Altınbaş University Institutional Repository
Not a member yet
5805 research outputs found
Sort by
A graph neural network assisted reverse polymers engineering to design low bandgap benzothiophene polymers for light harvesting applications
No full textIn this study, we present a novel approach to reverse polymer engineering utilizing a Graph Neural Network (GNN) framework to design low bandgap benzothiophene (BT) polymers for light harvesting applications. We have curated an extensive dataset comprising 57,556 structure-property pairs of BT-based compounds, leveraging expert knowledge to enhance the quality and relevance of the data. Our Transformer-Assisted Oriented pretrained model for on-demand polymer generation (TAO) demonstrates exceptional performance, achieving a chemical validity rate of 99.27 % in top-1 generation mode across a test set of 6000 generated polymers, marking the highest success rate reported among polymer generative models to date. Throughout the training process, the loss steadily decreased with each epoch, indicating that the model was learning effectively from the data. The model predictive accuracy is further validated by an impressive average R2 value of 0.96 for 15 defined properties, highlighting the TAO with its robust capabilities in polymer design. The newly designed polymers exhibit a bandgap range of 1.5–3.40 eV, making them promising candidates for light harvesting applications. Additionally, their highest Synthetic Accessibility Likelihood Index (SALI) scores reach up to 17 and also indicates that the majority of these polymers are amenable to synthesis. This work not only advances the field of polymer design but also provides a powerful tool for the targeted development of materials with specific electronic properties.Funding agency : Taif University
Grant number : TU-DSPP-2024-7
Effective expressions for the estimation of the first and third natural periods of minarets based on extensive and comprehensive parametric modal analyses
No full textThis paper investigates the modal behaviour of minarets by conducting parametric modal analyses with a comprehensive set of geometric and material properties and critically proposes effective expressions to estimate the first and third natural periods of minarets. To make this unprecedented attempt, firstly, previously studied 52 minarets were examined to systematically select the geometric and material properties of 189 representative minarets. Subsequently, numerical models of these 189 minarets were created and assigned with the selected values for height, cross-sectional area, moment of inertia, modulus of elasticity, and mass density. Then, modal analysis was applied to each representative minaret to obtain modal characteristics to see the effect of each variable. Next, an iterative optimization algorithm was used to establish four effective expressions for the first and third natural periods, considering the effect of five and three variables. Finally, the performance of the proposed expressions was checked with 35 of the previously studied minarets whose dynamic characteristics are available. The results demonstrated a satisfactory level of accuracy. For example, the expression for the fundamental period with 5 variables presented discrepancies mostly far less than 19 % for 28 minarets of a total of 35 minarets. © 2025 Institution of Structural Engineer
Biological activities of Tanacetum sp. essential oils on Tribolium castaneum (Herbst, 1797) (Coleoptera: Tenebrionidae) and Sitophilus granarius (Linnaeus, 1758) (Coleoptera: Curculionidae)
No full textTribolium castaneum (Herbst, 1797) (Coleoptera: Tenebrionidae) and Sitophilus granarius (Linnaeus, 1758) (Coleoptera: Curculionidae) are among the most destructive stored-grain pests worldwide. This study investigated the chemical composition and biological activities of essential oils (EOs) obtained from Tanacetum parthenium and T. vulgare as eco-friendly control agents. GC-MS analysis identified camphor (39.24%) as the major constituent in T. parthenium and alpha-thujone (57.56%) in T. vulgare. In contact toxicity assays, T. parthenium EO caused 92.9% mortality in S. granarius, while T. vulgare achieved 53%. Both oils were ineffective against T. castaneum. Repellency tests showed strong, dose-dependent effects, with both oils exhibiting >= 90% repellency against T. castaneum for up to 12 h and >80% against S. granarius at 5% concentration. T. vulgare EO significantly inhibited F1 progeny production of S. granarius (82.7%), outperforming T. parthenium (63%). In vitro enzyme inhibition assays revealed that T. vulgare EO strongly suppressed AChE (87.1%) and BChE (69.5%) activities, whereas T. parthenium showed weaker inhibition. Molecular docking analyses demonstrated that alpha- and beta-thujone, the main components of T. vulgare, directly bind to the catalytic sites of both AChE and BChE enzymes, confirming their potent neurotoxicity. sPLS analysis determined that beta-thujone is directly responsible for death and enzyme inhibition, while the more abundant alpha-thujone is statistically associated with sublethal and behavioral effects such as deterrence and progeny production inhibition. Although T. parthenium exhibits high contact toxicity against S. granarius but shows weak cholinesterase inhibition, sPLS analysis statistically proved a strong synergistic interaction between camphor, the main component driving toxicity, and cis-chrysanthenyl acetate, a minor component. Consequently, these findings demonstrate that essential oils act through both lethal physiological disruption and behavioral responses, highlighting T. vulgare as a promising botanical insecticide and underscoring the need for future work on optimized synergistic formulations.Funding agency: Ministry of Agriculture & Forestry - Turkey. General Directorate of Agricultural Research & Policies (TAGEM) - Republic of Turkiye Ministry of Agriculture & Forestry.
Grant number: TAGEM/BSAD/E/18/A2/P4/36
Correction to : The Psychometric Properties of Autism Mental Status Examination (AMSE) in Turkish Sample
No full text..
Genetic and circulating biomarkers of cognitive dysfunction and dementia in CKD
No full textChronic kidney disease (CKD) is commonly accompanied by cognitive dysfunction and dementia, which, in turn, increase the risk of hospitalization, cardiovascular events and death. Over the last 30 years, only four studies focused on genetic markers of cognitive impairment in CKD and kidney failure (KF), indicating a significant gap in research. These studies suggest potential genetic predispositions to cognitive decline in CKD patients but also underscore the necessity for more comprehensive studies. Seventeen reports have established connections between cognitive function and kidney disease markers such as estimated glomerular filtration rate (eGFR), Cystatin C and albuminuria. A rapid eGFR decline has been associated with cognitive deterioration and vascular dementia, and mild to moderate eGFR reductions with diminished executive function in elderly men. Various biomarkers have been associated to Alzheimer’s disease or dementia in CKD and KF. These include amyloid beta and phosphorylated tau proteins, uremic toxins, gut microbiota, metabolic indicators, hypertension, endothelial dysfunction, vitamins and inflammation. However, the causal relevance of these associations remains unclear. Overall, the available evidence points to a complex interplay between the different biomarkers and cognitive health in CKD patients, underscoring the need for more research to elucidate these relationships.Funding agency: European Cooperation in Science and Technology (COST)
Grant number: COST ACTION CA1912
A rapid UV/Vis assisted designing of benzodithiophene based polymers by machine learning to predict their light absorption for photovoltaics
No full textAs global energy demands escalate, developing high-performance photovoltaic (PV) materials through accelerated design methodologies is imperative. A machine learning (ML) assisted predictive models are used to accelerate the design of benzodithiophene (BDT)-based polymers for their PV applications. The current approach leverages a curated dataset of 191 compounds with experimental UV–Vis spectra, mapped to molecular electronic descriptors via RDKit. Random Forest modeling yields a predictive framework (R2 = 0.98) for predicting their maximum absorption (λmax). After it, their 5000 new designs as novel polymers, identifying top performers with Synthetic Accessibility Likelihood Index scores up to 57, ensuring synthesis feasibility have also been designed. Feature importance analysis highlights MaxPartialCharge and Aromatic rings as crucial descriptors. The designed materials exhibit optimal energy gaps (1.35–2.0 eV), paving the way for efficient PV devices. The computed UV–Vis spectra of best predicted polymers are studied with their λmax range of 487–987 nm showing a significant redshift behavior. The designed polymers presents and good potential towards and they can be good candidates for organic solar cell applications.Funding agency: Taif Universit
Activation of Proteolysis During Oocyte In Vitro Maturation
No full textIn vitro maturation (IVM) is a form of assisted reproductive technology (ART) applied to obtain mature oocytes in culture. Decline in IVM success rates by age has led consideration of novel approaches based on cellular dynamics. Our aim was to achieve proteostasis in old bovine oocytes from 13 to 16-year-old bovine with a lower potential for fertilization. Lysosomal activation was achieved through increasing concentrations of proton pump activators PIP2 (0.1, 0.5, 1, and 5 μM), PMA (0.1, 1, 10, and 50 μM), and DOG (0.1, 1, 10, and 50 μM) at 6, 12, 18, and 24 h of IVM in old bovine oocytes. Morphological analysis was performed and IVM rates were determined. DQ-Red BSA was applied to live oocytes to determine proteolytic activation while lysosome density was determined by Lysotracker probe. Protein carbonylation was detected through oxyblot analysis. Polar body extrusion (PBE), through which a haploid nonfunctional polar body is released in the perivitelline space after completion of the first meiotic division, was observed in PIP2-0.1 μM, -0.5μM-6h; PIP2-5μM-12h; PMA-0.1μM-18h; PIP2-0.1μM, -0.5μM-24h groups. Oocyte diameter was the highest in DOG-1μM-6h, PMA-0.1μM-12h, PIP2-1μM-18h, and PIP2-0.5μM-24h groups. Morphological scores of oocytes were higher in young and old control groups. PIP2, PMA, and DOG affected oocyte quality positively after 6 h of IVM yielding in oocyte scores similar to the control group oocytes. However, they had a negative impact on the oocyte scores in longer periods of IVM, except for lower doses PMA (0.1 and 1 μM) at 12 h and PIP2 (0.5 μM) and PMA (0.1 μM) at 18 h, which were able to maintain the scores relatively closer to the control oocytes. Proteolytic activation was achieved in all groups at 6 h of culture. At all other time points PIP2 and PMA groups showed a better response to proteolytic activation. Lysosome density was increased in PIP2-5μM-6h; PIP2-0.1μM, -1μM-12h; PIP2-1μM, -5μM-18h as well as PMA-0.1μM-6h; PMA-1μM, -10μM-12h; PMA-1μM-18h; DOG-50μM-6h and DOG-0.1μM-12h. Protein carbonylation was the lowest in PIP2-0.1 μM groups at 12, 18, and 24 h. This study suggests that proton pump activators PIP2 and PMA was found to have a positive impact on IVM in terms of both morphological scores and proteolytic activation in a time and dose dependant manner.Turkiye Bilimsel ve Teknolojik Arastirma Kurumu (TUBITAK); 118S77
On the quest for solar energy harvesters and nonlinear optics: a DFT exploration of A-D-D-A framework with varying sp2 hybridization
No full textIn response to address the constraints of fullerene analogues, scientists are constantly working on developing low-cost fullerene-free functionalization for nanoscale organic photovoltaics. During the present study, the computational design and analysis of 14 new non-fullerene dyes (IDIC-O-1 to IDIC-O-14) centered on indacenodithiophene (IDIC) core are proposed with sp2-hybridized nitrogen at varying positions. Regarding their UV–visible assessment, several long-range and range-separated functionals like B3LYP, CAM-B3LYP, ωB97XD, and APFD using the 6-311G + (d,p) basis set have been employed to identify their optimal level of density functional theory (DFT) with an impressive correlation at the CAM-B3LYP level. Their global hardness (η) and global electrophilicity (ω) natures show their persistent nature. The energy gaps (Egaps) are lesser than IDIC and IDIC-O to imply an easier electronic transition. When contrasted to the IDIC-O, the findings indicate that its broad absorption spectrum had a redshift. The efficient HOMO → LUMO-based CT was investigated, and an open-circuit voltage (Voc) study is done on HOMOIDIC → LUMOAcceptor. All dyes have their Voc values lower than reference (IDIC-O) except IDIC-O-11 with a positive value. These lower reorganization energies (RE) for holes and electrons indicate a greater charge transfer (CT). When contrasted to the IDIC-O, the newly designed dyes have better characteristics for solar cell performance
Optimizing β-carotene bioaccessibility predictions with advanced machine learning algorithms and feature selection strategies
No full textBioaccessibility is an important issue in designing functional food products, since in vitro retention of the bioactive compounds in the formulations has a potent impact on the health benefits in the target tissue. It was reported that artificial intelligence is of great attention for the optimizing the process parameters in food science and technology to produce affordable products with greater nutritional, sensorial, technological and functional characteristics. Hence, it is required to unveil how food processing variables affect the in vitro gastrointestinal digestion behavior of nutrients and optimize them mathematically. In the current study, it was aimed to utilize machine learning methods including TreeBoost (TB), Multilayer Neural Network (MLP) and Support Vector Machine (SVM) to predict the bioaccessibility of /3-carotene, one of the bioactive compounds, in the emulsion-based food matrix. The emulsion type (micro-or nano-), oil/water phase ratio, oil type, the type of the emulsifier (protein or carbohydrate), and /3-carotene concentration were selected as variables for the predicted models. Results demonstrated that TB-based model provides the best prediction for the bioaccessibility of /3-carotene with the values of R2 = 0.4325, RMSE = 17.2484, NMSE = 0.5675, and MAE = 13.3809. Besides, according to TB-based Model 7, the emulsion type (micro-or nano-), oil/water phase ratio, the oil type, the type of the protein, and /3-carotene concentration were found to be effective on the bioaccessibility of /3-carotene. Based on the comparison of method performance, The RMSE value of TB-based Model 7, showed an improvement of 9.27 % compared to the SVM method and 13.4 % compared to the MLP method. In this study, the importance of variables for the emulsion prescriptions and the predictive capability of existing machine learning tools in the area of /3-carotene release from that matrix were introduced. Considering the outcomes of the present study, it can be concluded that they will shed light to future experiments for the practical implementation of these models in food-related systems. Thereby, it could be possible to change the way for the design of delivery systems for food bioactives from time-consuming and costly trial-based traditional approaches to well-developed models, which can simulate the physical state of the process and results in reduced research and development time as well as financial investment and uniform product
A Novel Architecture Based on Weight Freezing and Random Forest for Website Phishing Detection
Get PDFPhishing is a severe cybersecurity threat that continues to cause significant economic losses and breaches of privacy globally. This paper presents a novel architecture for effective phishing website detection, integrating a unique weight freezing technique in the Random Forest algorithm with a feature pyramid scheme for feature selection. This innovative combination enables the model to focus on important features and discard irrelevant or redundant ones, thus improving detection accuracy while preventing overfitting. Using a dataset of 2456 instances with 30 attributes, the proposed model achieved an accuracy of 97.4%, an MSE of 0.05, and an F1 score of 96.2%, outperforming previous state-of-the-art methods. The promising results suggest the model's significant potential in the cybersecurity field, providing a robust tool against the escalating threat of phishing