Search in Publications
Found 6374 publications
Real-world performance analysis of a universal computational reasoning model for precision oncology in lung cancer
Dóra Mathiász
Christophe Le Tourneau
Dóra Tihanyi
Ákos Boldizsár
Júlia Déri
Maud Kamal
Dóra Lakatos
István Vályi-Nagy
László Urbán
Barbara Vodicska
Dóra Kormos
Anna Dirner
Róbert Dóczi
Viktor Kardos
Réka Szalkai-Dénes
Ákos Takács
Mária Kocsis-Steinbach
Gábor György Kalmár
Márton Bolyácz
Viktor Kardos
Christian Rolfo
Gábor Pajkos
István Peták
Richárd Schwáb
Edit Várkondi
Arkadiusz Z. Dudek
Publication Name: Npj Precision Oncology
Publication Date: 2025-12-01
Volume: 9
Issue: 1
Page Range: Unknown
Description:
Tumors harbor multiple genetic alterations, yet treatment decisions are commonly based on single biomarkers, leading to underutilization of genomic information by comprehensive molecular tests, uncertainty in clinical practice, and frequent treatment failures. Although molecular tumor boards can assist personalized treatments, this process is not scalable or standardized, resulting in highly discordant recommendations. Validated digital solutions for personalized decision support are highly needed. The Digital Drug Assignment (DDA) system is a computational reasoning model that scores treatment options based on the full tumor genomic data. We retrospectively analyzed data of 111 lung cancer patients and found that high-score MTAs (1000≦DDA score) provided significant clinical benefit over other treatments, in terms of ORR, PFS, and OS. These results demonstrate that the DDA system is predictive of relative benefit of the various agents used in lung cancer care. Digital drug assignment can potentially address challenges with complex molecular profiles in routine clinical settings.
Open Access: Yes
Flower fertilization optimization algorithm with application to adaptive controllers
Publication Name: Scientific Reports
Publication Date: 2025-12-01
Volume: 15
Issue: 1
Page Range: Unknown
Description:
This article presents the Flower Fertilization Optimization Algorithm (FFO), a novel bio-inspired optimization technique inspired by the natural fertilization process of flowering plants. The FFO emulates the behavior of pollen grains navigating through the search space to fertilize ovules, effectively balancing exploration and exploitation mechanisms. The developed FFO is theoretically introduced through the article and rigorously evaluated on a diverse set of 32 benchmark optimization problems, encompassing unimodal, multimodal, and fixed-dimension functions. The algorithm consistently outperformed 14 state-of-the-art metaheuristic algorithms, demonstrating superior accuracy, convergence speed, and robustness across all test cases. Also, exploitation, exploration, and parameter sensitivity analyses were performed to have a comprehensive understanding of the new algorithm. Additionally, FFO was applied to optimize the parameters of a Proportional-Integral-Derivative (PID) controller for magnetic train positioning—a complex and nonlinear control challenge. The FFO efficiently fine-tuned the PID gains, enhancing system stability, precise positioning, and improved response times. The successful implementation underscores the algorithm’s versatility and effectiveness in handling real-world engineering problems. The positive outcomes from extensive benchmarking and practical application show the FFO’s potential as a powerful optimization tool. In applying multi-objective PID controller parameter optimization, FFO demonstrated superior performance with a sum of mean errors of 190.563, outperforming particle swarm optimization (250.075) and dynamic differential annealed optimization (219.629). These results indicate FFO’s ability to achieve precise and reliable PID tuning for control systems. Furthermore, FFO achieved competitive results on large-scale optimization problems, demonstrating its scalability and robustness.
Open Access: Yes
ESG disclosure topics and reporting frameworks: exploratory research across automotive, construction, and energy industries
Publication Name: Discover Sustainability
Publication Date: 2025-12-01
Volume: 6
Issue: 1
Page Range: Unknown
Description:
Environmental, Social, and Governance (ESG) reporting and proper measurement of greenhouse gas emissions are becoming increasingly important for industries with substantial environmental impact. This research aims to assess the current state of ESG reporting practices and highlight areas for improvement across the automotive, construction and energy industries operating in the Central Eastern European (CEE) region. To achieve this aim, a multi-industry sustainability disclosure database was created and analyzed through a Python-based text-mining methodology, using term frequency-inverse document frequency and keyword-in-context analysis. The process involved extracting and preprocessing text from 60 sustainability reports for the year 2021, followed by constructing a custom dictionary of key ESG terms aligned with the European Sustainability Reporting Standards. The findings reveal considerable variance in the focus of qualitative disclosures across industries, particularly regarding climate change and biodiversity. The investigation underscores the need for enhanced transparency, consistent metrics, and rigorous validation in ESG reporting. The study also provides new insights into the technical possibilities of automated text analysis for sustainability reporting in the CEE region, and highlights key areas where improvement appears necessary.
Open Access: Yes
Scenario-driven decision models for rare element waste management by integrating koch snowflake fuzzy sets and euclidean expert weighting
Publication Name: Sustainable Futures
Publication Date: 2025-12-01
Volume: 10
Issue: Unknown
Page Range: Unknown
Description:
The most critical factors must be determined to effectively manage environmental wastes generated during the extraction of rare elements. Otherwise, businesses may not be able to effectively manage their limited financial and human resources. This situation negatively affects the financial performance of the projects. The limited number of existing studies in the literature causes environmental risks to be insufficiently managed and recycling processes to be unoptimized. This study aims to determine priority strategies to increase the effectiveness of rare element waste management processes. Comprehensive and original decision-making models are created under three different scenarios. Koch Snowflake fuzzy sets, Euclidean based expert weighting and cognitive information modelling and analysis system (CIMAS) approaches are integrated in this model. The main contribution of this study is that a new type of fuzzy numbers called Koch Snowflake fuzzy sets is developed by considering the concept of fractal numbers. Fractal geometry is a powerful tool for modelling complex and dynamic systems. Hence, these new sets provide more flexible and more detailed uncertainty modelling. Moreover, considering different scenarios dynamic strategies can be developed that can adapt to changing conditions, such as pandemics or trade wars. The findings denote that technological developments are determined as the most critical factor under normal conditions. In the scenario where trade wars occur, it is revealed that political and regulatory measures should be addressed as a priority. In the event of a new epidemic disease such as COVID-19, it is concluded that more importance should be given to long-term storage strategies.
Open Access: Yes
Biological and therapeutic implications of sex hormone-related gene clustering in testicular cancer
Publication Name: Basic and Clinical Andrology
Publication Date: 2025-12-01
Volume: 35
Issue: 1
Page Range: Unknown
Description:
Background: Gonadotropin dysregulation seems to play a potential role in the carcinogenesis of testicular germ cell tumor (TGCT). The aim of this study was to explore the expression of specific genes related to sex hormone regulation, synthesis, and metabolism in TGCT and to define specific hormonal clusters. Two publicly available databases were used for this analysis (TCGA and GSE99420). By means of hard-threshold regularized KMEANS clustering, we assigned TGCT samples into four clusters defined in respect to different expression of the sex hormone-related genes. We analysed clinical data, protein and gene expression, signaling regarding hormonal clusters. Based on whole-transcriptome gene expression, prediction of anti-cancer drug response was made by RIDGE models. Results: Cluster #1 (12–16%) consisted primarily of non-seminomatous germ cell tumor (NSGCT), characterized by high expression of PRL, GNRH1, HSD17B2 and SRD5A1. Cluster #2 (42–50%) included predominantly seminomas with high expression of SRD5A3, being highly infiltrated by T and B cells. Cluster #3 (8.3–18%) comprised of NSGCT with high expression of CGA, CYP19A1, HSD17B12, HSD17B1, SHBG. Cluster #4 (23–30%), which consisted primarily of NSGCT with a small fraction of seminomas, was outlined by increased expression of STAR, POMC, CYP11A1, CYP17A1, HSD3B2 and HSD17B3. Elevated fibroblast levels and increased extracellular matrix- and growth factor signaling-related gene signature scores were described in cluster #1 and #3. In the combined model of progression-free survival, S2/S3 tumor marker status, hormonal cluster #1 or #3 and teratoma histology, were independently associated with 25–30% increase of progression risk. Based on the increased receptor tyrosine kinase and growth factor signaling, cluster #1, #3 and #4 were predicted to be sensitive to tyrosine kinase inhibitors, FGFR inhibitors or EGFR/ERBB inhibitors. Cluster #2 and #4 were responsive to compounds interfering with DNA synthesis, cytoskeleton, cell cycle and epigenetics. Response to apoptosis modulators was predicted only for cluster #2. Conclusions: Hormonal cluster #1 or #3 is an independent prognostic factor regarding poor progression-free survival. Hormonal cluster assignment also affects the predicted drug response with cluster-dependent susceptibility to specific novel therapeutic compounds.
Open Access: Yes
Pathways to asbestos-free and sustainable cities using multi-level perspective approach
Publication Name: Discover Sustainability
Publication Date: 2025-12-01
Volume: 6
Issue: 1
Page Range: Unknown
Description:
Urban policymakers, researchers, and municipal planners increasingly face the challenge of managing complex sustainability transitions, particularly in contexts involving persistent environmental hazards such as asbestos contamination. This systematic review applies the Multi-Level Perspective (MLP), which examines interactions between niche innovations, socio-technical regimes, and broader landscapes, to the underexplored area of asbestos-free urban transitions. The concept of an “asbestos-free city” is introduced in this paper as a novel analytical lens to describe urban transitions aiming to eliminate asbestos-related risks through systemic, sustainable interventions. The review was conducted through a structured qualitative analysis of peer-reviewed academic literature, guided by predefined thematic criteria and relevance to urban asbestos-related transitions. The review highlights the factors that enable or hinder the adoption of asbestos-free and strong sustainable solutions, as well as the role of various actors, such as policymakers, industry, and civil society, in driving these transitions. Despite the growing body of work on sustainability transitions, the integration of MLP into asbestos-related urban transformation remains limited. This paper fills that gap by offering a structured synthesis and proposing a roadmap for future research and practice. Our findings provide actionable insights for actors across policy, civil society, and industry seeking to accelerate transitions toward asbestos-free and sustainable cities.
Open Access: Yes
Initial Validation of the Hungarian Version of Abridged Nutrition for Sport Knowledge Questionnaire (ANSKQ-HU)
Publication Name: Sports
Publication Date: 2025-12-01
Volume: 13
Issue: 12
Page Range: Unknown
Description:
Nutrition knowledge is essential for optimizing performance, recovery, and overall health in athletes. This study aimed to (1) adapt and validate the Hungarian version of the ANSKQ (Trakman et al., 2017) (ANSKQ-HU) and (2) assess the nutrition knowledge of Hungarian elite and recreational athletes. Following standard translation procedures and expert review, face validity was established. Data were collected from 1.335 athletes, and item difficulty, exploratory factor analysis (EFA), and reliability analyses were performed. A three-factor structure emerged: (1) Fundamentals of nutrition, energy needs, and prohibited substances; (2) Micronutrients and performance-enhancing sports nutrition; and (3) Utilization of macronutrients. While Cronbach’s alpha values were low (α = 0.41–0.62), this seemed acceptable given the dichotomous nature of the questionnaire. Most participants scored poorly (63.3%), with the lowest results in the micronutrients and performance-enhancing nutrition factor. Only 6.9% had formal nutrition education and most frequently respondents received help from coaches, family members, and friends. These findings highlight a significant gap in sports nutrition knowledge among Hungarian athletes and support the need for educative activities organized by sport nutrition professionals. The ANSKQ-HU is a reliable and valid tool for assessing nutrition knowledge in Hungarian athletes and can be a useful questionnaire for their support team (nutritionists, physicians).
Open Access: Yes
Numerical investigation of bonding in stone-clad Façades: comparative analysis with and without mechanical anchorage
Publication Name: Journal of Infrastructure Preservation and Resilience
Publication Date: 2025-12-01
Volume: 6
Issue: 1
Page Range: Unknown
Description:
Reliable simulation of bond behavior between stone façade panels and concrete substrates is crucial for safe façade design, particularly with mechanical anchorage. Conventional finite element models relying on tie constraints overestimate interface strength, especially in the absence of surface preparation or bonding agents. This study develops and validates a physically motivated, element deletion–based finite element methodology to accurately simulate crack initiation, propagation, and failure at the mortar–stone interface. The three-dimensional numerical models, implemented in ABAQUS and benchmarked against laboratory splitting shear tests, represent the composite system comprising a concrete substrate, sand-cement adhesive mortar, and a Travertine stone façade. Both unanchored and Z-type mechanically anchored configurations were examined. Results demonstrate the approach yields accurate predictions of failure loads and damage evolution: for unanchored specimens, the maximum numerical–experimental deviation was below 2%, while Z-type clip anchorage significantly enhanced the load-bearing capacity and altered the fracture mechanism. Compared to conventional tie or interface-layer models, the element deletion strategy provides a computationally efficient and transparent tool for capturing the failure behavior of stone–mortar–concrete composites. The findings offer insights for optimizing façade anchorage design and provide a validated numerical framework for future research.
Open Access: Yes
Enhancing Policy Insights: Machine Learning-Based Forecasting of Euro Area Inflation HICP and Subcomponents
Publication Name: Forecasting
Publication Date: 2025-12-01
Volume: 7
Issue: 4
Page Range: Unknown
Description:
Accurate inflation forecasting is of central importance for monetary authorities, governments, and businesses, as it shapes economic decisions and policy responses. While most studies focus on headline inflation, this paper analyses the Harmonised Index of Consumer Prices (HICP) and its 12 subcomponents in the euro area over the period 2000–2023, covering episodes of financial crisis, economic stability, and recent inflationary shocks. We apply a broad set of machine learning and deep learning models, systematically optimized through grid search, and evaluate their performance using the Normalized Mean Absolute Error (NMAE). To complement traditional accuracy measures, we introduce the Forecastability Index (FI) and the Interquartile Range (IQR), which jointly capture both the difficulty and robustness of forecasts. Our results show that RNN and LSTM architectures consistently outperform traditional approaches such as SVR and RFR, particularly in volatile environments. Subcomponents such as Health and Education proved easier to forecast, while Recreation and culture and Restaurants and hotels were among the most challenging. The findings demonstrate that macroeconomic stability enhances forecasting accuracy, whereas crises amplify errors and inter-model dispersion. By highlighting the heterogeneous predictability of inflation subcomponents, this study provides novel insights with strong policy relevance, showing which categories can be forecast with greater confidence and where uncertainty requires more cautious intervention.
Open Access: Yes
