Projects

Research area of Prof. P. Hobza

Non-covalent Interactions

Theoretical studies on stabilization energy, structure, geometry, properties and nature of stabilization on different types of noncovalent interactions: H-bonding; improper - blue shifting - H-bonding; dihydrogen bonding; halogen bonding; stacking; dispersion bonding.

Development of nonempirical ab initio quantum chemical methods for calculation of noncovalent interactions

With ever growing amount of information about the accuracy and applicability of various (low-order scaling) methods for calculation of noncovalent complexes, we can focus on the weak points of these methods and eventually propose a ways to increase their accuracy, applicability and reliability. Scaled perturbation theories, spin-component scaled methods and hybrids of density functional and wave function theories are currently investigated.

Development of semiempirical QM methods describing noncovalent interactions

Semiempirical QM methods are potentially very useful tool for study of large biomolecules, as their ability to describe quantum mechanical effects makes them superior to widely used molecular mechanics. However, these methods fail to describe properly noncovalent interactions - hydrogen bonds and dispersion, and these interactions often determine the structure and function of biomolecules. We have developed corrections to several semiempirical methods that allow to achieve chemical accuracy in description of these interactions.

Accurate calculation of noncovalent interaction energies and geometries

Highly accurate data on noncovalent complex properties, like the binding energy or its geometry present extremely valuable information not only for the proper understanding of the nature of the particular interaction, but also for development and testing of new, computationally more efficient approximate methods. In this respect, demanding coupled-cluster or other highly-correlated calculations have to be carried out. This is, considering the size of the biologically relevant species, a challenge for both the computer hardware and the software. Highly parallelized algorithms mainly in coupled-clusters and perturbation theories are being developed or optimized, utilization of the graphical processing units is also of a great interest for the future.

Many-body effects in noncovalent interactions

Proper description of many-body effects (also called "cooperativity" or "nonadditivity") in noncovalent complexes with a significant contribution form the dispersion energy is an extremely difficult task for computational chemistry, since the most sophisticated (thus the most demanding) computational methods have to be applied. A lot of the phenomena caused the many-body effects in, for instance, biology is still not revealed, mainly due to the oversimplification of models used in the past. Possible consequence of many-body effects on the geometry of the DNA or peptides are currently being studied.

Thermodynamic characteristics

The biologically relevant systems need to be described not only by means of the interaction energy but also by means of the free energy and entropy, respectively. These thermodynamic quantities are conveniently accessible through molecular dynamics (MD) simulations and advanced MD techniques. Particular attention is paid to an investigation of the DNA...ligand and protein...ligand complexes, to the role of solvent in the binding processes and solute conformational changes, all forward to the potential medicinal utilisation.

Systematic exploration of complex potential energy surfaces

Recent development in reaction path search algorithms allowed us to characterize reaction paths and transition states in systems of remarkable complexity. Systematic, combinatorial approach is used to determine all minimum energy paths connecting thermodynamically accessible minima. Currently, we are applying this methodology to conformation changes in biomolecular clusters and model peptides.

In silico drug design

Semiempirical PM6-D3H4X method, which was recently developed in our laboratory and which accurately describes H-bonding as well as dispersion energy, is used as a scoring function in virtual screening. Protein…ligand binding free energy is constructed as a sum of protein…ligand gas-phase binding free energy, change of hydration free energies, deformation energy and entropic term. When the crystal structure of a protein…ligand complex is not available some of docking algorithms are applied.

 


 

 

Research area of F. Lankaš

The diverse biological functions of RNA and DNA molecules cannot be fully understood from their chemical features. Rather, structure and mechanical properties of nucleic acids play a decisive role. For instance, the formation of nucleosomes, huge protein-DNA complexes involved in packaging eukaryotic genomes, depends on the sequence-specific DNA ability to deform. The ribosome, a large biomolecular machine synthesising proteins, contains functionally important stiff and flexible RNA elements. Thus, mechanical properties of RNA and DNA are important for their biological functioning. They also play a key role in artificial nucleic acids structures for nanoscience applications.

Our research is focused on multiscale modelling of nucleic acids structure and mechanical properties. In our approach, RNA and DNA molecules are modelled as ensembles of effective, interacting rigid bodies. The bodies may represent individual bases, base pairs, or larger groups of bases. The carefully formulated models contain unknown parameters whose values we infer from advanced large-scale, atomic-resolution molecular dynamics simulations, as well as from published structural data. This yields a mechanical description of RNA and DNA building blocks (motifs) from which properties of larger structures can be inferred. The results may help to understand the biological functioning of nucleic acids and may indicate ways for rational design of RNA and DNA nanostructures.

 


 

 

Research area of D. Nachtigallová

The behavior of nucleic acid bases in their excited states has been subject of several experimental and theoretical studies. The aim of these studies is to explain their photochemical behavior in order to prevent the nucleic acid bases against UV damage. In our group we are interested in the calculations of electronically excited states of nucleic acid bases in the stacked conformation to investigate the excited state energy transfer between DNA bases. The aim of these studies is to evaluate the interaction of DNA bases in their excited states depending on the sequence of bases and their mutual orientation. The calculations of the relevant small models are also performed to understand these phenomena based on the results obtained with precise methods which are not always possible in the calculations of DNA bases.

 

 


 

 

Research area of V. Špirko

Quantum-mechanical studying of large molecular rearangements opposed by nonharmonic (multiple minima) potential energy functions (conformational dynamics, isomerizations, proton transfers, dynamical corrections for molecular properties). Theoretical studying of highly excited and continuum ro-vibrational states of small molecular systems (density and statistical properties of molecular states, energy clustering and "hidden" symmetries, elementary chemical reactions). Approximate methods for quantum-mechanical calculations (adiabatic separation approaches, numerical integration of coupled Schrödinger equations).