M.S. Theses

Permanent URI for this collection

https://digitalarchive.library.bogazici.edu.tr/handle/123456789/11658

Browse

Now showing 1 - 20 of 306

Molecular dynamics study of the interactions between thymoquinone and lipid bilayers
(Thesis (M.S.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2022., 2022) Ülgey, Esra.; İleri Ercan, Nazar.
Thymoquinone, a phytochemical with antitumor activity, and its derivative aminothymoquinone were investigated using Molecular Dynamics (MD) Simulations to understand their interactions with simple and complex bilayer models. MD simulations of the models were performed using all atom (AA) and coarse-grained (CG) force fields, i.e., the OPLS-AA and MARTINI 3. Although the resolution of the molecules decreased during coarse-graining, the chemical and thermodynamic properties of the molecules were mostly retained. The bond and dihedral distributions validated the matching of AA and CG models, and the free energy calculations showed the reproducibility of new models apart from the agreement with the experimental logP values with less than 10% of error. AA and CG thymoquinone models were used with DOPC and POPC bilayers and the systems were compared with the sole bilayers. The structural properties of bilayers including area per lipid, bilayer thickness, order parameters, and lateral diffusion coefficients were computed. The interaction of CG molecules with two different normal and cancer membrane models was also investigated through the orientation of the molecules in the bilayers, the density distribution, and radial distribution function (RDF) in addition to the methods used for simple bilayer. Both molecules resulted in bilayer thinning with decreased bilayer thicknesses but increased the area per lipid values. Similarly, both molecules decreased the ordering of the bilayers, but the effect was slightly more significant with thymoquinone in normal membrane models. While thymoquinone diffused inside the model membranes, aminothymoquinone preferred to reside near the head groups of lipids. Overall, both molecules interacted similarly, in general, with the lipids of the model membranes, apart from small differences.
A study on mathematical modeling of direct synthesis of dimethyl ether in structured membrane reactors
(Thesis (M.S.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2022., 2022) Köybaşı, Hayrettin Hasan.; Avcı, Ahmet Kerim.
Catalytic transformation of CO2 into dimethyl ether (DME) is modeled in a membrane microchannel reactor. The reactor geometry includes an α-Al2O3 supported SOD membrane layer, capable of in-situ H2O removal, residing between a reaction channel and a parallel permeate channel. Reaction channels are washcoated with a physical mixture of Cu–ZnO/Al2O3 (CZA) and HZSM–5 catalysts, which are responsible for the synthesis and dehydration of methanol respectively. CO2-H2 reactant mixture (H2/CO2=3) is dosed to the reaction channel at space velocity=6x103 ml gcat-1 h-1 whereas, pure H2 at identical temperature and pressure, is fed to the permeate channel to sweep permeated steam. A steady-state isothermal reactor model based on the mass and momentum conservation equations is solved numerically using ANSYS software. Computed performance metrics show minimal deviation from the reported experimental data at 483–543 K, 30–50 bar. Incorporation of energy conservation to the model results in near-constant temperature profile and almost identical reactor performance, which validates isothermal operation. Owing to steam effux from and H2 influx to the reactive stream, membrane integration increases CO2 conversion (~60%) and DME yield (~30%) by a factor of >3 compared to membraneless operation. The benefits become more significant at higher temperature and pressures. Sending sweep H2 counter-currently offers superior H2O removal. Relative inlet velocity of the permeate inlet (vrat) affects membrane mass transfer dramatically. Adopting higher CZA/HZSM-5 mass ratio improves both performance metrics and the reactor capacity. With a ~7 m3 reactor system, 2.76x102 tons of DME can be synthesized from 1x103 tons of CO2 annually, provided that 1MW electrolyzer provides the required H2.
Subsets of slow dynamic modes reveal global information sources as allosteric sites
(Thesis (M.S.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2022., 2022) Altıntel, Bengi.; Haliloğlu, Türkan.
Allostery is a crucial biological regulation mechanism, and dynamic information flow offers a framework to characterize allosteric interactions in causal links. Here, using a novel application of the Transfer Entropy (TE) calculations based on the Gaussian Network Model (GNM), it has been demonstrated how the dissection of dynamic information into subsets of slow dynamic modes reveals various layers of multi-directional allosteric pathways that are intrinsic in a particular protein structure. The degree of collectivity (Col) in the information transfer of residues with their TE values (TECol score) in these subsets of slow modes identifies particular residues as potent effectors and global information sources having a strong dynamic capacity to collectively disseminate information to other residues in the protein structure. These information source residues are linked to known active and allosteric sites, as demonstrated by aspartate transcarbamoylase (ATCase), Na+ /K+ -adenosine triphosphatase (Na+ /K+ -ATPase), and human transient receptor potential melastatin 2 (TRPM2), along with a dataset of 20 proteins. These specific residues provide feasible binding sites for structure-based rational drug design since they together affect/control others and direct pathways of allosteric communication.
Structural dynamics of ABC transporters
(Thesis (M.S.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2022., 2022) Ersoy, Akarun Ayça.; Haliloğlu, Türkan.
The ATP-binding cassette transporters (ABC transporters) translocate sub strates across membranes by ATP binding and hydrolysis. To understand the com plex structural dynamics, transporters were investigated on three case studies: het erodimeric multidrug exporter TmrAB, cystic fibrosis transmembrane conductance regulator (CFTR), and manganese ion importer PsaBC. Anisotropic Network Model Langevin Dynamics (ANM-LD) simulations were performed between four different con formations of TmrAB. The analysis for each transition revealed the allosteric couplings between the ATP binding in nucleotide binding domains (NBD) and extracellular gates on transmembrane domains (TMD). Transfer Entropy (TE) calculations showed that TMDs and NDBs of TmrAB become drivers at different stages during the translocation cycle. The ANM-LD simulations of dephosphorylated to phosphorylated CFTR transi tion revealed that the initial movements in the NBDs are followed by the movement of the gating areas in TM. The causal relationships between these movements were sup ported by TE calculations. CFTR potentiator drug Ivacaftor has been shown to have a global effect, mimicking the effects of ATP binding and gating residues. Corrector drug Lumacaftor has a local effect on TM1, 2, 3, and 6. The dynamics of PsaBC was characterized using TE calculations, and testable predictions were made. Four possi ble gating residues were identified along with an allosteric pocket on TM8. Finally, TE calculations of many ABC transporters showed the differences and similarities be tween different types. Type I transporters were shown to be controlled by TMDs in the inward facing (IF) conformation, and NBDs and substrate binding protein (SBP) in the outward facing (OF) conformation. Type II transporters were always driven by NBDs, regardless of conformation. Type IV transporters showed more variability, but for most, TMDs are driving for IF and NBDs are driving for OF conformations. Dynamic characterization methods show potential in classifying ABC transporters.
Reconstruction of metabolic brain model with the aid of physiologically based pharmacokinetic modelling in the presence of autism spectrum disorder
(Thesis (M.S.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2022., 2022) Erimli, Sena.; Ülgen, Kutlu Ö.
Autism spectrum disorder (ASD) is a neurodevelopmental disorder with only symptomatic medical treatment with increasing prevalence. This study aims to reconstruct a genome-scale metabolic brain model that is expanded via genes and reactions related to ASD and observe changes in dopamine D2 receptors in the presence of antipsychotics. For this aim, iMS570, iNL403, and MODEL1608180000 are merged. After standardization, duplicate reaction removals, deletion of zero flux reactions, and expanding model with ASD related genes and reactions, the final model consists of 1638 reactions, 1358 metabolites, and 756 genes. For further investigations of the autism-specific brain model, transcriptome gene expression data GSE28475 is prepared and integrated. Autism-specific brain model shows mitochondrial and glutaminergic dysfunctions. Final 18360 unique genes are assigned as up-or down-regulated based on threshold determined by downregulated SHANK3 expression and 30% of averages of gene expression data. Additionally, the physiology-based pharmacokinetic (PBPK) model of risperidone and its metabolite paliperidone is simulated. Both drugs are antipsychotics that are used in ASD symptoms. First, simulations are performed to observe population density, CYP2D6 subtypes, ethnicity, total hepatic clearance, and P-gp concentration changes for 2 mg risperidone. PBPK simulation results of experimental articles were in the range of literature findings. Receptor occupancy simulations with chronic dosing underestimate plasma concentrations, but results are still in the range. With down-regulation of the SLC6A3 dopamine transporter gene, which is used to mimic DRD2 gene, proportional to receptor occupancy findings, the effect of risperidone on autism-specific brain model is investigated. Glutaminergic neurotransmission is decreased below healthy brain level, and mitochondrial dysfunction was relieved but still not in the healthy range. Reconstruction of ASD-specific brain model with PBPK modeling enhancement is promising work to better under disease mechanism.
A preliminary work on design and development of sour water gas shift catalysts for synthetic natural gas process
(Thesis (M.S.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2022., 2022) Özata, Hasretnur.; Aksoylu, Ahmet Erhan.
The main goal of this study is to design and develop a model sour water gas shift catalyst showing high performance in terms of activity, stability & selectivity and conducting performance screening by mimicking the syngas composition of the potential coal gasifier under ideal, sulfur-free conditions. According to the literature, dry powder entrained-bed type gasifier seems potential gasifier for the Synthetic Natural Gas (SNG) production processes, and it contains low steam/CO ratio, below 0.8. On the other hand, classical sour water gas shift (SWGS) catalysts work under high steam/CO ratio, specifically, above 2 or 3. In this context, from literature, one of the potential SWGS catalysts which is suitable for dry-powder syngas composition was chosen as a reference. Activated carbon supported KCoRe as the model SWGS catalysts were developed and these catalysts were tested under ideal, sulfur-free syngas composition. The effect of reaction conditions (i.e., reaction temperature and steam to carbon monoxide ratio) and catalyst preparation & support pre treatment methods (i.e., impregnation method, air & nitric acid pre-treatment on the Activated Carbon support) were investigated on the prepared catalyst. Activity and selectivity were determined in terms of conversions of CO and H₂, respectively. Experimental results showed that reaction conditions are very important parameters on the performance of the catalysts. At low steam/CO ratios, it is necessary to increase the temperature in order to get high activity. The effect of the impregnation method may vary from support to support since impregnation depends on the surface chemistry of the support material. Co-impregnation method may not be suitable for air oxidized AC supported catalysts. Both air and nitric acid treatments are helpful to increase the performance of the catalysts by increasing the oxygen bearing surface groups on the AC support.
Supercapacitors based on functionalized carbon materials
(Thesis (M.S.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2022., 2022) Korkut, Ayşe Saliha.; Uralcan, Betül.
Electrical double layer capacitors store energy in the form of electrical charges at the interface between an electrolyte and a high surface area electrode. As their energy storage mechanism relies on physical interactions, EDLCs have high power densities, unlike batteries, which are often limited by the slow charge- and mass- transfer kinetics. Additionally, EDLCs can sustain millions of charging/discharging cycles. Nevertheless, they suffer from low energy densities. For a high energy density EDLC, it is critical to maintain a compact architecture with large ion-accessible surface area while also ensur ing low ion transport and electrical resistance. We incorporate carbon quantum dots into thermally exfoliated graphene oxide sheets in the presence of a room temperature ionic liquid to form conductive carbon networks with improved ion transport networks, enhancing ion transport kinetics and storage. This yields an electrode in which both the carbon quantum dots and the ionic liquid serve as spacers to effectively separate the thermally exfoliated graphene oxide sheets, while the ionic liquid also functions as the electrolyte and carbon quantum dots provide a conductive network. Using this approach, we achieve a gravimetric capacitance of 165 F/g at 30/70 wt% CQD/TEGO composition with 4.5 M EMIM-BF4 electrolyte at 20 mV/s. The electrodes demon strate 70 % capacitance retention at 500 mV/s. When a 3.4 M EMIM-BF4 electrolyte is used instead, capacitance reaches 206 F/g, and the electrodes retain 70% of its ca pacitance at 500 mV/s. This demonstrates that we can simultaneously improve both energy and power density by tailoring the electrolyte composition.
Optimization of energy density in supercapacitors by utilizing a hybrid artificial neural networks-genetic algorithm based optimization algorithm
(Thesis (M.S.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2022., 2022) Kaya, Duygu.; Uralcan, Betül.; Aydın, Erdal.
Energy storage systems are required to meet the increased energy demand and re duce the need for fossil fuels. Energy storage devices are of interest to renewable energy systems and electric vehicles to provide a permanent energy supply. Supercapacitors, in particular, are eligible systems for energy storage owing to their unique properties such as very long cycle life, high reversibility, and high power density. Nonetheless, they have limited energy density. The ultimate goal is to increase the energy density of supercapacitors while maintaining high power density. In this thesis, hybrid artificial neural network-genetic algorithm (ANN-GA) model is utilized to increase the capac itance of supercapacitors. Several data preprocessing, feature selection, and machine learning algorithms are performed to predict the capacitance of supercapacitor by using experimental data. It is observed that ANN is a powerful method to capture nonlinear relationships concerning the physical and operational features of supercapacitors. GA is a promising method that examines search space for the optimal solution. The trained neural network model is used as the fitness function for genetic algorithm to achieve maximum capacitance within the feasible range. Selection, crossover, and mutation procedures are implemented in the reproduction step of GA to offer elaborate search space. In a nut shell, this study takes a step towards the rational design of superca pacitors by implementing a hybrid ANN-GA as an optimization tool to improve the capacitance. The results indicate that obtained optimal design parameters agree with the literature while improving the capacitance of supercapacitors significantly.
Investigation of medulloblastoma metabolism by tissue-specific genome-scale brain metabolic model and identification of therapeutic targets
(Thesis (M.S.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2022., 2022) Özbek, İlkay İrem.; Ülgen, Kutlu Ö.
Medulloblastoma (MB) is the most prevalent pediatric brain tumor arising in the cerebellum. Since conventional therapies decrease life quality and cause deleterious effects on children, computer models are urgently required to simulate cancer phenotypes and determine potential therapeutic targets with minimum side effects on healthy cells. In the present study, metabolic alterations specific to MB were reflected on the brain genome-scale metabolic model by employing transcriptome data. Moreover, the relation between metastasis and the Warburg effects and the pathways utilized by MB without carbon source were investigated. Flux sampling analysis was also performed to detect statistically different reactions in healthy and MB cases. Regulation, flux coupling, and essentiality analyses were conducted as well to find therapeutic targets for MB. Additionally, the antimetabolites which might lessen the use of substrates in cells by causing competitive inhibition were identified by using similarity scores and conducting FBA. To investigate sphingolipid metabolism in depth, 79 reactions were newly included in the MB model. Consequently, the MB model captured metabolic characteristics of MB successfully as confirmed by experimental studies. It was found that targeting proteins/enzymes related to fatty acid synthesis, mevalonate pathway in cholesterol synthesis and inhibition of cardiolipin production, and tumor inducing sphingolipid metabolites might be beneficial therapeutic strategies for MB. Furthermore, the suppression of GABA catalyzing and succinate-producing enzymes simultaneously might be a potential solution for metastatic MB. Using oleic acid as an antimetabolite owing to its structural similarity to linoleate and its downregulation in MB might be also a promising approach for this life-threatening disease.
Multi-scale modelling of supercapacitors : a combined simulation and machine learning approach
(Thesis (M.S.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2022., 2022) Sakallıoğlu, Sabri Hakan.; Uralcan, Betül.
Electrical double layer capacitors (EDLCs) store and release energy via re versible adsorption/desorption of ions at the electrode–electrolyte interface. Research on EDLCs mainly focus on improving their energy density while maintaining their at tractive properties such as high power density and long cycle life. EDLC performance is a complex function of the properties of its components, as well as the interactions between them. Given the large number of parameter combinations make traditional experiments remain infeasible for parameter optimization. To address this problem, we use molecular dynamics simulation data for a set of room temperature ionic liq uid/nanoporous carbon based EDLCs. By analyzing the charging kinetics and equi librium behavior of EDLCs using a transmission line model, we construct a simple data-driven method that is capable of quantitatively predicting energy density and time-dependent charging profile as a function of electrode micropore size and elec trolyte composition. In particular, linear and ridge regression, elastic networks, lasso, and neural network models are trained to predict gravimetric and volumetric capac itance (CG and CV ), charging time (τM), and electrical resistance (Rl). The elastic network model yields the best performance with a root mean square error of 3.10 F/g (CG), 0.15 s (τM), 1.09 F/cm3 (CV ), and 0.54 Ohm m (Rl). This model is then used to construct diagrams that show the dependence of the above-mentioned performance metrics to electrode pore size and electrolyte composition, and allow designing EDLCs with a set of predetermined performance criteria. This work can be extended to provide a framework that can quantify the effect of key factors on the EDLC performance.
Effect of temperature on collective dynamics of proteins: a time series analysis
(Thesis (M.S.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2008., 2008.) Türe, Özlem.; Turgut, Pemra Doruker.
Molecular Dynamics (MD) simulations are used to analyze the internal motions of proteins. In this thesis, the molecular dynamics trajectories of the apo form of dihydrofolate reductase (DHFR), and both apo and holo forms of triosephosphate isomerase (TIM) at three different temperatures, 200 K, 300 K and 400 K are examined. Analysis mainly consists of utilization of two methods: principal components analysis (PCA) to determine the collective protein fluctuations with high mean square fluctuations, and linear time series analysis to examine the collective vibrational motions in detail. Time series model parameters obtained for the free states of DHFR and TIM are similar, indicating the reliability of the analysis. It is found that at high temperatures collectivity reduces and global twisting motion seen in both proteins remarkably diminishes. At low temperatures, the important loop motions are reduced. Vibrational frequencies of the first 40 principal modes are extracted by time series analysis, and probability density functions of these frequencies are plotted to compare different MD runs. It is seen that simulations at higher temperatures have lower frequency distributions. Nevertheless, the difference between 300 K and 400 K is very small compared to the frequency shift from 200 K to 300 K. For its ligand bound form, TIM has higher frequencies than the free form at 200 K, as seen at 300 K in a previous study. However, ligand binding reduces the global twisting motion of the two monomers of TIM remarkably, which is opposite to what has been observed at 300 K. This shows that ligand binding may have different effects on the collective motions at different temperatures.
Au-based catalyst design for selective CO oxidation in hddrogen-rich streams
(Thesis (M.S.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2008., 2008.) Tezcanlı, Sadi T.; Yıldırım, Ramazan,
The objective of this study was to investigate the selective CO oxidation over Au/ -Al2O3 catalyst modified with various metal oxide promoters (such as Mg, Mn, Fe, Ni, Ce, and Co) in hydrogen-rich environment. All catalysts contained 1 wt.% Au, 1.25 wt.% promoter over Al2O3 as a support. The catalysts were prepared by the impregnation of the metal oxides to -Al2O3 support followed by the homogeneous deposition precipitation of the gold over composite MOx/ -Al2O3 support and tested in a microflow reaction system both in the absence and the presence of H2O and CO2. The temperature programmed oxidation (TPO) technique was performed in order to investigate the activity of catalyst in the temperature range of 80-150 oC. The Au loading of the catalysts were verified by Atomic Absorption Spectrometry using an ATI Unicam 929. In order to understand the surface structure and distribution of nanosized gold particles, catalysts were characterized by using high resolution transmission electron microscopy (HRTEM). Compare to Au/ -Al2O3 catalyst, the metal oxide promoted (especially MnOx and MgO) Au/MOx/ -Al2O3 exhibited higher catalytic activity towards CO oxidation at the gradient temperatures. The effect of CO2 in the reaction stream was negative as expected but it was balanced even improved by the addition of H2O for Au/MgO/ -Al2O3 catalyst, which elucidated the best performance under realistic reaction conditions (in presence of H2O and CO2). The catalysts containing 1.25 weight per cent and 2.5 weight per cent Mg exhibited comparable CO conversions while the conversion decreased drastically by the further increase of Mg content to 5 per cent. At the temperatures above 130 oC, the H2O loading of 0 vol.%, 5 vol.%, 10 vol.% demonstrated no sign of enhancement in CO conversion over Au/MgO/ -Al2O3 while the addition of H2O increased the conversion significantly at lower temperatures and this increase become more apparent at the higher H2O per cent. The CO conversion decreased as the CO2 content of the feed increased (mostly appeared in the temperatures of 120-150 oC). This effect was significant when the conversion in the absence and presence of 5 vol.% CO2 were compared, while there was no remarkable difference between 5 vol.% to 25 vol.% CO2 content at operational temperature of fuel cell.
Studies on the genes controlling glucose signaling in yeast
(Thesis (M.S.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2008., 2008.) Şanoviç, Aylin.; Kırdar, Betül.
This project was aimed to study the phosphorylation or dephosphorylation of the proteins involved in glucose sensing, signal transduction and glucose repression pathways by transcription of genes encoding these proteins. S. cerevisiae BY4743 strains (ho /ho , hap4 /hap4 , rip1 /rip1 and RIP1/rip1 ) were cultivated under carbon limited condition in order to identify the variations in the expression levels of genes involved in glucose sensing, signal transduction and glucose repression pathways as a response to system level perturbations. Metabolite profiles for glucose, ethanol, ammonia and glycerol were obtained. Expression profiles to be investigated were of CYC8, GRR1, MTH1, RGT1, RGT2, SKP1, SNF3, STD1,TUP1, YCK1, YCK2, ELM1, GLC7,HXK2, MIG1, PAK1, REG1, SNF1, SNF4, TOS3, HAP4, MBA1 genes. The highest cell density in steady state was seen in rip1 /rip1 strain. The respiratory deficient strains hap4 /hap4 and rip1 /rip1 produced maximum ethanol. The expression of HAP4 downregulated in ho / ho strain, and was upregulated in both deletion mutants of RIP1 of S. cerevisiae showing the lack of the response of HAP4 to glucose pulse caused by respiratory deficiency. MTH1, STD1, YCK1 and YCK2 genes were downregulated in the strains ho /ho , hap4 /hap4 and RIP1/rip1 , and upregulated in the homozygous deletion mutant of RIP1. REG1, GLC7, MIG1, SNF1 and SNF4 genes are upregulated with carbon pulse suggesting their involvement in the glucose repression pathway. MBA1 gene expression upregulated after carbon pulse in ho /ho , rip1 /rip1 and RIP1/rip1 strains, however for the HAP4 deletion mutant, the response to pulse injection wasn’t immediate.
The dynamics underlying enhancement of E2 ligase activity by E3 ligases in sumoylation
(Thesis (M.S.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2008., 2008.) Tozluoğlu, Melda.; Haliloğlu, Türkan.
Covalent attachment of SUMO (Small Ubiquitin Like Modifier) to proteins, sumoylation, is a posttranslational modification that can alter intracellular localization, interactions with other proteins or lead to modifications by other post-translational modifiers. Defects in sumoylation pathway are related to many neurological diseases, such as Huntington’s disease, Parkinson’s disease and more. Additionally, sumoylation is a part of cancer related pathways. Similar to other ubiquitin like modifier (Ubl) conjugation mechanisms, the conjunction of SUMO to targtes involves three groups of enzymes: E1 ligase, Aos1/Ub2 hetero-dimer; E2 ligase, Ubc9; E3 ligases, one of which is RanBP2. Differing from the other Ubl conjunction paths, the E2 ligase, Ubc9, can function without the E3 enzymes but with lower reaction efficiency. One of the target proteins that can be efficiently sumoylated by Ubc9 only, is RanGAP1. Although there are suggested models, it is not clear to date how the E3 enzyme, RanBP2, enhances sumoylation. This thesis mainly aims to identify the conformational/configurational restrictions and allosteric effects that RanBP2 may have on Ubc9-SUMO complex to increase sumoylation rate. Along, the structural motion that drives Ubc9-SUMO complex into association with RanBP2 is also addressed. For this, Ubc9-SUMO and Ubc9- SUMO-RanBP2 complexes from Ubc9-SUMO-RanGAP1-RanBP2 crystal structure are studied by Molecular Dynamics (MD) simulations. The conformational dynamics are elaborated by various means to reflect the equilibrium and dynamic behavior of these complex structures. The results in general suggest that RanBP2 restricts the conformational space of Ubc9-SUMO complex and as well as the orientational space of its monomers with respect to each other. The differences in the network of interactions between Ubc9 and SUMO residues in RanBP2 bound and unbound states suggest the determinants of the restriction in the motion observed. The correlations between the fluctuations of the residues associated with the catalytic activity and the residues that are responsible for the specific target recognition in Ubc9 are shown to be stabilized with RanBP2 binding. The comparative analysis of the dynamics with and without RanBP2 identifies a possible allosteric effect of RanBP2 binding on the mobility and flexibility of specific Ubc9 residues, Asp100 and Lys101, which are functional in target recognition. Additionally, it is seen that the dynamics of Ubc9-SUMO complex displays a pre-existing behavior for the binding to RanBP2. This may in general imply that the dynamics of structures set the sequence of events in the association with others to form complex structures.
Preparation of catalytic cordierite monoliths for the selective oxidation of carbon monoxide
(Thesis (M.S.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2008., 2008.) Döker, Yasemin Asiye.; Önsan, Zeynep İlsen.
The objective of this study was to investigate the state-of-the-art preparation methods for environmentally friendly, engineered, multi-channeled catalysts called monoliths and to develop a preparation procedure for cordierite monolithic Pt-Co-Ce-Al2O3 catalysts used for the preferential oxidation of CO in a hydrogen-rich environment. In this comparative study, monolithic catalysts were prepared by applying a layer of alumina support on to the walls of a cordierite monolithic carrier, termed as wash-coating, and then impregnating active components. The monoliths were wash-coated with alumina using three different methods, namely, colloidal coating, slurry coating and aluminum nitrate coating. Metal precursor solutions containing Pt-Co-Ce were impregnated over alumina wash-coated monoliths. Catalyst compositions were determined by atomic absorption spectrometry. Total surface area measurements, environmental scanning electron microscopy (ESEM) and optical microscopy were employed to physically characterize the catalysts prepared. The most promising monolithic catalyst preparation method was determined to be colloidal coating. Colloidal alumina coated and Pt-Co-Ce impregnated monolithic catalysts were tested for the preferential CO oxidation reaction in a micro-reactor flow system under a total flow of 100 cm3 min–1 using reaction temperatures in the 383-443 K range and typical feed compositions. The monolithic catalyst was found to exhibit the highest catalytic activity towards CO oxidation approaching 100 per cent CO conversion at a temperature of 443 K in a feed containing 1.0 volume per cent CO, 1.0 volume per cent O2 and 60.0 volume per cent H2 and balance helium.
Kinetic modeling of CO oxidation by genetic programming
(Thesis (M.S.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2008., 2008.) Yarangüme, Verda.; Yıldırım, Ramazan,
In this thesis, the kinetic modeling of CO oxidation was performed using genetic programming. A reaction rate equation was created from the experimental data, and then this equation was used to predict the mechanism of the reaction. Firstly; the generic reactions, of which both rate equation and mechanism were known, were studied in order to test the applicability of genetic programming in generating the rate expressions and to have a basic understanding about the method. The functions used in program were obtained from the functional terms commonly appears in catalytic rate equation. It was verified that the rate expressions derived using genetic programming were quite similar in terms of their structures and the groups that describes the main features of the rate equations in the literature. Then, the catalytic CO oxidation reaction was modeled to derive the model equations using three different experimental data sets; one of them was generated in our laboratory and the remaining two were obtained from the literature over various catalytic systems. After generating possible model equations, they were statistically evaluated by comparing with the experimental results and the other models proposed in the literature. The plausible models were then used to understand the mechanism of the reaction by analyzing the form of the rate expressions and the value of the parameters. The results were generally satisfactory, and it was concluded that genetic programming can help to understand the mechanism and the kinetics of the similar catalytic reactions.
Molecular dynamics of substrate recognition and co-evolution in HIV-1 protease
(Thesis (M.S.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2008., 2008.) Özen, Ayşegül.; Haliloğlu, Türkan.
Human Immunodeficiency Virus Type 1 (HIV-1) protease recognizes at least ten cleavage sites as its natural substrates. There is little sequence homology between these substrates and they are asymmetric around the cleavage site in both charge and size distribution. Thus, understanding of the molecular determinants of substrate recognition is a challenging task as well as of great importance in design of effective drugs. The protease-substrate complex crystal structures indicate that the substrates occupy a remarkable uniform region within the binding site, which has been termed as the substrate envelope. Nevertheless, the activities of proteins are intimately related to the dynamics, from local to global motion of the structure. To this end, an elaborated analysis on both structural and dynamic features of seven HIV-1 protease-substrate complex structures are carried out by molecular dynamics (MD) simulations in the present thesis. The conformations of the complex structures in time have been analyzed with respect to the interaction of the substrate with the protease in terms of the substrate volume, the changes in the van der Waals (vdW) contacts between the two, and the dynamics of both substrate and the protease in general. On the other hand, the co-evolution of the substrate peptides with the drug-resistant protease variants is also analyzed. The MD simulations for the p1-p6 substrates (wild-type and LP1’F) in complex with the protease variants (D30N, N88D, and D30N/N88D) were run and similar analysis to those in wild-type complex structures were made. In this work, the substrate recognition has been observed to be an interdependent event and the recognition mechanism may not be the same for all natural substrates. Also, the dynamic substrate envelope has been found to be smaller than the crystal structures suggest. The analyses of the mutant structures have shown that the substrate recognition is altered when there is drug resistance and this alteration is compensated by co-evolution. The results reveal that the conservation of the peptide conformational preferences and dynamic behavior of the complex structure appears to be important for protease substrate recognition.
The use of different statistical tools in the identification of perturbation-responsive transcription factors in yeast
(Thesis (M.S.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2008., 2008.) Çınar, Nil.; Kırdar, Betül.
S. cerevisiae transcription data, obtained under environmental perturbations, such as 16 different macronutrient (carbon, nitrogen, phosphorus and sulfur) limitation regimes in both aerobic and anaerobic conditions, and under genetic perturbations such as the deletions of MIG1 and both MIG1 and MIG2 genes, were integrated to transcriptional regulatory network in order to find the so-called key transcriptional factors (key TFs), meanining to transcriptional factors around which significant changes occur as a perturbation responsive behaviour. Key TFs were identified by integrating the processed transcriptional regulatory network, which consists of 8494 regulatory interactions between 144 transcriptional factors and 3399 target genes, with transcription data. Two probability methods, t-test and EDGE program were used to analyze the transcriptome data. The comparison of the key TFs identified by using two different statistical tools revealed that the application of these two different tools to the same triplicate data set can identify the same set of key TF responsive to genetic perturbations, and to carbon limitation between anaerobic and aerobic conditions. EDGE can therefore replace t-test in the application of the reporter TF algorithm. Dynamic non-replicate S. cerevisiae transcription data, consisting of expression levels obtained at different time points after the glucose pulse was given at the first steady state, resulting in totally 14 time point measurements until the second steady state, was analyzed in order to identify key TFs via cumulative Z-scores calculated using p-values by EDGE. Interacting key TF pairs were identified and their ranking was followed at the time points, and it was observed that interacting key TFs show highly similar changes in ranking order.
Determination of protein-protein binding sites using machine learning tools
(Thesis (M.S.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2008., 2008.) Sümbül, Fidan.; Haliloğlu, Türkan.
Protein-protein interactions are involved in almost all biological processes. Thus, the understanding of the principles underlying these interactions is of great significance. This is mainly to identify the functional sites in proteins and study how proteins function. The whole surface of the protein is not available for interaction with other proteins. There are some distinctive properties that differentiate binding residues from the rest of surface residues. To explore and further to predict the binding interfaces, the present work is composed of two sections. The first part is the identification of differentiating properties for three main groups of residues in a protein, namely, core, binding and non-binding surface residues on a database of 263 proteins. These properties are sequence and structure related characteristics, and as well dynamic peculiarities, of residues such as; the residue propensity, hydrophobicity, side chain polarity and charge, conservation, accessible surface area, and the fluctuations. Some residues prefer being at interface or core rather than the non-interface surface. The hydrophobic residues are favored at interface or in core of the protein. Positively charged polar residues are abundant at interface while the non-polar or polar but neutral ones are mostly found in the core. The interface and core residues have also higher conservation scores. The residues that have higher fluctuations with rest of the residues in the fastest and in the slowest modes by Gaussian Network Model (GNM) are mainly located at interface of proteins. These aforementioned properties are also analyzed in terms of the type of interactions, namely, homogeneous versus heterogeneous complexes and transient versus permanent complexes for a further understanding of the interaction sites. In the second part, these properties are used to predict the binding residues of proteins using support vector machines (SVM) and multiple kernels learning (MKL). Both of these methods are supervised classifier. The maximum accuracy obtained by SVM is 81.3 %, which is the highest observed accuracy in binding site prediction over the literature. The contributions of the grouped properties to the final results are determined by MKL. The type of amino acid, conservation score, accessible surface area and state of the amino acid (core or surface), relative correlations between fluctuations in both fast and slow modes, and the packing of the residue have the most contribution.
Adaptive Clarke-Gawthrop self-tuning control of blood glucose concentration in patients with Type I diabetes
(Thesis (M.S.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2008., 2008.) Çölmekçi, Ceylan.; Çamurdan, Mehmet Cihan.
Adaptive Clarke-Gawthrop type of self-tuning controller is used for regulating the blood glucose concentration in Type I diabetic patients. This control algorithm is proposed to be integrated into the automation pumps used by diabetic patients as a replacement for manual insulin injection. The controller is implemented on an educational software, GlucoSim which simulates the glucose-insulin dynamics in Type I diabetic patients with insulin infusion being the manipulated variable. The performance of the controller is investigated by changing some important parameters that should be specified at the beginning of the simulations. These parameters are the initial values of the controller parameters, the covariance matrix P, the setpoint of the blood glucose concentration, the constraint factor, the clamp value of the manipulated variable, the forgetting factor, and the control interval. The simulations are carried out for two cases: when the process and disturbance model parameters are unknown and when they are approximately known. The best performance is obtained when the process model parameters are unknown. The optimum parameter settings found are then as follows: the setpoint of the blood glucose concentration is 100 mg dl-1, the constraint factor is 1.5, the insulin infusion is clamped when it exceeds 50 000 mU min-1, the forgetting factor is 0.5 and the control interval is 5 minutes.

Browse

Recent Submissions