UMD Theses and Dissertations

Permanent URI for this collectionhttp://hdl.handle.net/1903/3

New submissions to the thesis/dissertation collections are added automatically as they are received from the Graduate School. Currently, the Graduate School deposits all theses and dissertations from a given semester after the official graduation date. This means that there may be up to a 4 month delay in the appearance of a given thesis/dissertation in DRUM.

More information is available at Theses and Dissertations at University of Maryland Libraries.

Browse

Subject: search.filters.subject.Computational chemistry

Search Results

Now showing 1 - 10 of 10
  • Thumbnail Image
    Item
    Optimized simulations of fermionic systems on a quantum computer
    (2024) Wang, Qingfeng; Monroe, Christopher; Chemical Physics; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
    Quantum computing holds promise for simulating microscopic phenomena, offering profound implications across disciplines such as chemistry, condensed matter physics, and high-energy physics, particularly in the accurate simulation of fermions. However, practical implementation requires the optimization of quantum programs to mitigate quantum noise and decoherence effects. Given the constraints of near-term quantum computers, the Variational Quantum Eigensolver (VQE) emerges as a key approach for estimating molecular ground state energies, crucial for determining chemical properties. This work aims to present advancements in optimizing VQE simulations to minimize quantum computational resources. Specifically, this work explores various optimization strategies, including the utilization of second-order perturbation correction to recover additional energy beyond VQE estimates and select critical ansatz terms. Additionally, circuit optimization techniques are investigated, focusing on achieving shorter equivalent ansatz circuits, particularly for physically-inspired VQE ansatz, through methods such as generalized fermion-to-qubit transformations and Pauli string orderings. Furthermore, this work demonstrates the advantage of a better initial state on a trapped-ion quantum computer.
  • Thumbnail Image
    Item
    TOWARD ENSEMBLE-BASED DRUG DISCOVERY THOUGH ENHANCED SAMPLING
    (2023) Smith, Zachary; Tiwary, Pratyush; Biophysics (BIPH); Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
    Quantitatively assessing protein conformational dynamics and ligand dissociation are two problems of critical importance for computer-aided drug discovery. Both of these problems involve larger shifts in the protein conformation than are ordinarily considered in drug discovery efforts. Even though it is well known that proteins are best described as a dynamic ensemble of states, actually acquiring a representative ensemble, especially one with probabilities attached to states, has remained an elusive problem. Molecular dynamics can in theory capture the full ensemble with a long enough simulation but it would take millions of years to simulate the timescale needed to study drug binding or unbinding. Given this timescale problem, it is necessary to develop software solutions to accelerate the sampling of these important rare events. A number of enhanced sampling methods such as metadynamics have arisen to deal with this problem but the methods that are able to attain the fastest speedup also require a low-dimensional description of the system's dynamics. In this thesis, I will develop methods to describe protein dynamics with low-dimensional functions that can be used with enhanced sampling and apply these methods in an enhanced sampling pipeline. The methods developed will both perform variable selection finding a small set of descriptors for the protein dynamics and perform manifold learning to find a low-dimensional representation of the dynamics using this set of descriptions. This pipeline will be used to tackle both problems of conformational dynamics and ligand dissociation in a relatively automated manner. I will then describe how solving these problems in a high throughput manner could impact structure-based drug design efforts, and the work remaining to attain that goal.
  • Thumbnail Image
    Item
    Exploring Mechanisms and Predicting Reactivity of Transition Metal-Catalyzed and Photocatalyzed Radical and Polar Organic Transformations
    (2023) Martin, Robert Thompson; Davis, Jeffery; Chemistry; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
    The creation of protocols to form novel C-C and C-heteroatom bonds has been the primary goal of organic synthesis since its inception. Chemists have long harnessed both radical and polar reactivities, often as complementary paths to construct these bonds to yield more complex molecular architectures. However, compared to the development of synthetic protocols, development of mechanistic models and enriching of mechanistic understanding of many organic reactions has been limited. Computational studies into the mechanisms of organic transformations provide an avenue by which mechanisms of reactions can be better understood and new patterns of reactivity can be predicted. Herein, quantum-mechanical computational methods e.g., density functional theory (DFT) have been employed in the pursuit of understanding the mechanisms of a series of radical and polar reaction schemes. Specifically, DFT calculations were employed to understand the mechanism and origins of selectivity of two nickel(I)-catalyzed olefin functionalizations. These studies demonstrate a catalyst-control scheme by which selectivity can be induced by the steric properties of the catalyst (Chapter 1). Following this work, two photocatalytic transformations which yield difluorinated products were studied thoroughly with computations. First, a synthesis of difluorinated lactone derivatives revealed a long-lived radical intermediate and motivated mechanistic experiments to isolate this radical. Next, a synthesis of difluorinated oxindole derivatives demonstrated the ability of arenethiols to act as photocatalysts (Chapter 2). Then, computations were used to rigorously explore a copper-catalyzed reductive cross-coupling of imine and allenamides. Specifically, computations were employed to explore the mechanism of the transformation and the origins of stereoselectivity and the divergent formation of urea and diamine products (Chapter 3). Finally, two computational investigations into the mechanisms of transformations catalyzed by first-row transition metals are detailed. In particular, a nickel-catalyzed hydroarylation of gem-difluoroalkenes is explored computationally to determine the order of steps in the reaction. In addition, the mechanism of a cobalt(I)-catalyzed allylic substitution is considered to ascertain the nature of the transformation as either radical or polar (Chapter 4). Given the complexity of the mechanisms of these transformations, computational studies provide an alternative route to acquire useful mechanistic understanding that can support or explain observed experiments and suggest further mechanistic experiments that could provide stronger evidence for a given mechanistic proposal.
  • Thumbnail Image
    Item
    Simulating membrane-bound cytoskeletal dynamics
    (2023) Ni, Haoran; Papoian, Garegin A.; Biophysics (BIPH); Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
    The cell membrane defines the shape of the cell and plays an indispensable role in bridging intra- and extra-cellular environments. The membrane, consisting of a lipid bilayer and various attaching proteins, mechanochemically interacts with the active cytoskeletal network that dynamically self-organizes, playing a vital role in cellular biomechanics and mechanosensing. Comprehensive simulations of membrane-cytoskeleton dynamics can bring insight in understanding how the cell mechanochemically responds to external signals, but a computational model that captures the complex cytoskeleton-membrane with both refined details and computational efficiency is lacking. To address this, we introduce in this thesis a triangulated membrane model and incorporate it with the active biological matter simulation platform MEDYAN ("Mechanochemical Dynamics of Active Networks"). This model accurately captures the membrane physical properties, showing how the membrane rigidity, the structure of actin networks and local chemical environments regulate the membrane deformations. Then, we present a new method for simulating membrane proteins, using stochastic reaction-diffusion sampling on unstructured membrane meshes. By incorporating a surface potential energy field into the reaction-diffusion sampling, we demonstrate rich membrane protein collective behaviors such as confined diffusion, liquid-liquid phase separation and membrane curvature sensing. Finally, in order to capture stretching, bending, shearing and twisting of actin filaments which are not all available with traditional actomyosin simulations, we introduce new finite-radius filament models based off Cosserat theory of elastic rods, with efficient implementation using finite-dimensional configurational spaces. Using the new filament models, we show that the filaments' torsional compliance can induce chiral symmetry breaking in a cross-linked actin bundle. All the new models are implemented in the MEDYAN platform, shedding light on whole cell simulations, paving way for a better understanding of the membrane-cytoskeleton system and its role in cellular dynamics.
  • Thumbnail Image
    Item
    EXPLORING THE EFFECTS OF PHYSIOLOGICAL ENVIRONMENT ON AMYLOID AGGREGATION
    (2022) Sahoo, Abhilash; Matysiak, Silvina; Biophysics (BIPH); Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
    Molecular level self-assembly/aggregation processes are common in biomolecular systems. Specifically, aggregation of protein molecules results in formation of amyloid deposits, that has been associated with neuronal dysfunction leading up to neurodegeneration. The protein aggregation is often influenced by several external physiological features, which can modulate this pathological process in a specific or non-specific manner. This thesis aims to elucidate the role of such factors in amyloid aggregation in the context of neurodegeneration. As test cases, we have focused on different fragments of Amyloid-beta peptide and Huntingtin protein and explored common interaction schemes in the presence of phospholipid membranes, solvated glucose molecules and added trailing sequences. Phospholipid membranes, composed of a heterogeneous distribution of lipid molecules, serve as packaging envelopes in cellular systems. But several studies have suggested a role of cellular membranes in abetting protein aggregation in neurodegenerative diseases. The first section of this thesis explores Amyloid-beta 16-22 aggregation in the presence of membranes. Lipid membranes have been shown to modulate peptide aggregation in a charge dependent manner with anionic membranes promoting faster peptide aggregation into ordered fibrillar structures compared to zwitterionic membranes. In this work, we evaluate the role of this electrostatic membrane headgroup charge on Amyloid-beta 16-22 peptide aggregation with model lipid membranes composed of POPC (1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine) and POPS (1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoserine) lipids. Beyond, membrane charge, membrane's physical organization can also affect peptide-peptide and peptide-membrane interactions. Here, we have curated the effects of applied surface-tension, as a proxy for membrane curvature, on peptide fibrillation propensities. Apart from ordered structures such as membranes, solvated small molecules are a large class of molecules that can affect aggregation patterning by affecting peptides through both specific and non-specific interactions. The second section of this thesis explores Amyloid-beta 16-22 aggregation in varying hyperglycemic conditions, to draw correlations between Alzheimer's disease and type 2 diabetes. Here, we discovered that the glucose prefers partitioning onto the aggregate-water interface in a specific manner, leading to a loss in rotational entropy that propels peptide aggregation. In the final section, we discuss the case of pathological peptide aggregation in the case of Huntington's disease. Broadly, Huntinting protein's N-terminal region which consists of 17-residue N-terminal domain (N17) and the following Glutamine repeat tract (Poly-Q) are our objects of interest and associated with pathological aggregation. The aggregation landscape of N17 is analyzed in presence of added different lengths of trailing Poly-Q tract and the presence of curved membranes. We have approached our research through a computational lens using molecular dynamics simulations. To address the relevant concerns of large spatio-temporal scales necessary to study peptide aggregation systems with molecular simulations, we have developed a coarse-grained forcefield (ProMPT: Protein Model with Polarizability and Transferability) that uses reduced spatial resolution to accelerate phase-space exploration. The forcefield can capture secondary and tertiary folding of protein structures with minimal constraints, and is transferable across biomolecular systems without a need for re-parametrization. My dissertation presents a holistic picture of peptide aggregation and various physiological factors that affect it, with biomolecular simulation across multiple scales.
  • Thumbnail Image
    Item
    EPIGENETICS TUNE CHROMATIN MECHANICS, A COMPUTATIONAL APPROACH
    (2021) Pitman, Mary; Papoian, Garegin A; Chemistry; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
    The base unit of DNA packaging in eukaryotes, the nucleosome, is adaptively modified for epigenetic control. Given the vast chemical space of chromatin and complexity of signaling and expression, much of our knowledge about genetic regulation comes from a biochemical or structural perspective. However, the architecture and function of chromatin also mechanically responds to non-equilibrium forces. Mechanical and biochemical properties are not independent of one another and the interplay of both of these material properties is an area of chromatin physics with many remaining questions. Therefore, I set out to determine how the material properties of chromatin are altered by biochemical variations of nucleosomes. All-atom molecular dynamics is employed coupled with new computational and theoretical tools. My findings and predictions were collaboratively validated and biologically contextualized through multiscale experimental methods. First, I computationally discover that epigenetic switches buried within the nucleosome core alter DNA accessibility and the recruitment of essential proteins for mitosis. Next, using new computational tools, I report that centromeric nucleosomes are more elastic than their canonical counterparts and that centromeric nucleosomes rigidify when seeded for kinetochore formation. We conclude that the material properties of variants and binding events correlate with modified loading of transcriptional machinery. Further, I present my theoretical approach called Minimal Cylinder Analysis (MCA) that uses strain fluctuations to determine the Young's modulus of nucleosomes from all-atom molecular dynamics simulations. I show and explain why MCA achieves quantitative agreement with experimental measurements. Finally, the elasticity of hybrid nucleosomes in cancer is measured from simulation, and I implicate this oncogenic variant in potential neocentromere formation. Together, these data link the physics of nucleosome variations to chromatin states' plasticity and biological ramifications.
  • Thumbnail Image
    Item
    EMERGENT NETWORK ORGANIZATION IN LINEAR AND DENDRITIC ACTIN NETWORKS REVEALED BY MECHANOCHEMICAL SIMULATIONS
    (2021) Chandrasekaran, Aravind; Papoian, Garegin A; Chemistry; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
    Cells employ networks of filamentous biopolymers to achieve shape changes and exert migratory forces. As the networks offer structural integrity to a cell, they are referred to as the cytoskeleton. Actin is an essential component of the cellular cytoskeleton. The organization of the actin cytoskeleton is through a combination of linear and branched filaments. Despite the knowledge of various actin-binding proteins and their interactions with individual actin filaments, the network level organization that emerges from filament level dynamics is not well understood. In this thesis, we address this issue by using advanced computer simulations that account for the complex mechanochemical dynamics of the actin networks. We begin by investigating the conditions that stabilize three critical bundle morphologies formed of linear actin filaments in the absence of external forces. We find that unipolar bundles are more stable than apolar bundles. We provide a novel mechanism for the sarcomere-like organization of bundles that have not been reported before. Then, we investigate the effect of branching nucleators, Arp2/3, on the hierarchical organization of actin in a network.By analyzing actin density fields, we find that Arp2/3 works antagonistic to myosin contractility, and excess Arp2/3 leads to spatial fragmentation of high-density actin domains. We also highlight the roles of myosin and Arp2/3 in causing the fragmentation. Finally, we understand the cooperation between the linear and dendritic filament organization strategies in the context of the growth cone. We simulate networks at various concentrations of branching molecule Arp2/3 and processive polymerase, Enabled to mimic the effect of a key axonal signaling protein, Abelson receptor non-tyrosine kinase (Abl). We find that Arp2/3 has a more substantial role in altering filament lengths and spatial actin distribution. By looking at conditions that mimic Abl signaling, we find that overexpression mimics are characterized by network fragmentation. We explore the consequence of such a fragmentation with perturbative simulations and determine that Abl overexpression causes mechanochemical fragmentation of actin networks. This finding could explain the increased developmental errors and actin fragmentation observed in vivo. Our research provides fundamental self-assembly mechanisms for linear and dendritic actin networks also highlights specific mechanochemical properties that have not been observed earlier.
  • Thumbnail Image
    Item
    PHOSPHOLIPID BEHAVIOR AND DYNAMICS IN CURVED BIOLOGICAL MEMBRANES
    (2020) JING, HAOYUAN; Das, Siddhartha SD; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
    Curvature in biological membranes defines the morphology of cells and organelles and serves key roles in maintaining a variety of cellular functions, enabling trafficking, recruiting and localizing shape-responsive proteins. For example, the bacterial protein SpoVM is a small amphipathic alpha-helical protein that localizes to the outer surface of a forespore, the only convex surface in the mother bacteria. Understanding several of these membrane curvature dependent events rely on a thorough understanding of the properties, energetics, and interactions of the constituent lipid molecules in presence of curvatures. In this dissertation, we have used molecular dynamics (MD) simulations to explore how the curvature of the lipid bilayer (LBL), a simplified mimic of the cell membrane, affects the packing fraction and diffusivity of lipid molecules in the LBL, energetics of lipid flip flop in the LBL, and lipid desorption from the LBLs. We have also investigated the interaction between LBLs and a small bacterial protein, SpoVM, which was previously shown to preferentially embed in positively curved membranes. Our work started with simulating convex surface, represented by the nanoparticle supported lipid bilayers (NPSLBLs) in MD. We first quantified the self-assembly, structure, and properties of a NPSLBL with a diameter of 20 nm and showed how the type of the nanoparticle (NP) affects the properties of the NPSLBLs. Second, we studied the energetics of lipid flip flop and desorption from LBLs for the cases of planar substrate supported lipid bilayer (PSSLBL) and NPSLBL. Finally, we investigated the energetics of SpoVM desorption from the PSSLBL and the NPSLBL providing clues to the fundamental driving forces dictating the curvature sensing of SpoVM. In Chapter 1, we discuss the motivation, methods, biological relevance, and the overall structure of this thesis. In Chapter 2, the structure and properties of a pre-assembled NPSLBL were studied. In Chapter 3, we report the MD simulation results on the structure and properties, such as diffusivity, of the lipid molecules within the LBLs of the NPSLBLs formed through the self-assembly route. We compare our findings with that of unsupported lipid bilayer nanovesicles (NVs). Our results show that the structure of the NPSLBLs, although affected by the type of the NPs, is still similar with the free NV consisting of identical number and species of lipid. On the other hand, the properties such as the diffusivity of the lipid molecules within the LBL are significantly different between the cases of NPSLBL and the free vesicle. Results are provided for different combinations of the lipid molecules and the NP materials. The findings described in Chapters 2 and 3 will be eventually useful in long-term for designing new generation of NPSLBLs as drug carrier. In Chapter 4, we focus on the lipid flip-flop and desorption from the LBLs for NPSLBLs and PSSLBLs. We investigated the energetics of a lipid molecule traversing through the lipid bilayer (from inner-to-outer and outer-to-inner leaflet) as a function of the position of the hydrophilic head group of the lipid within the LBL. We obtained the potential of mean force (PMF) by using umbrella sampling. Most importantly, we observed little effect of the curvature in the variation of the lipid flip-flop PMF, establishing that the energetics of lipid migration within the supported bilayer, which implies that energy changes associated with bilayer fluctuations, is independent of the shape of the supported bilayer. The conclusion is supported by the reported experimental results. Next, in Chapter 5, MD simulations are carried out to reveal the energetics of a single SpoVM protein undergoing desorption from LBLs of NPSLBLs and PSSLBLs. The free energy comprises of five different contributions: 1) the free energy change for deforming the protein in the bilayer with respect to the conformation of the protein in the membrane, 2) the free energy change for reorienting the protein in the bilayer about the first Euler angle with the conformation of the protein restrained, 3) the free energy change for reorienting the protein in the bilayer about the second Euler angle with the conformation and the first Euler angle restrained, 4) the free energy change for changing the position of the center of the protein from the membrane to the bulk water with conformation and both Euler angles restrained, and 5) the free energy change for deformation of the protein in the bulk water with respect to the conformation of the protein in the membrane. Through these simulations, we confirmed that SpoVM prefers NPSLBLs rather than PSSLBLs, indicating by a lower free energy change. Additionally, we revealed that the SpoVM membrane sensing is based on the interplay between the packing of the hydrophilic head groups of the lipids and the packing of the acyl chains of the lipids. Our findings reported in Chapter 5 might be helpful in the development of diagnosis and treatment of diseases associated with protein mislocalization.
  • Thumbnail Image
    Item
    Effect of electrostatic interactions on biomolecular self-assembly processes
    (2018) Xu, Hongcheng; Matysiak, Silvina; Biophysics (BIPH); Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
    Molecular level self-assembly processes are not only ubiquitous in living cells, but also widely applied in industry to synthesize and fabricate a variety of nanoscale biomaterials. The emergence of ordered aggregates from disordered components typically requires driving forces from electrostatic interactions to hydrophobic-hydrophilic effects. This thesis aims to elucidate the effect of electrostatic interactions, and the intricate balance between electrostatic and hydrophobic interactions in dictating spontaneous self-assembly processes with three case studies covering various types of biomolecules. For the first case study, we have examined the pH-induced polysaccharide hydrogel network formation. The polysaccharide molecule chitosan forms hydrogels composed of water-filled cross-linking polymer chains. The pH-responsive selfassembly behavior of chitosan hydrogel has been utilized in fabricating nanomaterials with a wide range of applications. To investigate the role of electrostatic interactions in the chitosan hydrogel network formation, we have developed a novel coarse-grained (CG) chitosan polymer model that captures the pH-dependent self-assembly behavior. The structural, mechanical, and thermodynamical properties of chitosan polymer hydrogel have been characterized well in the simulations and agree very well with experimental observations. For the second case study, the anticancer peptide folding induced by phospholipid membrane was investigated. Peptide folding in an aqueous environment is a self-assembly process that has been well studied over the years. However, the folding in a membranous environment is complicated by the heterogeneity in phospholipid distributions and membrane-peptide interactions. To provide information about the driving forces behind membrane peptide folding and the effect of lipid composition on folding behavior, my work has combined our recently developed Water-Explicit Polarizable Protein (WEPPRO) and Membrane (WEPMEM) model to explore the driving forces behind model anticancer peptide SVS-1 folding and how they can be affected by changing the membrane composition. For the third case study, we have studied the formation of nanodomains in mixed lipid bilayers. Phospholipid membranes are essential components in animal cells. The heterogeneous distribution of phospholipids on the membrane bilayer plays an important role in cellular structure and function such as signal transduction and membrane fusion. Interactions between a mixture of lipids and different ligands give rise to interesting patterns that are yet to be understood. Model lipid bilayers with a content of anionic lipids have been shown experimentally to be sensitive to the presence of certain ions. Monovalent cation Li+ induces membrane phase transition similarly as Ca2+ and Mg2+, while distinctive from other monovalent cations like Na+ and K+. We have evaluated the role of electrostatics interactions in the sizedependent cation-induced lipid nanodomain formation with binary mixed bilayers composed of zwitterionic and anionic lipids.
  • Thumbnail Image
    Item
    Atomistic Modeling of Solid Interfaces in All-solid-state Li-ion Batteries
    (2018) Zhu, Yizhou; Mo, Yifei; Material Science and Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
    All-solid-state Li-ion battery based on solid electrolyte is a promising next-generation battery technology, providing intrinsic safety and higher energy density. Despite the development of solid electrolyte materials with high ionic conductivity, the high interfacial resistance and interfacial degradation at the solid electrolyte–electrode interfaces limit the electrochemical performance of the all-solid-state batteries. Fundamental understanding about the solid-solid interfaces is essential to improve the performance of all-solid-state batteries. In this dissertation, I perform first principles computation to bring new understanding about these solid interfaces. Using our developed computation approach based on large materials database, I calculated the intrinsic electrochemical stability window of solid electrolytes and predicted interphase decomposition products. I revealed the effects of different types of interphase layers on the interface stability and battery performance, and also provided interfacial engineering strategies to improve interface compatibility. Lithium metal anode can provide significantly higher energy density of Li-ion batteries. However, only a limited number of materials are known to be stable against lithium metal due to its strong reducing nature. Using first-principles calculations and large materials database, I revealed the general trend of lithium reduction behavior in different material chemistry. Different from oxides, sulfides, and halides, nitride anion chemistry exhibits unique stability against lithium metal, which is either thermodynamically intrinsic or a result of stable passivation. Therefore, many nitrides materials are promising candidate materials for lithium metal anode protection. Since solid electrolytes in all-solid-state batteries are often polycrystalline, the grain boundaries can have an important impact on the ion diffusion in solid electrolytes. I performed molecular dynamics simulations to study the ion diffusion at grain boundaries in solid electrolyte materials, and showed the distinct diffusion behavior at grain boundaries different from the facile ion transport in the bulk. In addition, I studied the order-disorder transition induced by mechanical strain in lithium garnet. Such transition can lead to orders of magnitude change in ionic diffusivity. This series of work demonstrated that computational modeling techniques can help to gain critical fundamental understandings of the solid interfaces in all-solid-state Li-ion battery, and to provide practical engineering strategies to improve the battery performance.