AKAli Onur Kaya

Ali Onur Kaya

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Ali Onur Kaya has a total of 11 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Development of an Interpretable QSAR Model for Predicting Coagulation Factor XIIa Inhibitors Using Ensemble Machine Learning

    by Ali Onur Kaya, Mert Can Emre

    Background/Objective: Activated coagulation factor XII (FXIIa) is a component of the contact activation pathway and a pharmacologically relevant target in contact-system-associated processes. In this study, scaffold-aware and interpretable machine-learning QSAR models were developed for human FXIIa activity. Methods: Bioactivity records for the human single protein target CHEMBL2821 were retrieved from ChEMBL release 37. Modeling was restricted to exact IC50 measurements from assays explicitly referring to FXIIa, Factor XIIa, or activated Factor XII. Median-consolidated pIC50 values and two-dimensional Mordred descriptors were evaluated using leakage-safe preprocessing, scaffold-disjoint validation, Y-randomization, applicability domain analysis, structural similarity auditing, and SHAP interpretation. Results: The regression dataset comprised 424 compounds and 166 Bemis–Murcko scaffolds in this study. The Gradient Boosting regressor achieved R2 = 0.7560, RMSE = 0.6924, and MAE = 0.4965 on the locked scaffold-disjoint test set (n = 85); across 50 repeated scaffold partitions, the mean R2 was 0.6892 ± 0.1515. The classification model achieved ROC-AUC = 0.9453, PR-AUC = 0.9807, balanced accuracy = 0.7561, and MCC = 0.5972 (n = 73). Y-randomization supported nonrandom predictive signals (empirical p = 0.0099). Conclusions: The models support computational prioritization within the represented FXIIa chemical domain, while prospective evaluation of independently generated compounds remains necessary.

  • Open Access

    An Explainable Machine Learning-Based QSAR Framework for Predicting Thrombin Inhibitory Activity

    by Ali Onur Kaya, Mert Can Emre

    Background/Objectives: Public thrombin bioactivity records contain heterogeneous endpoints, replicate measurements, related chemical series, and potentially reactive compounds that may bias quantitative structure–activity relationship models. In this study, we developed an explainable, assay-aware, and leakage-safe machine learning framework for predicting thrombin-inhibitory activity. Methods: Exact Ki and IC50 records for human thrombin (CHEMBL204) were standardized, converted to pActivity, and aggregated using predefined criteria. The final dataset comprised 5189 unique compounds represented by development-filtered Mordred descriptors and Morgan fingerprint. The models were optimized using development-only out-of-fold validation and evaluated using random and scaffold-disjoint held-out tests. Results: In the random-split analysis, ConsensusAll achieved an out-of-fold R2 of 0.7588 and a held-out test R2 of 0.7480, with an RMSE of 0.7360, MAE of 0.5307, and concordance correlation coefficient of 0.8540. In the scaffold-disjoint locked test, ConsensusTop5 achieved R2 = 0.5831, RMSE = 0.9484, and MAE = 0.7290, respectively. The applicability domain covered 92.68% of the locked test compounds and yielded R2 = 0.6041. One hundred Y-randomization runs produced a mean R2 of −0.1233 (empirical p = 0.0099). Conclusions: The framework provides useful predictions within the represented chemical space and measurable generalization for unseen scaffolds. This supports compound prioritization, although prospective biochemical validation remains necessary.

  • Open Access

    Machine Learning-Driven QSAR Modeling of FXIa Inhibitors for Virtual Screening and Rational Drug Design

    by Ali Onur Kaya, Mert Can Emre, Nesrin Emre

    Background/Objectives: Coagulation factor XIa (FXIa) has emerged as a promising therapeutic target for the development of safer anticoagulant therapies with reduced bleeding risk. This study aimed to develop an interpretable machine learning-driven quantitative structure–activity relationship (QSAR) framework for predicting the inhibitory activity of FXIa inhibitors and supporting virtual screening applications. Methods: A total of 3026 curated compounds retrieved from the ChEMBL database were used for regression modeling, whereas 2119 compounds were retained for classification modeling after excluding intermediate-activity molecules. Molecular descriptors were generated using RDKit, Mordred, and Morgan fingerprint representations. Following preprocessing and feature selection, multiple machine learning algorithms were systematically benchmarked. Model robustness and reliability were further evaluated using 5-fold cross-validation, scaffold-aware validation, applicability domain analysis, and Y-randomization testing. Results: Nonlinear ensemble learning approaches consistently outperformed conventional linear algorithms. The optimized HistGradientBoostingRegressor achieved the best regression performance, with an independent test-set R2 value of 0.711 and an RMSE value of 0.759, whereas the optimized classification model achieved accuracies approaching 95%. SHAP analysis identified lipophilicity-related descriptors, aromatic scaffold organization, electrostatic surface properties, and molecular topology as major contributors to FXIa inhibitory activity prediction. In addition, a proof-of-concept virtual screening workflow successfully identified several candidate compounds exhibiting high predicted pKi values and elevated active-class probabilities. Conclusions: The proposed framework provides a robust, interpretable, and reproducible machine learning-driven QSAR strategy for FXIa inhibitor discovery and may facilitate future virtual screening campaigns and medicinal chemistry optimization studies targeting FXIa-associated anticoagulant drug discovery.


  • Top Cited Papers
    Open Access

    Interpretable machine learning-driven QSAR modeling for coagulation factor X inhibitors: from molecular descriptors to predictive potency

    by Ali Onur Kaya

    Inhibition of Coagulation Factor X (FXa) is a clinically validated therapeutic strategy; however, developing safer and more selective inhibitors remains a major challenge. In this study, we developed an interpretable machine learning–based QSAR framework to predict both the inhibitory potency and activity class of small molecules targeting FXa. A structurally curated dataset of 6400 compounds was retrieved from ChEMBL, standardized, and encoded using 391 non-redundant Mordred descriptors following systematic filtering. Benchmarking of 42 regression and 42 classification algorithms identified ExtraTreesRegressor and XGBoostClassifier as the most robust models. The regression model achieved an R 2 of 0.760 and an RMSE of 0.831 on the independent test set, while the classification model reached an accuracy of 0.91 with balanced precision, recall, and an ROC-AUC of 0.962. SHAP (SHapley Additive exPlanations) analysis further enhanced interpretability by revealing that electrostatic, topological, and polar surface descriptors were the dominant contributors to FXa inhibitory potency. Applicability domain assessment using Williams plots confirmed that most compounds in both the training and test sets lay within the model’s reliable prediction space. Overall, the proposed QSAR pipeline integrates strong predictive performance with valuable mechanistic interpretability and rigorous validation, offering a practical computational tool for the virtual screening and rational design of novel FXa inhibitors.


  • Open Access

    Interpretable Machine Learning-Driven QSAR Modeling for Coagulation Factor X Inhibitors: From Molecular Descriptors to Predictive Potency

    by Ali Onur Kaya

    The inhibition of Coagulation Factor X (FXa) is a clinically validated strategy in anticoagulant therapy; however, the development of safer and more selective inhibitors remains a critical challenge. In this study, we present a machine learning–enhanced quantitative structure–activity relationship (QSAR) modeling framework to predict the inhibitory potency (pKi) of small molecules targeting FXa. Bioactivity data were curated from the ChEMBL database and standardized, resulting in a filtered dataset of 6400 structurally validated compounds. The molecular descriptors were calculated using the Mordred platform and filtered for statistical robustness. Two predictive approaches were employed: regression using the ExtraTrees Regressor and binary classification using the XGBoost Classifier. The regression model achieved an R² of 0.760 and an RMSE of 0.831 on the test set. The classification model demonstrated strong performance across all key metrics, achieving an accuracy of 0.91, precision of 0.92 (class 0) and 0.89 (class 1), recall of 0.89 (class 0) and 0.92 (class 1 ) , and an F1-score of 0.91 for both classes. These results indicate a balanced and robust predictive capability across active and inactive compounds. SHAP (SHapley Additive exPlanations) analysis enabled the interpretation of key structural features driving activity, revealing that electrostatic and topological descriptors were the most dominant. The applicability domain analysis was conducted using the leverage approach, and the Williams plots indicated that all compounds in both the training and test sets fell within the reliable prediction space of the regression model. We are confident that the models developed in this study provide not only strong predictive performance but also interpretable insights and can be effectively used to guide the rational design and screening of novel FXa inhibitors in anticoagulant drug discovery.

  • Comparative microwave absorption of dual-layer coatings containing CoFe2O4, NiFe2O4, and manganese-based iron oxide Mn0.176Fe1.824O3 nanoparticles

    by Ali Onur Kaya, Mert Can Emre, Kayhan Ateş, Ferhat Aslan, Ahmet Tumbul, Şükrü Özen, İbRahim Halil Mutlu

    Electromagnetic interference (EMI) shielding effectiveness (SE) and radar-absorbing materials (RAMs) are increasingly vital in aerospace, defense, and telecommunications. In this study, cobalt ferrite (CoFe2O4-CFO), nickel ferrite (NiFe2O4-NFO), and manganese-based iron oxide (Mn0.176Fe1.824O3-MFO) nanoparticles were synthesized via a hydrothermal method and incorporated, along with the dual-layer paint formulations, to investigate their EMI shielding performance. The coatings were applied to lightweight fiberglass substrates pre-treated with a primer layer and tested across the 3.2–5.0 GHz frequency range. X-ray diffraction (XRD) analysis confirmed the formation of spinel structures for CFO and NFO, whereas the Mn-based sample crystallized into a rhombohedral Mn–Fe–O-phase. Scanning electron microscopy (SEM) revealed distinct morphological features, including differences in agglomeration among the samples. Vibrating sample magnetometry (VSM) revealed that MFO exhibited the highest coercivity (HC) and squareness ratio, indicating a stronger magnetic anisotropy. The shielding effectiveness was primarily dominated by absorption effects. Among the samples, MFO at 10 wt% exhibited the highest total SE (∼6–7 dB), whereas CFO at 5 wt% showed reasonable performance (∼3–5 dB) with minimal material usage. The results demonstrate that material selection and filler concentration are critical for optimizing EMI shielding effectiveness. These findings contribute to the development of scalable, lightweight, and cost-effective ferrite-based coatings for targeted frequency applications.


  • Open Access

    Microwave Absorption Properties of Paint Containing CoFe₂O₄ Nanoparticles and Wood Shaving Additive

    by Ali Onur Kaya, Mert Can Emre, İbRahim Halil Mutlu

    The demand for effective electromagnetic interference (EMI) shielding has driven the development of advanced materials combining magnetic and dielectric properties. This study presents a composite paint with cobalt ferrite (CoFe₂O₄) nanoparticles (5 wt%) and wood shavings (10 wt%) for enhanced EMI shielding in the 3.2–5.0 GHz range. Wood shavings improved electromagnetic wave absorption through their porous structure and dielectric properties, complementing CoFe₂O₄’s magnetic loss characteristics. The paint achieved a peak absorption shielding effectiveness (SE) of 12 dB, surpassing the 10 dB of the CoFe₂O₄-only system. Reflection SE values were reduced to -1 to +1 dB. This eco-friendly and cost-effective material offers significant potential for industrial and defense applications.


  • Open Access

    Investigation effect of pulsed magnetic fields on pepper (Capsicum annuum L.) plant growth

    by Ali Onur Kaya, Mert Can Emre, Ozer Calıs, İlhami Tozlu, İbrahim Halil Mutlu

    Pepper (Capsicum annuum L.) plants are extensively utilized in culinary and spice industries, rendering their cultivation pivotal in agricultural production. Enhancing their growth and yield is a critical research area for producers and horticulturists. Recent investigations have delved into the use of pulsed magnetic fields (PMFs) as a potential growth stimulant. Unlike static magnetic fields, PMFs are characterized by transient, high-intensity magnetic bursts, potentially eliciting varied responses in plants. To assess PMFs' impact on pepper plants, several experiments were setup comprising two solenoids, each wound around an 18-cm-diameter rigid plastic pipe but with differing coil turns, one with 40 and the other with 80 turns. These solenoids were utilized to generate PMFs at a frequency of 1 kHz with two intensities: 17 micro-Tesla (µT) and 34 µT. The pepper plants were situated within the PMF zone under controlled conditions, ensuring consistency in light, temperature, and moisture levels. The experimental design included three plant groups: a control group with no PMFs exposure except that of the Earth’s magnetic field, and two groups subjected to 17 µT and 34 µT PMFs intensities with Earth’s magnetic field, ranging between 25-65 µT. The treatment spanned 15 days, involving 6 hours of daily continuous exposure. Key growth indicators such as plant height, stem diameter, leaf area, and fresh and dry weights of both shoot and root systems were measured and analyzed. This analysis revealed significant increases in plant height, leaf area, and fresh and dry weights of the shoot, but not in root systems. Further research is warranted to deepen the understanding of PMFs' effects on pepper plants.