Fluoropolymers are employed in countless end-user applications across several industries. One such fluoropolymer is polytetrafluoroethylene. This research is concerned with studying and understanding the thermal behavior of polytetrafluoroethylene. Such understanding is critical to predict its behavior in diverse service environments as the polymer ages and for allowing bottom up design of improved polymers for specific applications. While a plethora of experiments have investigated the thermal properties of polytetrafluoroethylene, examining these properties using molecular dynamics simulations remains in its infancy. In particular, the current body of molecular dynamics research on polytetrafluoroethylene has primarily focused on studying polytetrafluoroethylene phases, its physical nature, and its helical conformational structure. The present study is the first molecular dynamics simulations research to study polytetrafluoroethylene behavior near the glassy transition temperature. Specifically, the current research utilizes molecular dynamics simulations to achieve the following objectives: (a) model and predict polytetrafluoroethylene glassy transition temperature at different molecular weights, (b) examine the impact of glassy transition temperature on the volume-temperature and thermal properties, (c) study the influence of molecular weight on polytetrafluoroethylene melt and glassy state, and (d) determine the governing forces at the molecular level that control polytetrafluoroethylene glassy transition temperature. Achieving the aforementioned objectives requires performing four major tasks. Motivated by the scarcity of polytetrafluoroethylene force fields research, the first task aims to generate and test polytetrafluoroethylene force fields. The parameters were produced based on the Optimized Potentials for Liquid Simulations All Atom model. The intramolecular parameters were generated using the automated frequency matching method while the torsional terms were fitted using the nonlinear least squares algorithm. The intermolecular partial atomic charges were obtained using Northwest Computational Chemistry software and fitted using the restrained electrostatic potential at (MP2/6-31G*) level of theory. The final set of parameter was tested by calculating polytetrafluoroethylene density using molecular dynamics simulations. The second task involves building polytetrafluoroethylene amorphous structure using molecular dynamics at periodic boundary conditions for polytetrafluoroethylene cell at different molecular weights. We use the amorphous structure in the molecular dynamics simulations in consistence with research evidence which reveals that polymer properties such as the specific volume will differ as the polymer passes the glassy transition when it is in the amorphous phase structure whereas no variation occurs when the polymer passes the glassy transition while it is in the crystalline structure. The third task includes testing polytetrafluoroethylene melt phase properties: density, specific heat, boiling point, and enthalpy of vaporization. In the fourth and final task, we performed molecular dynamics simulations using NAnoscale Molecular Dynamics program. This task involves the polymer relaxation process to predict polytetrafluoroethylene mechanical behavior around the glassy transition temperature. Properties that are affected by this transition such as density, heat capacity, volumetric thermal expansion, the specific volume, and the bulk modulus were examined and the simulated results were in good agreement with experimental findings.

The Thesis deals with MD simulations of solutions of chiral solutes in chiral solvents. These solutions consist of 2,2,2-trifluoro-1-phenylethanol, 1-phenylethanol and 1-phenyl- ethanamine.The differences in NMR properties between different combnations of solvent and solute absolute configuration were modeled. Indeed, differences in radial distribution functions and conformer abundances of solute calculated by the WHAM method were found. These results correlated with experimental differences in NMR shifts. Additionally, a method of cluster preselection was developed. It significantly decreased the amount of clusters needed for computations of NMR shieldings and hence the computer time. Keywords: chirality, molecular dynamic, nuclear magnetic resonance

Title: Ion-Protein Interactions Author: Mgr. et Mgr. Jan Heyda Department: Physical and Macromoleculer Chemistry Advisor: Prof. Pavel Jungwirth, DSc., IOCB AS CR, v.v.i. Advisor's e-mail address: pavel.jungwirth@uochb.cas.cz Abstract: Conventional molecular dynamics simulations in combination with ad- vanced methods of analyses were used to improve the understanding of the interac- tion between ions and proteins in salt solutions. Thus systems of diverse complexity and size were investigated, starting with simple (and molecular) salt solutions with small fragments that mimic the various functional groups of amino acids such as N-methylacetamide representing the peptide bond or alkylated ammonium cations. Continuing with individual positively charged amino acids (arginine, histidine, ly- sine) a strong binding interaction with small fluoride anion that is significantly weak- ened for larger halides (Cl− , Br− , I− ) was described. This observation was extended by detecting the strong sensitivity of fluoride to charge distribution on ammonium, lysine side chain, and the N-terminal of glycine while sensitivity of iodide was found to be low. Later it was shown that the attractive side chain-side chain interactions are significant for short positively charged peptide fragments in polyarginine and dihistidine, while...

The number of three-dimensional structures of potential protein targets available in several platforms such as the Protein Data Bank is subjected to a constant increase over the last decades. This observation should be an additional motivation to use structure-based methodologies in drug discovery. In the recent years, different success stories of Structure Based Drug Design approach have been reported. However, it has also been shown that a lack of druggability is one of the major causes of failure in the development of a new compound. The concept of druggability can be used to describe proteins with the capability to bind drug-like compounds. A general consensus suggests that around 10% of the human genome codes for molecular targets that can be considered as druggable. Over the years, the protein druggability was studied with a particular interest to capture structural descriptors in order to develop computational methodologies for druggability assessment. Different computational methods have been published to detect and evaluate potential binding sites at protein surfaces. The majority of methods currently available are designed to assess druggability of a static structure. However it is well known that sometimes a few local rearrangements around the binding site can profoundly influence the affinity of a small molecule to its target. The use of techniques such as molecular dynamics (MD) or Metadynamics could be an interesting way to simulate those variations. The goal of this thesis was to design a new computational approach, called JEDI, for druggability assessment using a combination of empirical descriptors that can be collected ‘on-the-fly’ during MD simulations. JEDI is a grid-based approach able to perform the druggability assessment of a binding site in only a few seconds making it one of the fastest methodologies in the field. Agreement between computed and experimental druggability estimates is comparable to literature alternatives. In addition, the estimator is less sensitive than existing methodologies to small structural rearrangements and gives consistent druggability predictions for similar structures of the same protein. Since the JEDI function is continuous and differentiable, the druggability potential can be used as collective variable to rapidly detect cryptic druggable binding sites in proteins with a variety of MD free energy methods.

Methods to advance the understanding of water and other aqueous systems are devel- oped. This work falls into three areas: The creation of better interaction potentials for water, improved methods for sampling configurational space, and the applications of these methods to understand systems of interest. Charge transfer has been shown by ab initio methods to be important in the water–water and water–ion interactions. A model for treating charge transfer in liquid water and aqueous systems is presented in this manuscript. The model is called Discrete Charge Transfer (DCT) and is based on the commonly-used TIP4P/2005 model, which represents the charge distribution of water molecules with three charge sites. Such models have been very successful in reproducing many of the physical properties of water. Charge transfer is introduced by transferring a small amount of charge, -0.02e, from the hydrogen bond acceptor to the hydrogen bond donor, as has been indicated by electronic structure calculations. We have parameterized both polarizable and non-polarizable potentials, optimized to include charge transfer. Methods to surmount the obstacles incurred by the introduction of charge transfer, which involve the amount of charge transfer at large distances and implementation into Molecular Dynamics simulation, is presented, along with our results assessing the importance of charge transfer in liquid water and aqueous systems. Also presented is a method for improving eciency of a sampling technique, Replica Exchange, by reducing the number of replicas. The improved method is called Replica Exchange with Driven Scaling (REDS2).

The presence of charge transfer (CT) interactions is clear in a variety of systems. In CT, some electron density is shifted from one molecule to another (non-bonded) molecule. The importance of this CT interaction is unclear. Previous attempts to look at the conse- quences of CT required the use of ab initio molecular dynamics (AIMD), a computationally intensive method. Herein, a method for including CT in force field (FF) simulations is described. It is efficient, produces charges in agreement with AIMD, and prevents long- ranged CT. This CT MD method has been applied to monatomic ions in water. When solvated, ions do not have an integer charge. Anions give up some electron density to their ligands, and cations receive some electron density from their ligands. In bulk, the first solvation shell does not compensate for all CT, i.e. the charge is not smeared out over the first solvation shell. Rather, some charge is also found in the second solvation shell and further into the bulk. The charge of the first solvation shell depends on the balance between ion-water and water-water CT. When an interface is present, the charge outside of the second solvation shell will reside at the interface. This occurs even when the ion is over 15 Å away from the surface. The effect of long-ranged CT is mediated by changes in the hydrogen bonding patterns in water induced by the ions (not direct CT from the ions to distant waters). The model has also been applied to water’s ‘‘self-ions’’ hydronium and hydroxide. Trajectories from the multi-state empirical valence bond model (MS-EVB3) are analyzed. The differences between monatomic and molecular ions are explored. The direction of CT and the effect of hydrogen bonding with the ion are considered. The damping of CT as ligands are added is discussed and a method to improve the MD model, in order to account for damping, is proposed.

Philosophiae Doctor - PhD / The HIV-1 envelope (Env) glycoprotein is the primary target of the humoral immune response and a critical vaccine candidate. However, Env is densely glycosylated and thereby substantially protected from neutralisation. Despite the importance of the HIV- 1 Env glycans, limited computational analyses have been employed to analyse these glycans. Here, the Env glycans of two HIV-1 wild-type subtype C isolates are examined, in detail, using computational approaches. These particular strains were used since in vitro data showed that the removal of a single glycan had a substantially different impact on the neutralisation sensitivity of the two strains. Molecular dynamics simulations, and the subsequent analyses, were carried out on the computationally determined, fully glycosylated, Env structures of these two wild-type strains and their N301A mutant counterparts. Detailed comparison of the molecular dynamics simulations demonstrated that unique glycan dynamics and conformations emerged and that, despite shared HXB2 reference sequence positions, the glycans adopted distinct conformations specific to each wild-type model. Furthermore, different changes in conformations were observed for each wild-type model compared to its N301A mutant counterpart and, interestingly, these N301A mutant model-specific glycan conformations were directly associated with the protein residues ultimately found to be exposed, which may explain the varied resistance to neutralising antibodies observed, in vitro, for the two N301A mutant strains.

Dengue é uma doença tropical negligenciada que atualmente ameaça mais da metade da população mundial e representa custo anual de bilhões de dólares para as áreas afetadas. Avanços recentes na elucidação estrutural de proteínas humanas e vitais contribuíram para o entendimento do ciclo de replicação vital e das interações vírus-hospedeiro a nível molecular. Em particular, a camada mais externa do vírus é composta por 180 monômeros da glicoproteína do envelope. Cada um desses monômeros apresenta duas glicosilações que são reconhecidas por lectinas cuja atividade depende de Ca2+. A interação entre o vírus e essas lectinas favorece a infecção ou o aparecimento de sintomas mais severos. Neste trabalho aplicamos simulações de dinâmica molecular e métodos de estimativa de afinidade para avançar nosso conhecimento sobre os mecanismos moleculares do reconhecimento de glicosilações high-mannose pelas lectinas DC-SIGN e MR, ambas já experimentalmente validadas como alvos biológicos para desenvolvimento de novos antivirais. Adicionalmente, através de virtual screening usando um conjunto de programas de implementação própria e uma biblioteca de moléculas já aprovadas para uso como fármacos, encontramos uma molécula (lohexol) que apresenta alto potencial de interação com ambos os receptores. Embora testes experimentais ainda sejam necessários para validar esse achado, nossos resultados sugerem que essa molécula, e eventualmente moléculas similares, podem agir como inibidor da infecção do vírus da Dengue por um mecanismo duplo. / Dengue is a tropical neglected disease that currently threatens more than half the world\'s population and represents a yearly cost of billions of dollars to the affected areas. Recent advances in the elucidation of 3D structures of human and viral proteins has contributed to the understanding of the viral replication cycle and virus-host interactions at the molecular level. In particular, the outermost layer of the virus is composed by 180 monomers of the envelope glycoprotein. Each one of these monomers displays two glycosylations that are recognized by lectins which have Ca2+-dependent activity. The interaction between the virus and these lectins favors infection or severe disease onset. In this work we apply molecular dynamics simulations and affinity estimation methods to advance our knowledge about the molecular mechanisms underlying the recognition of the high-mannose glycosylation by the DC-SIGN and MR lectins, both of which are already validated as targets for the development of new antivirals against Dengue. Additionally, by running virtual screening assays using a set of programs implemented by us and a library containing molecules which are already approved to be used as drugs, we found one molecule (lohexol) which presents high potential of interaction with both receptors. Although experimental testing is still necessary to validate this finding, our results suggest that this molecule, and eventually similar ones, can act as an inhibitor of the Dengue virus infection by a dual mechanism.

<div>Atomistic simulations are used in this project to study the deformation mechanism of polycrystalline and bicrystal of pure Al and Al-Mg alloys. Voronoi Tessellation was used to create three-dimensional polycrystalline models. Monte Carlo and Molecular</div><div>Dynamics simulations were used to achieve both mechanical and chemical equilibrium in all models. The first part of the results showed improved strength, which is included the yield strength and ultimate strength in the applied tensile loading through the addition of 5 at% Mg to nanocrystalline aluminum. By viewing atomic structures, it clearly shows the multiple strengthening mechanisms related to doping in Al-Mg alloys. The strength mechanism of dopants exhibits as dopant pinning grain boundary (GB) migration at the early deformation stage. At the late stage where it is close to the failure of nanocrystalline materials, Mg dopants can stop the initiation of intergranular cracks and also do not let propagation of existing cracks along the GBs. Therefore, the flow stress will improve in Al-Mg alloy compared to pure Al. In the second part of our results, in different bicrystal Al model, Σ 3 model has higher strength than other models. This result indicates that GB structure can affect the strength of the material. When the Mg dopants were added to the Al material, the strength of sigma 5 bicrystal models was improved in the applied shear loading. </div><div><br></div><div>However, it did not happen for Σ 3 model, which shows Mg dopants cannot affect the behavior of this GB significantly. Analysis of GB movements shows that Mg dopants stopped GBs from moving in the Σ 5 models. However, in sigma 3 GBs, displacement of grain boundary planes was not affected by Mg dopants. Therefore, the strength and flow stress are improved by Mg dopants in Σ 5 Al GBs, not in the Σ 3 GB.</div>

Neste trabalho, a relação entre a adesão e o grau de desordem de superfícies em contato é estudada por meio de simulações por dinâmica molecular. O sistema em estudo é composto por um indentador cilíndrico rígido e um bloco deformável. Um reservatório térmico é colocado logo abaixo do bloco, de forma a manter o sistema a temperatura ambiente. Os sólidos são feitos do mesmo material e são modelados por intermédio do potencial genérico de Lennard-Jones. A adesão entre as superfícies é variada gradualmente por meio da variação do raio de corte do potencial que descreve a interação entre as superfícies indentador-bloco. Cada simulação se inicia com um recozimento, após o qual são realizadas as simulações de contato propriamente ditas, até que ocorra penetração de 1,7 raios atômicos. A força normal, a energia potencial, a temperatura e a energia cinética são acompanhadas ao longo do processo. Os resultados reforçam a importância da adesão no jump-to-contact, estando este fenômeno relacionado à geração de defeitos cristalinos. Há indícios de que a distribuição da carga, bem como a dissipação de energia, seja influenciada pela adesão na interface. Mostra-se que a desordem apresentada pelos átomos do bloco é proporcional à contribuição da adesão. / In this work, the relationship between adhesion and the degree of disorder of surfaces in contact is studied by means of molecular dynamics simulations. The system consists of a rigid cylindrical indenter and a deformable block. A heat reservoir is placed just below the block, in order to maintain the system at room temperature. Both solids are made of the same material and are modeled through the generic Lennard-Jones potential. Adhesion between the surfaces is gradually varied by varying the cut-off radius of the potential describing the interaction between the indenter surface and the block. Each contact simulation is preceeded by an annealing step, and is conducted until the penetration reaches 1.7 atomic radii. The normal force, the potential energy, kinetic energy and temperature are monitored throughout the process. The results reinforce the importance of adhesion in the phenomenon of jump-to-contact, which is also related to the generation of crystalline defects. The results indicate that the load distribution and power dissipation is influenced by the adhesion at the interface. It is shown that the disorder presented by the atoms of the block is proportional to the contribution of adhesion.