Graduate School of Engineering and Science

Permanent URI for this collectionhttps://hdl.handle.net/11693/115678

Browse

Now showing 1 - 20 of 3467

Open Access
Using zebrafish disease model in molecular research on neurodevelopmental disorder generation
(2025-01) Başcı, Ayşe Reyyan Kutan
In this thesis, zebrafish were used as a model organism to study two different projects: issues stemming from the dopaminergic system during the neurodevelopmental period, specifically ADHD, and a Parkinson’s disease model representing dopaminergic system disorders in old age. This approach allowed for an investigation of the lifelong effects of disorders caused by dopaminergic system dysfunction. Attention-deficit/hyperactivity disorder (ADHD) is a prevalent condition characterized by persistent inattention, hyperactivity, and impulsivity, often leading to significant impairments in daily functioning. Numerous medications have been developed to manage ADHD, with methylphenidate (commonly known as Ritalin or Concerta) being the most widely used active compound. While methylphenidate effectively improves attention, focus, and emotional regulation, its long-term use can lead to behavioral issues such as increased depression and anxiety, as well as physical side effects like sleep disturbances and decreased sensitivity to rewards. These side effects are more challenging to compensate for in adults compared to children, which may leave permanent effects. This study aims to investigate the long-term effects of methylphenidate use in adult women, with a focus on its role in sleep disorders, circadian rhythm disruption, and the potential implications for pregnancy, specifically on the susceptibility of offspring to accelerated brain aging. Our hypothesis is that methylphenidate use during pregnancy may contribute to changes in offspring telomere length and gene methylation patterns associated with brain aging, thereby increasing their vulnerability to neurodegeneration. To evaluate this, telomere length and methylation analyses were conducted on genes linked to brain aging in second-generation offspring. Additionally, our hypothesis is that therapeutic interventions, such as melatonin for regulating sleep disturbances and oxytocin as an alternative to mitigate methylphenidate’s side effects, may have protective effects. Zebrafish were used as the model organism in this study due to their high genetic similarity to humans and their ease of egg production, enabling multi-generational studies. As a result of this study, it was determined that neuroinflammation caused by circadian rhythm disruption and maternal stress was reduced through melatonin & methylphenidate and oxytocin & methylphenidate combinational treatments. In the offspring of the next generation, it was observed that the telomere length inherited at birth was shorter when treated with methylphenidate only. In the second research project of this thesis, the origins of Parkinson’s disease from two distinct locations—the brain and the gut—were investigated. Parkin-son’s disease is a neurodegenerative disorder characterized by the loss or reduction of dopaminergic neurons in the central nervous system. It is associated with aggregation of alpha-synuclein fibrils and resulting in motor function impairments. In the previous phase of this research, transgenic zebrafish models of Parkinson’s disease were developed by integrating human alpha-synuclein gene into the zebrafish genome at the single-cell stage. In this phase of thesis, six-month-old transgenic zebrafish were used to test the hypothesis that Parkinson’s disease can originate from either the brain or the gut. Alpha-synuclein protein was injected into the brain and gut, and its migration between these two organs was analyzed. The migration of alpha-synuclein fibrils was validated using immuno-histochemistry techniques. Behavioral changes and motor function impairments were assessed using novel tank tests, swim endurance tests, and hyposmia tests. As a result of the study, it was validated that alpha-synuclein fibrils injected into the brain-to-gut and gut-to-brain migrate at different speeds.
Embargo
Investigation of the role of IKK-related kinases in colitis-associated cancer
(2025-01) Özkurt, Ahmet Göktuğ
Colitis-associated cancer (CAC) is a serious consequence of persistent colonic inflammation, predominantly occurring in individuals with prolonged ulcerative colitis. The molecular processes underlying CAC include chronic inflammation, immunological dysfunction, and abnormal activation of oncogenic signaling pathways. IKK-related kinases are pivotal regulators of immunological signaling, inflammation, and cancer. Nonetheless, their precise functions in CAC remain inadequately understood. In this study, to understand the effects of IKK-kinase inhibition on the onset and the development of CAC, three experimental setups that mimic acute, chronic and cancer phases of the disease were employed by AOM/DSS induced CRC mouse model. Clinical and histological activity scoring were conducted across acute, chronic, and cancer phases to assess the effect of pharmacological inhibition of IKK-related kinases on disease severity, tumor burden, and mouse survivals. Furthermore, we examined cytokine and chemokine profiles, immune cell infiltration, and oxidative stress markers to ascertain the role of these kinases in the inflammatory tumor microenvironment. Finally, to establish the clinical relevance of our findings, publicly available single-cell RNA sequencing data of ulcerative colitis patients were analyzed to identify potential parallels in human disease. It has been demonstrated that the modulation of immune response and the initiation and development of tumors in the CAC are significantly impacted by IKK-related kinase activity.
Open Access
Transcriptomic investigation of nanoparticle toxicity on HepaRG spheroids
(2025-01) Erden, Merve
Nanoparticles (NPs) are used in diverse fields such as cosmetics, drug delivery, food industry and medicine. They are easy to produce and easy to modify according to the needs of the application. However, there is an arising concern about the adverse effects of the nanoparticles. Negative impacts of the nanoparticles on human body needs further analysis because nanoparticle exposure of human body increases every day with the widespread usage of these materials in daily products. In literature, the studies about the nanoparticle toxicity remain inadequate since they generally use insufficient number and non-variety of nanoparticles, inappropriate cell line or cell culture method. This study focuses on evaluation of nanoparticle toxicity with transcriptomic analysis. With the help of these analysis, we aim to find the genes that respond to NP toxicity, which might help us better understand the mechanism of the toxicity. HepaRG cells were used as the best alternative to primary hepatocytes. Additionally, to get the results most similar to the in vivo liver organ, HepaRG cells were cultured in 3D spheroids then they were incubated with eleven different nanoparticles of varying materials and sizes. To reveal the genes that responds to nanoparticle toxicity and to find the toxicity mechanisms of nanoparticles on HepaRG spheroids, transcriptome sequencing (RNA-seq) was conducted. Anticipatedly, cellular processes such as drug metabolism, sensitivity to metal ions, oxidative stress, EMT and apoptosis pathways showed enrichement in the DEGs of all nanoparticles. The common and unique genes inbetween different types and sizes of NPs were noted. These results offer novel findings about the toxicity mechanisms of HepaRG spheroids in response to NPs.
Open Access
Enhanced direct quadrature control implementation for a three-phase inverter
(2025-01) Giziroğlu, Batuhan Berkay
Power electronics is widely utilized in various fields, such as the military, daily life, and industry. The efficiency of power conversion applications is important. Three-phase inverters are primarily used in motors and UPS systems. However, the design of a three-phase inverter requires advanced knowledge of software, hardware, and algorithms. The digital control of the three-phase inverter provides flexibility for the adjustment of output voltage and frequency. The combination of Sinusoidal Pulse Width Modulation (SPWM) and Direct Quadrature (DQ) control increases the stability of output voltage. In this thesis, we present an implementation of a three-phase voltage source inverter using the Digital Signal Processor (DSP) based controller of TMS320F28379D. There are three electronic cards in the system. The first card includes Switched-Mode Power Supplies (SMPS), current sense circuits, and three half-bridges using discrete MOSFETs. The second card comprises of a low-loss filter and a voltage sensing circuit. The input voltage of the three-phase inverter is 30 V, and the output is a three-phase inverter with an output voltage that can be adjusted up to 12 V with an output frequency of 50 Hz. The maximum power of the inverter is 98.5 W. The efficiency of the inverter is 90.72 % at the full load.
Open Access
Investigation of IRE1 and JNK signaling pathways in DSS induced ER stress in the mouse cerebral cortex
(2025-01) Sobacı, Sanem
Endoplasmic reticulum (ER) stress plays a critical role in cellular homeostasis and is implicated in various neurodegenerative diseases. The inositol-requiring enzyme 1 (IRE1) is one of the arms of the unfolded protein response (UPR), which is activated in the presence of ER stress. IRE1 pathway activates c-Jun N-terminal kinase (JNK) downstream under prolonged stress. This thesis explores the activation of the IRE1 and JNK signaling pathways in the mouse cerebral cortex following DSS-induced colitis, focusing on their roles as markers of ER stress in the context of the gut-brain axis. While DSS-induced intestinal inflammation and ER stress are well-established, neurological effects remain less understood. Employing a murine model, the study explored ER stress markers in the cerebral cortex resulting from intestinal pathology. Despite evidence of DSS-triggered systemic inflammation and ER stress in intestinal tissues, our study revealed no significant differences in the expression levels of IRE1, p-IRE1, or the p-IRE1/IRE1, nor in JNK, p-JNK, or the p-JNK/JNK between the control and DSS-treated groups. Additionally, these results are supported with correlational and linear discriminant analyses (LDA). These findings suggest that acute DSS-induced colitis did not elicit a detectable ER stress response in the mouse cerebral cortex under the conditions used. Possible explanations include tissue-specific reactions to the ER stress, transient activation of the IRE1-JNK pathway that returned to baseline by the time of analysis, or potential survivor's bias. Despite its limitations, this thesis provides a novel investigation into the effects of DSS-induced colitis on ER stress in the cerebral cortex.
Open Access
Novel observer based friction estimation and control methods for simple nonlinear mechanical systems
(2025-01) Orta, Emre
Friction is a common nonlinear phenomenon in inertial mechanical systems, often leading to undesirable effects such as stick-slip motion, hysteresis, and reduced tracking accuracy. Effective friction compensation is essential for enhancing robustness and achieving high precision in controlling such systems. Both model-based and non-model-based approaches have been widely utilized for friction compensation, with friction observers playing a significant role in estimating the friction acting on the system. This study focuses on observer-based adaptive estimation techniques, specifically employing the Friedland-Park observer to estimate the parameters of a friction model in an inertial system. The proposed approach aims to mitigate stick-slip motion and enhance tracking performance. The research evaluates various applications of the Friedland-Park observer for friction compensation and its implementation in high-order systems. Both two-state and single-state observers are examined, alongside different system models and controllers, to ensure the robustness of the proposed methods. Performance metrics include velocity reference tracking accuracy and the compensator's responsiveness during velocity sign reversals, assessed under diverse reference input conditions. Simulation results demonstrate a significant improvement in velocity tracking accuracy, with up to a 65% reduction in tracking error. The proposed friction compensation methods effectively handle varying friction conditions, ensuring system robustness and precision. This study contributes a comprehensive approach to friction compensation, offering various design options for inertial mechanical systems. Future work will extend this research by incorporating a gyroscope model to address the noisy nature of inertial sensors and performing the hardware implementation of the proposed methodology on gyro-stabilized platforms to validate its practical applicability. This work can also be extended within the scope of system identification by developing a novel approach for estimating plant parameters through the adaptation of Friedland-Park observer equations.
Open Access
Transition metal dimers on GAAS(110) surface: a DFT study
(2025-01) Alaybeyoğlu, Ahmet
Dilute magnetic semiconductors (DMSs) have gained appreciable interest in the past two decades. This is due to the perspectives of both fundamental physics and novel applications. They are useful in the context of single dopants (solotronics) because studying individual dopants and their interactions with the host pro-vides rich physics. Here we carry out density functional theory calculations for transition metal element dimer dopants on GaAs(110) surface using Quantum Espresso software by employing ultra-soft pseudopotentials (USPP) and report properties such as exchange energy, spin-resolved density of states, spin-resolved projected density of states, structural and magnetic relaxations, and STM images of surface with dimers. We model the GaAs(110) surface with the BURAI software package, and we study 4 transition metal dimers on near and far configurations on the surface, Fe, Cr, V, and Co. Since the dopants are magnetic, we consider both the ferromagnetic and antiferromagnetic alignments. We show that magnetic configurations greatly alter the relaxed positions of dopants. We also report the exchange energy between ferromagnetic (triplet) and antiferromagnetic (singlet) states for near and far configurations. At the end of our calculations, we found the relationship between exchange energy and distance and reported a dramatic change in the magnitude of the exchange energy as a function of dimer separation. We also observe that except for the Co dimers, Fe, Cr, and V dimers have a reasonable exchange energy between ferromagnetic and antiferromagnetic alignments. This is due to nonzero magnetic moments on these transition metal dopants. For less than two percent of dopant ratio, we observed a metallic system for Fe dimer, a semiconducting system for Cr, and a half-metallic system for V dimers. Electronic structure calculations such as DOS and PDOS projections are in parallel with the expectations of ferromagnetic and antiferromagnetic alignments. We also study scanning tunneling microscopic images of these dimers on the surface showing a large contrast under a bias voltage of 1V. Semiconducting surfaces such as GaAs hosting transition metal dopants with d-orbitals in this thesis have the potential to be excellent platforms for novel quantum applications such as fast-switching qubits. Also, reported spin-resolved density of states (DOS) and spin-resolved projected density of states can be used to understand the hybridization of bands, exchange splitting on orbitals, and design materials for future spintronic and optical applications.
Open Access
Deep learning based decoders for concatenated codes over insertion and deletion channels
(2025-01) Kargı, Eksal Uras
Channels with synchronization errors, including insertion/deletion channels, are of significant importance, as they are encountered in various systems, such as communication networks and various storage technologies, including DNA data storage. Serially concatenated codes where the outer code is a powerful channel code, such as a low-density parity-check (LDPC) or convolutional code, and the inner code is a watermark or marker code, are shown to be effective solutions over such channels. In particular, the use of marker codes, referring to insertion of preselected sequences in the transmitted data stream periodically, are shown to work well in regaining synchronization at the receiver and achieving improved error rate performance compared to other alternatives. In the current literature, maximum a posteriori (MAP) detector realized by the well-known forward-backward algorithm is commonly employed to decode the inner marker code and estimate the log-likelihood ratios (LLRs) of the bits encoded by the outer code, and the resulting log-likelihood estimates are fed to the outer decoder to estimate the transmitted data. Alternative to the MAP detector, this thesis proposes deep learning-based solutions to estimate the LLRs of the coded bits in the paradigm of concatenated codes, exploiting the marker information and addressing some limitations of conventional methods. Bit-level deep learning-based detectors offer good alternatives when the channel statistics are not perfectly available at the decoder, degrading of the performance of the MAP detector. They can also be employed for one-shot decoding when the outer code is a convolutional code. Also developed are symbol-level deep learning-based detectors to exploit the correlations among adjacent bits at the detector output. Contrary to the existing symbol-level decoders for insertion/deletion channels, the newly proposed approaches can go beyond the case of combining three bits, offering further enhancements in performance while keeping the complexity tolerable. As a final contribution, deep learning-based detectors are developed for insertion and deletion channels that are further exacerbated by inter-symbol interference, e.g., modeling bit-patterned media recording channels, and their performance is studied via numerical examples.
Open Access
Addressing encoder-only transformer limitations with graph neural networks for text classification
(2025-01) Aras, Arda Can
Recent advancements in NLP have been primarily driven by transformer-based models, which capture contextual information within sequences, revolutionizing tasks such as text classification and natural language understanding. In parallel, GNNs have emerged as powerful tools for modeling structured data, leveraging graph representations to capture global relationships across entities. However, significant challenges persist at the intersection of these fields, limiting the efficacy and scalability of existing models. These challenges include the inability to seamlessly integrate contextual and structural information, computational inefficiencies associated with static graph construction and transductive learning, and the underperformance of models in low-labeled data scenarios. This thesis explores and addresses these challenges by developing novel methodologies that unify transformers and GNNs, leveraging their complementary strengths. The first contribution, GRTE, introduces an architecture that combines pre-trained transformer models with heterogeneous and homogeneous graph representations to enhance text classification in both inductive and transductive settings. Compared to state-of-the-art models, GRTE achieves significant computational efficiency, reducing training overhead by up to 100 times. The second contribution, Text-RGNN, proposes a relational modeling framework for heterogeneous text graphs, enabling the nuanced representation of diverse interactions between nodes and demonstrating substantial accuracy improvements of up to 10.61% over existing models, particularly in low-labeled data settings. Finally, the third contribution, VISPool, introduces a scalable architecture that dynamically constructs vector visibility graphs from transformer outputs, enabling seamless integration of graph-based reasoning into transformer pipelines while improving performance on NLP benchmarks such as GLUE, with performance improvements of up to 13% in specific tasks. Through comprehensive experimentation and benchmarking against state-of-the-art models, this thesis establishes the efficacy of these proposed methodologies. The results demonstrate the potential for improved performance, scalability, and the ability to address long-standing challenges in NLP and GNN integration. These contributions lay a robust foundation for future research and applications at the intersection of graph-based and transformer-based approaches, advancing the state of the art in text representation and classification.
Open Access
Fabrication of mesoporous nickel oxide based thin film electrodes and their electrochemical properties
(2025-01) Katırcı, Assel Amirzhanova
In this thesis, robust electroactive mesoporous Ni1-xMnxO thin-film electrodes were synthesized on FTO and graphite rod substrates. Molten salt-assisted self-assembly (MASA) synthesis method was employed to produce uniform thin films. The synthesis started with preparing ethanol solutions, containing various molar ratios of [Mn(H2O)4](NO3)2 and [Ni(H2O)6](NO3)2 (between 1.0 to 0.1 Ni(II)/Mn(II) ratio ) and surfactants (C12H25(OCH2CH2)10OH, C12E10 and C16H33N(CH3)3Br, CTAB). Then, these solutions are coated over conductive substrates to obtain the salt-surfactant lyotropic liquid crystalline (LLC) mesophase. The thin mesophase is calcined in order to produce mesoporous Ni1-xMnxO thin-films on the FTO or graphite. The thin-films form solid solutions with the x value of up to 0.7. The Ni1-xMnxO thin-films transform to NiMnO3, Mn3O4, and Mn2O3 phases at increased Mn ratios and annealing temperatures. The films are mesoporous and were confirmed by N2 adsorption-desorption analysis and typical type IV isotherms characteristic for mesoporous materials. Pore sizes varied from 2.8 to 17.6 nm from Ni-rich to Mn-rich oxides. The surface area reaches to 211 m2/g in Ni0.9Mn1O, while the pure NiO has a BET surface area of 164 m2/g at 350 oC calcination temperature. The FTO and graphite-coated electrodes (FTO-Ni1-xMnxO and G-Ni1-xMnxO) display high charge capacities, but the FTO coated electrodes are unstable and undergo to degradation over extended time of usage. In the first few CV cycles of the FTO-based electrodes, they show an increased capacity, however, decline in further cycles. On the other hand, graphite-based electrodes show better stability and high charge capacity. Origin for increasing charge capacity with cycling is attributed to a transformation of the metal oxides to metal hydroxides. Thus, the electrochemical CV cycling of both pure NiO and Ni1-xMnxO electrodes results in a structural change into a NiO(core)/Ni(OH)2(shell) or Ni1-xMnxO(core)/Ni(OH)2(shell) configurations. The shell thickness is ranged from 2.0 nm (pure NiO) to 1.1 nm (Ni0.9Mn0.1O) at 350°C. Moreover, the shell thicknesses and charge capacities are affected by the pore-wall thicknesses, which increases with increasing annealing temperature. Despite these changes, the manganese addition improves the stability of the electrodes, but there is no improvements on the overpotential on oxygen evolution reaction (OER). Moreover, the annealing temperature reduces the charge capacity, whereas the OER performance remains the same. By using the same MASA method, m-NiO-SiO2 electrodes were synthesized using [Ni(H2O)6](NO3)2 and tetramethyl orthosilicate (TMOS) with CTAB and C12E10 surfactants at different Ni to TMOS ratios. Silica acts as hard template support for NiO, and the film is formed in good quality with bimodal pore size distribution. The sample pore size that was observed is 2.6 nm, which originates from the m-SiO2 domains. The second pore system had also mesopores; the average pore size is 15 nm, calcined at 350 oC. That property helps better infiltration of electrolytes, which is advantageous during electrochemistry. During electrochemical analysis, silica is etched out in basic electrolyte. These electrodes, prepared on graphite substrate have specific surface area of around 130 m2/g. The electrodes show an overpotential 381 mV in the CP experiment at 10 mA/cm2 current density.
Open Access
Aerospace metamaterials and functional coatings
(2025-01) Astarlıoğlu, Aziz Taner
Optically transparent and electrically conductive thin-film coatings are widely used to functionalize surfaces of various high-technology platforms, including mobile phones, displays, detectors, and LEDs. Their integration into aviation transparencies, such as canopies, windshields, and windows, is widely known and used for de-icing purposes. However, there are limited reports or information available about using such thin-film coatings for electromagnetic interference (EMI) shielding, low observability (LO), and solar irradiation protection features. This thesis aims to study and demonstrate optically transparent aviation structures possessing the properties of EMI shielding, LO features, and solar radiation protection altogether. To this end, in this thesis, we specifically addressed the problem of achieving high EMI shielding and solar protection, which require high electrical conductivity, resulting in a trade-off reducing the optical quality and LO performance. Transparent engineering polymers are widely used in structural parts in aviation thanks to their enhanced mechanical performance. However, good-quality films require high-temperature processes, which is not applicable to transparent aviation structures. Therefore, the architecture of layered films can be applied to meet the requirements of well-featured aviation transparencies. For these purposes, in the thesis we also designed monolithic and laminated aviation transparencies with surface modification based on stratified films and their patterned ones using numerical and experimental methods. We developed numerical approaches for the design of aircraft transparencies, including both the optical and electromagnetic requirements and validated our results. We successfully conducted experimental studies for uniform large-area thin-film coatings onto aviation transparencies. The results revealed that EMI shielding and solar control performance were achieved with minimal optical losses for planar structures. LO requirement was incorporated into prototypes built on curved or laminated transparent structures instead of monolithic ones to sustain optical, solar protection, and EMI shielding performance to a possible extent. We showed that the low observability performance of such patterned structures, the metamaterials, is enhanced in terms of bandwidth and attenuation compared to the planar thin-film-coated monolithic counterparts. These multi-functional thin-film coatings are essential in aviation, especially for high-performance 5th-generation fighter jets and other civil applications. This thesis paves the way for thin-film-coated transparent aviation structure designs across different domains, including visible, infrared, and microwaves, to enable their multi-functionality at large scales. The experimental large-area coating method guides the coating of a large and complex area to remove the limitation of metamaterial applications at the industrial level. We believe that our findings in this thesis will help to replace traditional planar thin film coatings with metamaterials at the industry scale, aiming to outperform traditional counterparts.
Open Access
Generation of transgenic flies to uncover the role of the intrinsically disordered regions in transcriptional regulation using Drosophila melanogaster transcription factor Bicoid
(2025-01) Akdoğan, Dilan
Multicellular organisms develop from a single cell into a complex organism. Their development is strictly controlled by transcription factors that control the gene expression in a context-dependent manner, so that each gene is expressed at the right time and place. The specificity of transcription factors is determined mainly by their DNA-binding domains; however, their activity and interaction with DNA, proteins, and small molecules are modified by the effector domains (EDs). The EDs are generally low-complexity regions with high flexibility, often intrinsically disordered. Intrinsically disordered regions (IDRs), with their flexible and adaptable nature, help navigate protein activity in a context-dependent manner. The maternal morphogen Bicoid is a transcription factor responsible for the anterior development of Drosophila melanogaster embryos that regulates the expression of hundreds of genes responsible for Drosophila segmentation. It is absolutely required for embryo development and its absence results in the replacement of anterior structures by posterior ones. It consists of a 60 amino acid long structured homeodomain (HD) flanked by IDRs on both amino- and carboxyl termini. Through full Bicoid and HD swap experiments between D.melanogaster, Lucilia sericata, and Calliphora vicina, we found disordered EDs are required for Bicoid's full developmental functions. We used the Drosophila genetic manipulation tool recombinase-mediated cassette exchange to generate Bicoid ED mutant flies. We aim to elucidate the role of IDRs in regulating the activity of Bicoid via phenotypical and molecular analyses.
Embargo
Spider orb web, investigating structural features and using biomimicry for lattice design
(2024-12) Yavuz, Koray
Spider orb web has evolved to fulfil multiple roles, such as catching prey and acting as a sensing tool. For orb-weavers, the web must stop and retain prey, which can fly into it with considerable momentum. Considering orb weaver spiders are mostly blind, the web must also transmit accurate information about the prey’s location. There are many web features aiding these roles; some are known, and some are waiting to be shed light on. Considering these two cases, there are also two parts in this thesis, the first part is about a particular web-building behaviour of spiders and how it affects the signal transmittance, and the other is about creating a new lattice design for energy absorption utilizing some of the known structural characteristics of the spider web. The first part of this study is about designing a new lattice (SW) for energy absorption inspired by the structure of spider webs. Spider orb web comprises four structural elements: anchor, frame, radial, and spiral threads. The first three are the main components that provide structural integrity. These components have a hierarchical nature; the anchors bind to the environment and are generally thickest, while radii form the innermost part with the thinnest threads. The frames make up the connection between the anchor and radii; thus, there is no direct connection between them, and they generally have a thickness value between the radii and anchor threads. These features help the spider orb web to be a resilient, efficient structure for energy absorption, so using the same properties, a 3D lattice was designed for energy absorption. This design is then optimized for improved Energy Absorption Efficiency(EAE) and Energy Absorption(EA) value. The second part is about web-building behaviour that seems counterintuitive, heavier spiders increase the pretension of the threads as they get heavier, which diminishes the ability of the web to stop and retain prey. To investigate this behaviour, a spider web model with controllable pretension is needed; thus, a pretension-adjusting algorithm has been developed. A realistic spider web model was created using non-linear material properties to describe the mechanical behaviour of the spider silk and web pretension values seen in nature. Using this model, different scenarios with changing spider weight and web pretension were simulated using a numerical method based on Solid Mechanics. Our results show that this behaviour is likely related to the signal transmittance on the spider web. Spider web evolved to withstand damaging environmental factors such as wind and rainfall while preserving its functionality for trapping prey. Understanding spider web structure could lead us to improve engineering designs by implementing similar resiliency. This thesis presents a study investigating spider webs and a biomimicking application inspired by spider web structure. So, while the two areas are different in the sense that one is closer to biology while the other is to mechanical engineering, they serve the same purpose: understanding how this structure, spider orb-web, functions and how we can take ideas from it.
Open Access
Leveraging file significance in bus factor estimation
(2025-01) Haratian, Vahid
Software projects often face developer turnover for various reasons. Since develop-ers are key sources of knowledge in these projects, their absence inevitably leads to some degree of knowledge loss. The Bus Factor (BF) is a metric used to assess the impact of this knowledge loss on a project’s continuity. Traditionally, BF is defined as the smallest group of developers whose departure would result in a loss of more than half of the project’s knowledge. Current state-of-the-art methods calculate developers’ knowledge based on the number of files they have authored, using data from version control systems (VCS). However, numerous studies have highlighted that not all files in software projects hold the same level of significance. In this study, we investigate the impact of weighting files based on their significance on the performance of two widely used BF estimators. Significance scores are calculated using five established graph metrics derived from the project’s De-pendency Graph: PageRank, In-/Out-/All-Degree, and Betweenness Centralities. Additionally, we introduce BFSig, a prototype implementing our approach. Lastly, we present a new dataset featuring BF scores reported by software practitioners from five prominent GitHub repositories. Our findings show that BFSig surpasses the baseline methods, achieving up to an 18% reduction in Normalized Mean Absolute Error (NMAE). Additionally, BFSig reduces False Negatives by 18%when identifying potential risks linked to low BF. Furthermore, our respondents validated BFSig’s versatility, highlighting its capability to evaluate the BF of individual project subfolders. In conclusion, we believe that when estimating BF from authorship, software components of greater significance should be given higher weight.
Embargo
Elucidation of the role of IKK-related kinases in ulcerative colitis
(2025-01) Borucu, Ümran
Ulcerative colitis is an inflammatory disease of the colon resulting from the imbalance of the inflammatory mechanisms. Its onset and development depend on multiple factors, which are not yet fully uncovered. Understanding the factors involved in the pathophysiology of ulcerative colitis is important for the management of the disease. IKK-related kinases are involved in numerous pathways in cellular signaling. They play significant roles in the modulation of the inflammatory responses. In this study, we investigated the roles played by IKK-related kinases in ulcerative colitis. We generated DSS-induced colitis mouse models in acute and chronic phases. We treated these mice with a dual IKK-related kinase inhibitor, Amlexanox, to understand the role played by these kinases on the onset and development of the condition. Moreover, we performed biochemical analyses on the samples to gain molecular insights. We observed an increased inflammatory response both phenotypically and biochemically upon inhibition of IKK-related kinases.
Embargo
Do visual cortices in the blind activate to language processing demands?
(2024-12) Varol, Ayşe Betül
The fate of the occipital cortices in the blind has been of considerable interest in neuroscience to investigate whether and how much brain regions change with experience. Neuroimaging shows that the occipital cortices in the blind are activated during all kinds of non-visual tasks. In our previous study, we recently showed that all visual regions become MD regions and activate in response to any kind of control demand in the blind. Therefore, we hypothesized that the activation of the deprived visual cortex in language tasks in the literature is related to high cognitive control demand rather than language processing. This means that if V1 is involved in cognitive control demand, then it should not be activated in the passive listening task. Therefore, the study investigates whether the primary visual cortex (V1) in blind subjects participates in cognitive functions, particularly language processing, during a passive listening task with minimal cognitive control demand. Using fMRI, we compared whole brain and region of interest (ROI) activations in blind, and sighted participants across multiple areas, including occipital, language, and multiple demand (MD) regions. Results revealed no significant activation in V1 across all groups during the passive listening task, suggesting task-dependent limitations in the integration of V1 into higher-order cognitive processes rather than language. Furthermore, no significant differences in language or MD area activity were observed between blind and sighted participants. These findings challenge previous evidence on the occipital cortex becoming a language region in the blind and emphasize the importance of task complexity in modulating such interpretation. All of these suggest that blind occipital regions primarily become MD regions, and their activation during language tasks is limited to those that require domain-general cognitive control.
Embargo
Laser micro drilling of de Laval nozzles using a hybrid hexapod system
(2024-12) Bakan, Aykut
In recent years, miniaturization, one of the most intensely focused topics in science and technology, has enabled the use of macroscale engineered components at the microscale. Although components at this scale are extremely small, they can exhibit complex geometries. Therefore, precise manufacturing setups are required for their production or for processing existing ones. An example of such components is micro de Laval nozzles, which are used in the propulsion systems of micro-satellites. In this thesis, a novel method is presented for the production of micro de Laval nozzles, in addition to a method that enables drilling on freeform surfaces using a nanosecond laser and performing angled drillings on planar surfaces, all facilitated by a hybrid hexapod. The effects of parameters such as laser power, laser frequency, scan speed, number of loops, and rotation speed on the geometry and morphology of the nozzle were investigated for the method developed for micro de Laval nozzles. Experimental results demonstrated that critical geometric parameters for micro de Laval nozzles, such as throat diameter, outlet diameter, and outlet half angle, meet the requirements specified in the literature. The findings suggest that the developed method has the potential to provide a single-step, cost-effective solution for fabricating micro de Laval nozzles. In the future, efforts will focus on minimizing the sidewall roughness of the produced nozzles and automating the method for practical applications.
Open Access
Plastic degradation using genetically engineered microorganisms
(2025-01) Polat, Cem Dirse
The usage of PET plastics in daily life have excessively increased in the last decade. The increased usage of PET is accompanied with the massive amount of PET waste accumulating rapidly. Environmental pollution caused by this waste has reached a critical point with pollutants being found even in the most remote parts of the world. Causing massive damage to ecosystems and even human health, PET plastic waste needs to be handled urgently. Although there are ongoing PET recycling and treatment efforts, the current methods in use are insufficient. The techniques currently used are either costly, leave a significant carbon footprint or are lacking in their ability to recycle microplastics. However, with the discovery of microorganisms which have the ability of degrading PET, biodegradation of PET products has emerged as a promising green alternative. In this thesis we designed bacterial tools to utilize the PET hydrolyzing enzyme, PETase. For this purpose, living bacterial platforms were engineered. The first system employed E. coli as the host to display PETase on the cellular surface. With PETase molecules anchored on its surface, aiding in the stability and the activity of the enzyme, the system will be a useful tool for PET degradation. For the surface display system, the Ag43 autotransporter protein is used. The system was cloned, and expression was analyzed using immunocytochemistry labeling. The activity of the system was analyzed with chromatography and mass spectrometry. The second system proposed uses E. coli once again as a workhorse for PETase secretion, creating a simple yet effective tool for the bioremediation of PET. For secretion of the enzyme, the disruption of Braun’s lipoprotein to create a leaky outer membrane is exploited. The system was cloned, and the cloning was verified. Also, the activity of native PETase was analyzed with HPLC and mass spectrometry. With this analysis, the PET degrading activity of PETase was confirmed.
Embargo
Distributed waveguide design for reducing thermal load in semiconductor high power lasers
(2025-01) Saadi, Osama Aadil
Semiconductor lasers lead laser technology due to their high efficiency, compact size, and cost-effectiveness. Among these, GaAs-based laser diodes (LDs) are the most efficient light sources, but are still constrained by self-heating, which elevates internal temperatures and degrades performance, output power, and de-vice lifetime. Traditionally, increasing the cavity length has mitigated this issue by improving thermal conductivity, facilitated by advances in epitaxial growth, design, and device packaging. However, the cavity lengths of high-power GaAs LDs are now limited to approximately 5 mm, beyond which the output power declines because of intrinsic physical constraints. This work presents a new type of waveguide design, called distributed waveguide (DWG), that overcomes conventional cavity-length limitations. The DWG integrates lasing and secondary sections along the waveguide, which are electrically isolated to control current injection, yet optically connected for efficient beam transport. The laser section is electrically pumped to generate output, while the secondary section operates near-threshold to dissipate heat effectively. Extending the cavity length from 4 to 8 mm, DWG LDs exhibit significantly improved thermal management with favorable device characteristics. Experimental results, corroborated by numerical analysis, demonstrate that DWGs achieve approximately 1.8× lower junction temperature change while delivering high output power. Additionally, the DWG platform and its fabrication process are fully com-patible with standard semiconductor laser manufacturing techniques, ensuring industrial adoption. This work provides clear evidence that innovative waveguide designs can effectively mitigate self-heating, promising enhanced performance, output power, and reliability in semiconductor lasers.
Embargo
Interpretable holistic manipulation strategies in household environments for task and motion planning
(2025-01) Yenicesu, Arda Sarp
Interpretable Responsibility Sharing (IRS) introduces a novel heuristic for Task and Motion Planning (TAMP), leveraging holistic manipulation strategies to enhance planning efficiency and interpretability in household environments. By systematically incorporating auxiliary objects such as trays and pitchers—common in human-constructed spaces—IRS simplifies and optimizes task execution. The heuristic is based on the concept of Responsibility Sharing (RS), where auxiliary objects share task responsibilities with robotic agents, dividing complex tasks into manageable sub-problems. This division not only mirrors human usage patterns but also aids robots in navigating and manipulating within human-designed spaces more effectively. By integrating Optimized Rule Synthesis (ORS) for decision-making, IRS ensures that the use of auxiliary objects is both strategic and context-aware, enhancing the interpretability and effectiveness of robotic planning. Experiments across diverse household tasks, including serving, pouring, and handover, demonstrate that IRS significantly outperforms traditional methods, reducing effort in task execution and improving decision-making. This approach aligns with human-inspired strategies while offering a scalable framework adaptable to the dynamic complexities of household environments.