Shota Rustaveli State University

Gastric cancer (GC) is a leading cause of cancer mortality due to late diagnosis, metastasis, and therapy resistance. Epigenetic modifications rewire the cytokine network, facilitating the development, progression, and chemoresistance of GC. Cytokines regulate immune and inflammatory responses through various pathways, including the IL-6/JAK/STAT pathway, MAP Kinase, NF-κB signaling, IL-8/NF-κB, and the TGF-β axis, which sustain persistent inflammation and oncogenic signaling. The expression of these regulatory molecules is tightly controlled by epigenetic changes, including DNA methylation, histone modification, chromatin remodeling, and miRNA sponging. In addition, this is influenced by long non-coding RNAs (lncRNAs) and circular RNAs (circRNAs), which impact the tumor microenvironment and tumor survival. H. pylori and EBV further remodel cytokine epigenetic networks, promoting tumor progression. LncRNA and circRNAs can play dual roles, acting as both oncogenic and tumor-suppressive agents, where oncogenic ncRNAs, via drug efflux, apoptosis inhibition, autophagy, PD-L1 stabilization, and CSC maintenance, whereas tumor-suppressive ncRNAs restore chemosensitivity and antitumor immunity by reactivating PTEN, MHC-I, and apoptosis, proving to be promising biomarkers and therapeutic targets. Various clinical trials are also discussed in this review.

The paper is devoted to an empirical analysis of the relationship between the digital economy and innovative competitiveness, based on the case of the Black Sea region countries. The relevance of the study is обусловed by the fact that, in the context of globalization, the formation of economic competitiveness is increasingly grounded in digital transformation, technological progress, and the development of innovative potential. This issue is of particular importance in the Black Sea region, where economies with differing levels of development coexist and pronounced structural asymmetries are evident, creating additional challenges for both stimulating economic growth and fostering the formation of innovative ecosystems. In this context, understanding the role of the digital economy as a key determinant of countries’ positions in the global competitive landscape is especially significant. The aim of the study is to analyze the impact of digital economy development on the formation of innovative competitiveness. Within this objective, the research focuses on assessing how the intensity of digital technology use, the quality of digital infrastructure, and the level of digital skills influence innovation activity and competitiveness. In addition, particular attention is paid to the interaction among these factors and the potential synergistic effects arising from their combined development. The study also considers the role of the institutional environment and the education system in strengthening the digital economy, which ultimately contributes to enhancing innovation capacity and ensuring economic resilience. Methodologically, the research is based on a quantitative approach, ensuring the objectivity and generalizability of the findings. Data were collected through a survey of 100 respondents using an online questionnaire platform. The sample included representatives from diverse professional and social groups, allowing for a more comprehensive reflection of the multifaceted impact of the digital economy. The collected data were analyzed using descriptive and correlational statistical methods, enabling the identification of the strength and direction of relationships between variables. Additionally, elements of comparative analysis were employed to better assess regional differences and emerging trends. The empirical results indicate significant positive relationships among the key components of the digital economy. In particular, digital skills demonstrate a strong influence on both professional and economic competitiveness, confirming that human capital is one of the most critical resources in the modern economy. Furthermore, the intensive use of digital technologies significantly stimulates the growth of innovative activity, enhances productivity, and supports the development of new business models. The findings also reveal that access to digital infrastructure is an important precondition for economic development opportunities; however, it is not the sole determining factor, as its effective utilization depends on both the institutional environment and the digital competencies of the population. The results of the study confirm that the digital economy constitutes one of the fundamental bases for the formation of innovative competitiveness. At the same time, its effective development is determined by a multifactorial environment, including the quality of human capital, the effectiveness of the education system, and the robustness of institutional arrangements. The paper also highlights the importance of cooperation between the public and private sectors, the support of innovation, and the implementation of appropriate policies in the process of digital transformation. The practical significance of the study lies in the fact that its findings can be used in the development of economic and innovation policies, particularly for the countries of the Black Sea region, where digital transformation remains in an active phase of development and requires targeted and coordinated approaches.

Objective: Amyotrophic lateral sclerosis (ALS) is a progressive and fatal neurodegenerative disease that generates significant fear in both patients and the general population. In recent years, widespread access to artificial intelligence (AI)–driven health information tools—such as symptom checkers, large language models, and automated risk interpretation platforms—has transformed how individuals seek medical knowledge. While these tools offer educational benefits, they may also contribute to heightened ALS anxiety- CyberALS -the term we invented for this matter. Our objective is to explore the phenomenon of AI-driven anxiety related to ALS in non-ALS patients. Examining how AI-mediated health information influences symptom interpretation and what factors contribute to a higher risk of developing anxiety towards ALS. Methods: Between 2021 and 2025, 582 consecutive patients presenting with neuromuscular complaints were referred to the ALS clinics at the First University Clinic of TSMU and Medcenter Batumi with fear of having ALS. Following comprehensive neurological examination and appropriate longitudinal diagnostic investigations, 220 individuals were determined to have benign symptoms without evidence of neuromuscular disease and were included in the analysis. Participants were stratified according to self-reported use of AI-based symptom checkers or generative AI platforms: 143 patients reported repeated AI exposure, while 77 patients with comparable clinical presentations had not used AI tools. Demographic variables (age, sex, educational level) and personal experience with neurological illness were recorded. Anxiety severity was assessed in all participants using the Hamilton Anxiety Rating Scale (HAM-A), and anxiety levels were compared between AI users and non-users. Statistical analysis was performed using binary logistic regression models, separately for AI users and non-AI users. Statistical significance was defined as p < 0.05. Results: Among 220 patients evaluated (2021–2025), the majority presented with diffuse fasciculations and subjective weakness without objective neurological deficits. No patient fulfilled clinical or electrophysiological criteria for motor neuron disease. Anxiety assessment revealed elevated levels across the AI using cohort (N143), with mean HAM-A scores in the moderate-to-severe (24–30) range. Higher anxiety scores were significantly more frequent among younger patients (age 22–28) and those with higher educational attainment. Comparative analysis reveals a statistically significant increase in anxiety levels among patients who utilized AI platforms, relative to a control group (N77) that abstained from AI-assisted self-diagnosis. Personal or familial history of neurological illness further amplified anxiety severity and disease-related fear. Conclusion: This study demonstrates that exposure to AI chatbots may contribute to clinically significant health anxiety and persistent fear of ALS in patients presenting with benign neuromuscular symptoms. Despite the absence of objective evidence for motor neuron disease, elevated anxiety levels were universal and frequently disproportionate. Addressing cyberALS is essential to ensure that digital health technologies support, rather than undermine, psychological well-being.

The increasing complexity of financial markets and the limitations of traditional stress-testing methods have created a need for more adaptive approaches to systemic risk assessment, particularly in unstable economic environments. This study develops and validates an ensemble artificial intelligence framework for dynamic stress-testing and systemic risk management using quarterly macro-financial data from 1999 to 2024 drawn from the Bank for International Settlements, the International Monetary Fund, the Federal Reserve Economic Data system, and NYU Stern’s V-Lab. Five ensemble methods − Random Forest, XGBoost, CatBoost, Histogram-based Gradient Boosting, and a Stacking ensemble − are estimated and compared against linear regression and naive forecast baselines, with the credit-to-GDP gap, the debt service ratio, and SRISK as target variables. The Stacking ensemble achieves the lowest out-of-sample prediction errors, reducing root mean squared error by 32 per cent relative to linear regression for both the credit gap and SRISK, with Diebold-Mariano tests confirming statistical significance at the 1 per cent level. SHAP analysis identifies the lagged credit-to-GDP gap, the debt service ratio, the VIX, GDP growth, and corporate bond spreads as the most important predictors, revealing non-linear thresholds at a credit gap of 10 per cent and a debt service ratio of 18 per cent. Dynamic stress-testing simulations show that an extreme scenario pushes the representative emerging economy above the 10 per cent crisis threshold within eight quarters, while the advanced economy remains below it. The early warning evaluation achieves an area under the ROC curve of 0.91 at the four-quarter horizon and a signal-to-noise ratio exceeding the 2:1 policy-useful threshold. The study concludes that ensemble methods substantially outperform traditional models in systemic risk prediction, that the identified non-linear thresholds provide empirically grounded anchors for macroprudential monitoring, and that emerging economies face higher systemic risk exposures under identical shock scenarios.

Background: Obesity and overweight are major global health concerns and it is obvious, that the trend of the cases is growing over time. These states are closely linked to type 2 diabetes mellitus, cardiovascular disease, hypertension, heart failure with preserved ejection fraction and various malignancies. Excess adiposity promotes insulin resistance, chronic inflammation, dyslipidemia, endothelial dysfunction, and neurohormonal activation. Growing interest in natural product–based therapeutics has highlighted polyphenol-rich plant extracts as potential adjunctive strategies for metabolic regulation. Cherry laurel (Prunus laurocerasus) is a polyphenol-rich fruit containing bioactive compounds such as quercetin, kaempferol, chlorogenic acid, and anthocyanins with reported anti-obesity effects. Methods: Extracts of fresh, frozen, and dried Prunus laurocerasus fruits, collected from different regions of Georgia were analyzed for polyphenolic content. The biological activities of the identified bioactive compounds were contextualized based on previously published preclinical studies evaluating their effects on adipogenesis, lipid metabolism, glucose homeostasis, inflammatory signaling and energy regulation. Results: All extracts demonstrated consistently high polyphenolic content regardless of processing method or geographic origin. Chlorogenic acid and anthocyanins were identified as the predominant constituents, followed by substantial levels of quercetin and comparatively lower concentrations of kaempferol. Conclusion: The polyphenolic profile of Prunus laurocerasus suggests its potential as a multifaceted adjunctive strategy in metabolic syndrome management, warranting further translational and clinical investigation.