath of adsorption of small molecules in zeolites or on the surfaces of 2D materials, and defining a CCSD(T)-quality finite-T benchmark set to test the accuracy of electronic structure methods.
Dicsovery of Antiviral Compounds Against SARS-CoV-2 Blocking Viral RNA Synthesis
AUTOR: Vladimír Frecer
Due to recurring outbreaks of highly transmissible infectious diseases caused by coronaviruses, the identification of new antiviral therapeutics should be a global research priority of medicinal chemistry. The proposed project focuses on the basic research of new antiviral agents against the SARS-CoV-2 coronavirus with a new mechanism of action. The goal of the project is to design and optimize inhibitors of the viral enzyme guanine-N7-methyltransferase, a key component of the life cycle of coronaviruses, and a validated pharmacological target. The designed inhibitors will derive from known bisubstrate sulfonamide (BSSA) analogs of S-adenosyl-L-methionine (SAM) extended by a nucleobase isostere occupying the RNA cap-binding site, which were shown to block the synthesis of viral RNA. The project uses a modern rational molecular design approach based on computer-aided combinatorial chemistry methods, molecular modeling, and optimization of active molecules, based on available 3D crystal structures of the viral receptor protein-ligand complexes and quantitative structure-activity relationships (QSAR). The proposed project will use advanced computer simulations (molecular dynamics) and quantum chemical methods (QM/MM) to calculate the binding affinities of new molecules to the viral receptor and evaluate the stability of drug-receptor complexes involving reversible (non-covalent) inhibitors. The expected result of the project is the identification of new viral guanine-N7-methyltransferase inhibitors that are more potent than known BSSAs and can lead to the development of new antiviral therapeutics against SARS-CoV-2.|basic_html|Due to recurring outbreaks of highly transmissible infectious diseases caused by coronaviruses, the identification of new antiviral therapeutics should be a global research priority of medicinal chemistry. The proposed project focuses on the basic research of new antiviral agents against the SARS-CoV-2 coronavirus with a new mechanism of action. The goal of the project is to design and optimize inhibitors of the viral enzyme guanine-N7-methyltransferase, a key component of the life cycle of coronaviruses, and a validated pharmacological target. The designed inhibitors will derive from known bisubstrate sulfonamide (BSSA) analogs of S-adenosyl-L-methionine (SAM) extended by a nucleobase isostere occupying the RNA cap-binding site, which were shown to block the synthesis of viral RNA. The project uses a modern rational molecular design approach based on computer-aided combinatorial chemistry methods, molecular modeling, and optimization of active molecules, based on available 3D crystal structures of the viral receptor protein-ligand complexes and quantitative structure-activity relationships (QSAR). The proposed project will use advanced computer simulations (molecular dynamics) and quantum chemical methods (QM/MM) to calculate the binding affinities of new molecules to the viral receptor and evaluate the stability of drug-receptor complexes involving reversible (non-covalent) inhibitors. The expected result of the project is the identification of new viral guanine-N7-methyltransferase inhibitors that are more potent than known BSSAs and can lead to the development of new antiviral therapeutics against SARS-CoV-2.
Nano-layers and nano-particles based on group IV elements Si, Ge and Sn
AUTOR: Marek Mihalkovič
Ultra-thin layers and small particles of Silicon or Germanium will – according to our preliminary study – take form of clathrate, which is stabilized by surface energy against the well know stable bulk form – diamond-type structure, and has considerably lower internal energy than another competitor – amorphous phase. Computational prediction of these new forms requires intensive atomistic simulations. Machine-learning force fields fitted to DFT energetics – if they can capture bulk and surface interactions simultaneously – will substantially lower the computational requirements. We also ask whether aluminia or zirconia substrates can reproduce free-standing slab results.|basic_html|Ultra-thin layers and small particles of Silicon or Germanium will – according to our preliminary study – take form of clathrate, which is stabilized by surface energy against the well know stable bulk form – diamond-type structure, and has considerably lower internal energy than another competitor – amorphous phase. Computational prediction of these new forms requires intensive atomistic simulations. Machine-learning force fields fitted to DFT energetics – if they can capture bulk and surface interactions simultaneously – will substantially lower the computational requirements. We also ask whether aluminia or zirconia substrates can reproduce free-standing slab results.
Exploring text generation LLMs with implications for disinformation tackling
AUTOR: Robert Moro
Machine text generation has significantly progressed in the past years thanks to the emergence of a new generation of large language models (LLMs). They had a profound impact on the area of NLP, while pushing forward several avenues of research and opening new challenges. At the same time, they opened new opportunities for their misuse, including generation of disinformation in multiple languages. Our objective is to research LLMs’ capabilities (incl. in-context learning and synthetic data augmentation) to tackle multilingual and multimodal disinformation. Namely, we will focus on the following tasks: 1) multilingual claim matching and claim detection, 2) machine generated text detection, and 3) textual/multimodal content credibility assessment. To perform the planned tasks, we will need to fine-tune pretrained classification models (e.g. XLM-RoBERTa, MDeBERTa), run inference on selected LLMs (in zero-shot or few-shot setting) covering various model sizes, architectures, and means of pre-training (e.g. LLaMA-2, Platypus, LLaMA-65B, Alpaca-LoRa-30B, etc.) and fine-tune some of the selected LLMs to make them work efficiently with custom knowledge bases and data. The planned duration of the project is 12 months and we request 50 000 CPU-core hours and 400 000 GPU-core hours. The expected results will improve our understanding of advantages and limitations of LLMs and provide several improved methods that can be used to tackle disinformation online.|basic_html|Machine text generation has significantly progressed in the past years thanks to the emergence of a new generation of large language models (LLMs). They had a profound impact on the area of NLP, while pushing forward several avenues of research and opening new challenges. At the same time, they opened new opportunities for their misuse, including generation of disinformation in multiple languages. Our objective is to research LLMs’ capabilities (incl. in-context learning and synthetic data augmentation) to tackle multilingual and multimodal disinformation. Namely, we will focus on the following tasks: 1) multilingual claim matching and claim detection, 2) machine generated text detection, and 3) textual/multimodal content credibility assessment. To perform the planned tasks, we will need to fine-tune pretrained classification models (e.g. XLM-RoBERTa, MDeBERTa), run inference on selected LLMs (in zero-shot or few-shot setting) covering various model sizes, architectures, and means of pre-training (e.g. LLaMA-2, Platypus, LLaMA-65B, Alpaca-LoRa-30B, etc.) and fine-tune some of the selected LLMs to make them work efficiently with custom knowledge bases and data. The planned duration of the project is 12 months and we request 50 000 CPU-core hours and 400 000 GPU-core hours. The expected results will improve our understanding of advantages and limitations of LLMs and provide several improved methods that can be used to tackle disinformation online.
Electronic structure computations for material and drug design
AUTOR: Lukáš Bučinský
The electronic structure of transition metal complexes and of their ligands will be explored. Materials related to hydrogen storage will be studied and druglikeness will be evaluated using molecular docking.
The work related to transition metal complexes will be based on DFT and CASSCF/NEVPT2 calculations. The electronic structure will be compared with experimentally derived charge density from single crystal X-ray diffraction measurements. Hydrogen storage capacity of graphene like materials, metal organic frameworks and transition metal complexes will be calculated on smaller sized clusters and under periodic boundary conditions. Molecular docking calculations will be utilized to evaluate a potential usage of the studied transition metal complexes in medicinal applications.|basic_html|The electronic structure of transition metal complexes and of their ligands will be explored. Materials related to hydrogen storage will be studied and druglikeness will be evaluated using molecular docking.
The work related to transition metal complexes will be based on DFT and CASSCF/NEVPT2 calculations. The electronic structure will be compared with experimentally derived charge density from single crystal X-ray diffraction measurements. Hydrogen storage capacity of graphene like materials, metal organic frameworks and transition metal complexes will be calculated on smaller sized clusters and under periodic boundary conditions. Molecular docking calculations will be utilized to evaluate a potential usage of the studied transition metal complexes in medicinal applications.
Computational modelling for photochemistry
AUTOR: Šimon Budzák
The rapidly growing field of photoswitching is based on photoisomerization reactions, which involve the conversion of a molecule from one isomer to another upon absorption of light. This phenomenon finds utility in selectively activating drugs within specific regions of the human body or encoding information in molecular form using light. To function effectively as a photoswitch, a molecule must exhibit a high molar absorption coefficient and large photoisomerization quantum yield (PQY), indicative of the transformation’s efficiency. Our project aims to better understand PQY within two distinct photoswitch families: hydrazones and dithienylethenes. To accomplish this, we will use both static and dynamic methods of computational chemistry. We will analyze the topology of the excited state surface and identify the position and effectiveness of photochemical funnels. In addition, we will use molecular dynamic simulations to track the development of the excited state over the time. To achieve the objectives significant computational resources are needed with the potential to advance the development of more efficient photoswitches.|basic_html|The rapidly growing field of photoswitching is based on photoisomerization reactions, which involve the conversion of a molecule from one isomer to another upon absorption of light. This phenomenon finds utility in selectively activating drugs within specific regions of the human body or encoding information in molecular form using light. To function effectively as a photoswitch, a molecule must exhibit a high molar absorption coefficient and large photoisomerization quantum yield (PQY), indicative of the transformation’s efficiency. Our project aims to better understand PQY within two distinct photoswitch families: hydrazones and dithienylethenes. To accomplish this, we will use both static and dynamic methods of computational chemistry. We will analyze the topology of the excited state surface and identify the position and effectiveness of photochemical funnels. In addition, we will use molecular dynamic simulations to track the development of the excited state over the time. To achieve the objectives significant computational resources are needed with the potential to advance the development of more efficient photoswitches.
Interactions of albumin with poly(ethylene oxide) coated graphene
AUTOR: Zuzana Benková
Prevention or control of protein adsorption onto surfaces have applications in diverse fields from medicine to industrial coatings. Despite the important role of surface coating by end-grafted polymer chains, the mechanism of protein repulsion remains elusive. There is no theoretical model that satisfactorily explains the ambiguities observed in the experiments. Bridging theoretical and experimental approaches, atomistic molecular dynamics simulations will be performed to scrutinize interactions of poly(ethylene oxide) (PEO) chains grafted onto graphene with human serum albumin (HSA), whose secondary structure is rich in α-helices. To investigate the role of water molecules in these interactions, the simulations in water and in vacuum will be carried out. A large range of grafting densities of PEO chains will be considered to cover different conformational regimes of the grafted polymer layers. Attention will also be focused on the effect of these conformations on the interactions with HSA and on the secondary structure of HAS. Specifically, the question arises whether the soft layer of grafted chains stabilizes or destabilizes the α-helices of HSA. The systems will be studied in two setups, one corresponding to biological conditions in living systems (setup I) and the other corresponding to the surface forces apparatus technique (setup II). Of interest is the comparison of whether the adsorption/repulsion of HSA in setup I coincides with the presence/absence of a negative minimum in the free energy profile of HSA in setup II. In other words, this comparison should reveal whether the two setups are generally consistent or whether there are conditions under which the two setups do not provide consistent conclusions. The propensity of the grafted PEO layers to repel or attract HSA will be compared for all simulated systems. |basic_html|Prevention or control of protein adsorption onto surfaces have applications in diverse fields from medicine to industrial coatings. Despite the important role of surface coating by end-grafted polymer chains, the mechanism of protein repulsion remains elusive. There is no theoretical model that satisfactorily explains the ambiguities observed in the experiments. Bridging theoretical and experimental approaches, atomistic molecular dynamics simulations will be performed to scrutinize interactions of poly(ethylene oxide) (PEO) chains grafted onto graphene with human serum albumin (HSA), whose secondary structure is rich in α-helices. To investigate the role of water molecules in these interactions, the simulations in water and in vacuum will be carried out. A large range of grafting densities of PEO chains will be considered to cover different conformational regimes of the grafted polymer layers. Attention will also be focused on the effect of these conformations on the interactions with HSA and on the secondary structure of HAS. Specifically, the question arises whether the soft layer of grafted chains stabilizes or destabilizes the α-helices of HSA. The systems will be studied in two setups, one corresponding to biological conditions in living systems (setup I) and the other corresponding to the surface forces apparatus technique (setup II). Of interest is the comparison of whether the adsorption/repulsion of HSA in setup I coincides with the presence/absence of a negative minimum in the free energy profile of HSA in setup II. In other words, this comparison should reveal whether the two setups are generally consistent or whether there are conditions under which the two setups do not provide consistent conclusions. The propensity of the grafted PEO layers to repel or attract HSA will be compared for all simulated systems.
Analysis of the structure of biomolecules using quantum chemical methods.
AUTOR: Michal Hricovíny
This project deals with theoretical analysis of biomolecular structure by applying quantum-chemical methods (QM) and methods of artificial intelligence (AI). We will focus on biologically important carbohydrates and their derivatives, as well as proteins and protein-carbohydrate complexes. The emphasis will be placed on the biologically important carbohydrate group – glycosaminoglycans, which are involved in a number of essential biological processes. Thus, the aim of the project is to reveal the details of the influence of structure on the interaction between carbohydrates and proteins, the influence of substituents, ions and solvents. Different quantum chemical methods will be tested in order to find the most optimal and efficient method that would give reliable results. |basic_html|This project deals with theoretical analysis of biomolecular structure by applying quantum-chemical methods (QM) and methods of artificial intelligence (AI). We will focus on biologically important carbohydrates and their derivatives, as well as proteins and protein-carbohydrate complexes. The emphasis will be placed on the biologically important carbohydrate group – glycosaminoglycans, which are involved in a number of essential biological processes. Thus, the aim of the project is to reveal the details of the influence of structure on the interaction between carbohydrates and proteins, the influence of substituents, ions and solvents. Different quantum chemical methods will be tested in order to find the most optimal and efficient method that would give reliable results.
Polymers with active chiral topology and machine learning frameworks.
AUTOR: Dušan Račko
Specific: The project seeks to explore a novel category of knotted and catenated polymers while tackling a novel problem that extends beyond current knowledge, namely, macromolecular chirality and chiral confinements. This will be achieved through the utilization of state-of-the-art molecular simulations, harnessing leading-edge high-performance computing techniques and software packages. Furthermore, the project seeks to attract and engage young researchers in pioneering research. We propose the following specific scientific objectives to serve as the foundation for the realization of these incentives:
Investigate Polymer Properties: Explore the physical and biophysical characteristics of polymer knots, with a focus on understanding how their topological states including chirality influence these properties.
Develop Novel Methods: Innovate techniques for the preparation and characterization of polymer knots and links with chirality of choice.
Simulate Nanotechnological Devices: Employ computational simulations to evaluate the efficacy of nanotechnological devices in both detecting and creating knots and determine chiral composition.
Analyse Chirality and Environment: Gain valuable insights into the intricate relationship between polymer knot topology, chirality, and chiral environments, particularly their significance in emerging domains like chiral nanotechnology and superstructures.
Explore Glass Transition Effects: Study the influence of alterations in topological states on the glass transition phenomenon in polymers, providing critical insights into influence of knottedness for future material applications of knotted polymers.
Examine Geometric Patterns: Investigate correlations between geometric patterns within polymer chains and knottedness, with emphasis on distinguishing correlations also in chiral environments.
Machine Learning for Topology: Harness machine learning frameworks to develop an experimentally accessible descriptor that can train neural networks to understand and identify polymer topologies and identify also chirality in complex environments.
Measurable: Progress towards these objectives will be quantified through the successful execution of specific tasks and the achievement of defined milestones, that involve training of early career researcher (ECR), publications in Open Access (OA) journals, contributions on scientific meetings and developing collaborations.
Assignable: The project team, is led by the experienced excellent researcher, who is also the Head of the Department of Molecular Simulations of Polymers (Department) at the Polymer institute of the SAS. The research team accounts for full employment of freshly defended PhD, Dr. Renáta Rusková, but the team can be operationally reinforced reaching for the experienced researchers from inside the Department.
Realistic: Given the long-term team’s expertise in using and performing computer molecular simulations beyond the state-of-the-art, access to high-performance computing resources, and skills with leading-edge software, the objectives are well within reach. However, they are ambitious enough to drive significant contributions to the emerging field of chiral nanotechnology.
Time-Related: The project’s objectives will be achieved within the defined project timeline, completion of partial tasks takes into account the duration of the period of this call, which is 1 year (that can be extended).|basic_html|Specific: The project seeks to explore a novel category of knotted and catenated polymers while tackling a novel problem that extends beyond current knowledge, namely, macromolecular chirality and chiral confinements. This will be achieved through the utilization of state-of-the-art molecular simulations, harnessing leading-edge high-performance computing techniques and software packages. Furthermore, the project seeks to attract and engage young researchers in pioneering research. We propose the following specific scientific objectives to serve as the foundation for the realization of these incentives:
Investigate Polymer Properties: Explore the physical and biophysical characteristics of polymer knots, with a focus on understanding how their topological states including chirality influence these properties.
Develop Novel Methods: Innovate techniques for the preparation and characterization of polymer knots and links with chirality of choice.
Simulate Nanotechnological Devices: Employ computational simulations to evaluate the efficacy of nanotechnological devices in both detecting and creating knots and determine chiral composition.
Analyse Chirality and Environment: Gain valuable insights into the intricate relationship between polymer knot topology, chirality, and chiral environments, particularly their significance in emerging domains like chiral nanotechnology and superstructures.
Explore Glass Transition Effects: Study the influence of alterations in topological states on the glass transition phenomenon in polymers, providing critical insights into influence of knottedness for future material applications of knotted polymers.
Examine Geometric Patterns: Investigate correlations between geometric patterns within polymer chains and knottedness, with emphasis on distinguishing correlations also in chiral environments.
Machine Learning for Topology: Harness machine learning frameworks to develop an experimentally accessible descriptor that can train neural networks to understand and identify polymer topologies and identify also chirality in complex environments.
Measurable: Progress towards these objectives will be quantified through the successful execution of specific tasks and the achievement of defined milestones, that involve training of early career researcher (ECR), publications in Open Access (OA) journals, contributions on scientific meetings and developing collaborations.
Assignable: The project team, is led by the experienced excellent researcher, who is also the Head of the Department of Molecular Simulations of Polymers (Department) at the Polymer institute of the SAS. The research team accounts for full employment of freshly defended PhD, Dr. Renáta Rusková, but the team can be operationally reinforced reaching for the experienced researchers from inside the Department.
Realistic: Given the long-term team’s expertise in using and performing computer molecular simulations beyond the state-of-the-art, access to high-performance computing resources, and skills with leading-edge software, the objectives are well within reach. However, they are ambitious enough to drive significant contributions to the emerging field of chiral nanotechnology.
Time-Related: The project’s objectives will be achieved within the defined project timeline, completion of partial tasks takes into account the duration of the period of this call, which is 1 year (that can be extended).
Modern challenges of molecular biology thorough the prism of residue network modelling
AUTOR: Vladimír Sládek
The primary goals of the project are to develop and integrate modern theoretical methods into current workflows established in molecular biology with particular focus on knowledge-based drug design.
Disruption of natural protein-protein signalling pathways is a cornerstone for the mechanisms of many diseases. Viral mode of action is often the mimicking of natural binding partners in protein complexes. Such is the case of the Human cytomegalovirus protein UL141, which blocks the binding of the TRAIL-R2 death receptor to the TRAIL protein preventing the infected cell to undergo cell death.
The understanding of the physical nature of the protein-protein binding interfaces is of tremendous importance when one desires to modulate the formation of the protein complex. Due to the highly complex nature of these interaction one cannot rely on heuristics. The realm of protein residue networks has established itself as a suitable mathematico-theoretical model to address this kind of problems. It allows to identify binding hotspots and hubs. An important development of these methods was the combination of network models and physical models which resulted in the emergence of pair interaction energy weighted protein residue networks (PIE-PRN), to which we contributed.
The PIE-PRN models can be based on residue pair interaction data obtained from specialised quantum-chemical calculations such as FMO. The computational resources requested in this project will be used mostly for this purpose.|basic_html|The primary goals of the project are to develop and integrate modern theoretical methods into current workflows established in molecular biology with particular focus on knowledge-based drug design.
Disruption of natural protein-protein signalling pathways is a cornerstone for the mechanisms of many diseases. Viral mode of action is often the mimicking of natural binding partners in protein complexes. Such is the case of the Human cytomegalovirus protein UL141, which blocks the binding of the TRAIL-R2 death receptor to the TRAIL protein preventing the infected cell to undergo cell death.
The understanding of the physical nature of the protein-protein binding interfaces is of tremendous importance when one desires to modulate the formation of the protein complex. Due to the highly complex nature of these interaction one cannot rely on heuristics. The realm of protein residue networks has established itself as a suitable mathematico-theoretical model to address this kind of problems. It allows to identify binding hotspots and hubs. An important development of these methods was the combination of network models and physical models which resulted in the emergence of pair interaction energy weighted protein residue networks (PIE-PRN), to which we contributed.
The PIE-PRN models can be based on residue pair interaction data obtained from specialised quantum-chemical calculations such as FMO. The computational resources requested in this project will be used mostly for this purpose.
Advancing Lung Ultrasound Analysis: Innovative Approaches for Diagnostics and Data Enhancement
AUTOR: Maroš Hliboký
This research conducted by a group of researchers focuses on the design and development of artificial neural methods for processing lung ultrasound images. The study comprises five phases, addressing classification, time-dependent feature extraction, and data augmentation. The research aims to improve patient diagnosis by creating models for detecting pathological findings and designing new approaches for extracting time-dependent features. All phases are in an advanced stage, facilitating model performance and processing efficiency comparison.|basic_html|This research conducted by a group of researchers focuses on the design and development of artificial neural methods for processing lung ultrasound images. The study comprises five phases, addressing classification, time-dependent feature extraction, and data augmentation. The research aims to improve patient diagnosis by creating models for detecting pathological findings and designing new approaches for extracting time-dependent features. All phases are in an advanced stage, facilitating model performance and processing efficiency comparison.