Bing Yang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Bing Yang has a total of 205 co-authors. Here are the top co-authors by number of shared publications:
Santosh Chokkakula
Department of Microbiology, Chungbuk National University College of Medicine and Medical Research Institute
4 shared publications- Bommireddy Naveen
Department of Chemical and Biological Engineering, Gachon University
3 shared publications - Chang Xu
Tianjin Key Laboratory of Radiation Medicine and Molecular Nuclear Medicine, Institute of Radiation Medicine, State Key Laboratory of Advanced Medical Materials and Devices, Chinese Academy of Medical Sciences and Peking Union Medical College, Tianjin, China
3 shared publications Manman Qiu
College of Life Sciences, Nankai University, Tianjin, China
3 shared publications- Qiang Liu
Institute of Radiation Medicine, Chinese Academy of Medical Sciences & Peking Union Medical College, Tianjin, China
3 shared publications Qiaoman Fei
Department of Cellular Biology, School of Basic Medical Sciences, Tianjin Medical University, Tianjin 300070, China
3 shared publications
Latest Publications
- Open Access
- 10 November 2026
- Published by Frontiers Media SA in Frontiers in Artificial Intelligence
- Vol. 9, 1988857
An explainable ensemble learning approach for liver disease risk prediction with interpretable clinical summaries
Background: Liver disease causes more than 2 million annual deaths, often asymptomatic until the early stages. Routine liver function tests (LFTs) are typically interpreted univariately. We developed an explainable machine learning (ML) model for liver disease risk prediction using LFTs. Methods: Models (logistic regression, random forest, gradient boosting, soft-voting ensemble) were trained on the Kaggle Liver Disease Patient Dataset (LPD, n = 28,948 after removing 1,743 overlaps) and externally validated on the UCI Indian Liver Patient Dataset (ILPD, n = 583). Performance was assessed via AUC and DeLong’s test; SHAP provided interpretability and patient-level summaries. Results: The gradient boosting model achieved the best discrimination, with an AUC of 0.96 in the internal holdout set and 0.967 (95% CI 0.953–0.980) in the ILPD validation set, significantly outperforming random forest and the soft-voting ensemble (both AUC 0.939) and logistic regression (AUC 0.759, all p < 0.001). At the conventional 0.50 threshold, the model identified 410/416 liver disease cases (sensitivity 98.6%, accuracy 88.5%). At an optimised threshold of 0.55, accuracy improved to 91.6%, with sensitivity 97.4%, specificity 77.2%, precision 91.4%, and F1-score 0.943; average precision was 0.99 on ILPD. Five-tier risk stratification showed a stepwise increase in observed prevalence from 0% (Very Low) to 98.7% (Very High); the two lowest tiers contained only 1/85 cases, while the two highest tiers captured 97.4% of all cases. SHAP analysis identified direct bilirubin (mean |SHAP| 0.427), alanine aminotransferase (0.367), and alkaline phosphatase (0.355) as the strongest predictors; age and sex contributed minimally. Automated SHAP-based summaries highlighted key contributing biomarkers per patient and flagged discordant findings for review. Conclusion: An explainable gradient boosting model using routine LFTs delivers high-accuracy liver disease prediction, with transparent risk stratification and SHAP-based clinical summaries to aid early detection and triage.
- 10 November 2026
- Published by Frontiers Media SA in Frontiers in Artificial Intelligence
- Vol. 9, 1988857
Open AccessAn explainable ensemble learning approach for liver disease risk prediction with interpretable clinical summaries
Background: Liver disease causes more than 2 million annual deaths, often asymptomatic until the early stages. Routine liver function tests (LFTs) are typically interpreted univariately. We developed an explainable machine learning (ML) model for liver disease risk prediction using LFTs. Methods: Models (logistic regression, random forest, gradient boosting, soft-voting ensemble) were trained on the Kaggle Liver Disease Patient Dataset (LPD, n = 28,948 after removing 1,743 overlaps) and externally validated on the UCI Indian Liver Patient Dataset (ILPD, n = 583). Performance was assessed via AUC and DeLong’s test; SHAP provided interpretability and patient-level summaries. Results: The gradient boosting model achieved the best discrimination, with an AUC of 0.96 in the internal holdout set and 0.967 (95% CI 0.953–0.980) in the ILPD validation set, significantly outperforming random forest and the soft-voting ensemble (both AUC 0.939) and logistic regression (AUC 0.759, all p < 0.001). At the conventional 0.50 threshold, the model identified 410/416 liver disease cases (sensitivity 98.6%, accuracy 88.5%). At an optimised threshold of 0.55, accuracy improved to 91.6%, with sensitivity 97.4%, specificity 77.2%, precision 91.4%, and F1-score 0.943; average precision was 0.99 on ILPD. Five-tier risk stratification showed a stepwise increase in observed prevalence from 0% (Very Low) to 98.7% (Very High); the two lowest tiers contained only 1/85 cases, while the two highest tiers captured 97.4% of all cases. SHAP analysis identified direct bilirubin (mean |SHAP| 0.427), alanine aminotransferase (0.367), and alkaline phosphatase (0.355) as the strongest predictors; age and sex contributed minimally. Automated SHAP-based summaries highlighted key contributing biomarkers per patient and flagged discordant findings for review. Conclusion: An explainable gradient boosting model using routine LFTs delivers high-accuracy liver disease prediction, with transparent risk stratification and SHAP-based clinical summaries to aid early detection and triage.
- Open Access
- 10 September 2026
- Published by Springer Nature in Molecular Cancer
The abscopal effect and systemic immune reprogramming: from phenomenon to therapeutic design
- 10 September 2026
- Published by Springer Nature in Molecular Cancer
Open AccessThe abscopal effect and systemic immune reprogramming: from phenomenon to therapeutic design
- Open Access
- 24 July 2026
- Published by Frontiers Media SA in Frontiers in Cellular and Infection Microbiology
- Vol. 16, 1859541
From neuraminidase inhibitors to novel mechanisms: expanding therapeutic strategies against influenza
Seasonal influenza causes approximately one billion infections and up to 650,000 respiratory deaths annually. Despite decades of antiviral development, the therapeutic arsenal remains constrained by time-dependent efficacy (clinical benefit confined largely to treatment initiated within 48 hours of symptom onset), low genetic resistance barriers (resistance arising from a single point mutation), and critical evidence gaps in high-risk populations. This review critically evaluates the antiviral landscape, focusing on next-generation cap-dependent endonuclease inhibitors (CENIs), resistance implications across all approved drug classes, translational barriers in host-directed therapies and biologics, and H5N1 pandemic preparedness. Suraxavir marboxil, ZX-7101A, and the first approved polymerase basic protein 2 (PB2) inhibitor, onradivir, confirm that scaffold-level optimization (incremental structural refinement within an existing drug class) can meaningfully reduce resistance emergence rates. Across all approved classes, a shared failure pattern is evident: single-target dependence, low-fitness-cost resistance mutations (mutations that confer drug resistance without measurably impairing viral replication or transmission), and a therapeutic window that most patients do not reach. Combination antiviral therapy shows promise for resistance prevention, though the FLAGSTONE Phase 3 trial demonstrated that virological benefit does not consistently translate to clinical outcomes in hospitalized patients. Host-directed therapies and biologics face a shared translational obstacle: the biological window of maximal activity consistently precedes the clinical window of patient presentation, and systemic biologics cannot achieve inhibitory concentrations at the respiratory mucosal surface. For H5N1, preparedness relies on a single observationally supported drug class; baloxavir is absent from most national stockpiles, and no randomized trial data exists for any antiviral in human H5N1 infection. Adaptive trial infrastructure, diversified stockpiling to include next-generation CENIs, and equitable global antiviral access are the most urgent priorities in influenza therapeutics.
- 24 July 2026
- Published by Frontiers Media SA in Frontiers in Cellular and Infection Microbiology
- Vol. 16, 1859541
Open AccessFrom neuraminidase inhibitors to novel mechanisms: expanding therapeutic strategies against influenza
Seasonal influenza causes approximately one billion infections and up to 650,000 respiratory deaths annually. Despite decades of antiviral development, the therapeutic arsenal remains constrained by time-dependent efficacy (clinical benefit confined largely to treatment initiated within 48 hours of symptom onset), low genetic resistance barriers (resistance arising from a single point mutation), and critical evidence gaps in high-risk populations. This review critically evaluates the antiviral landscape, focusing on next-generation cap-dependent endonuclease inhibitors (CENIs), resistance implications across all approved drug classes, translational barriers in host-directed therapies and biologics, and H5N1 pandemic preparedness. Suraxavir marboxil, ZX-7101A, and the first approved polymerase basic protein 2 (PB2) inhibitor, onradivir, confirm that scaffold-level optimization (incremental structural refinement within an existing drug class) can meaningfully reduce resistance emergence rates. Across all approved classes, a shared failure pattern is evident: single-target dependence, low-fitness-cost resistance mutations (mutations that confer drug resistance without measurably impairing viral replication or transmission), and a therapeutic window that most patients do not reach. Combination antiviral therapy shows promise for resistance prevention, though the FLAGSTONE Phase 3 trial demonstrated that virological benefit does not consistently translate to clinical outcomes in hospitalized patients. Host-directed therapies and biologics face a shared translational obstacle: the biological window of maximal activity consistently precedes the clinical window of patient presentation, and systemic biologics cannot achieve inhibitory concentrations at the respiratory mucosal surface. For H5N1, preparedness relies on a single observationally supported drug class; baloxavir is absent from most national stockpiles, and no randomized trial data exists for any antiviral in human H5N1 infection. Adaptive trial infrastructure, diversified stockpiling to include next-generation CENIs, and equitable global antiviral access are the most urgent priorities in influenza therapeutics.
- Open Access
- 20 May 2026
- Published by Frontiers Media SA in Frontiers in Molecular Biosciences
- Vol. 13, 1873649
Editorial: New tumor immune checkpoints and their applications in tumor immunotherapy
More than a century ago, William Coley observed that bacterial products could trigger tumor regression, a finding that laid the foundation for cancer immunotherapy despite the era's limited immunological knowledge. Today, we continue to pursue the same fundamental goal: equipping the immune system to recognize and eradicate malignant cells. The advent of immune checkpoint inhibitors (ICIs) targeting PD-1/PD-L1 and CTLA-4 has profoundly reshaped the therapeutic landscape for numerous malignancies. Yet durable clinical benefit remains limited to a subset of patients, and both primary and acquired resistance pose enduring challenges. This has spurred an intensive search for novel immune checkpoints and robust predictive biomarkers to cancer patients receiving ICIs and find no significant changes in liver attenuation or liver enzyme levels, offering reassuring evidence that these agents may not exacerbate hepatic steatosis or subclinical liver injury-a finding particularly relevant for managing patients with pre-existing metabolic liver disease. These clinical vignettes and systematic reviews underscore the need for a multidimensional evaluation framework that incorporates baseline patient characteristics, dynamic biomarker monitoring, and vigilant assessment of immune-related adverse events.The articles collected in this Research Topic illustrate the breadth and depth of current investigation into novel tumor immune checkpoints and their applications.Looking ahead, several priorities warrant attention. The functional overlap and contextdependent expression of immune regulatory molecules call for systematic combinatorial strategies grounded in mechanistic understanding, potentially pairing novel checkpoint blockade with epigenetic modulators, metabolic inhibitors, or rational vaccine platforms. Translating emerging targets from preclinical validation to clinical deployment will require coordinated efforts, innovative trial designs, and the integration of integrative multi-omics approaches-from genomics and transcriptomics to single-cell analyses-to delineate the interplay among tumor metabolism, immune regulation, and therapeutic response. The continued incorporation of artificial intelligence into diagnostic and predictive pipelines is poised to accelerate progress across these domains. We extend our sincere appreciation to all the authors, reviewers, and the editorial team for their contributions, and hope this collection will serve as a catalyst for further investigation, interdisciplinary collaboration, and ultimately the advancement of effective, personalized immunotherapy for patients with cancer.
- 20 May 2026
- Published by Frontiers Media SA in Frontiers in Molecular Biosciences
- Vol. 13, 1873649
Open AccessEditorial: New tumor immune checkpoints and their applications in tumor immunotherapy
More than a century ago, William Coley observed that bacterial products could trigger tumor regression, a finding that laid the foundation for cancer immunotherapy despite the era's limited immunological knowledge. Today, we continue to pursue the same fundamental goal: equipping the immune system to recognize and eradicate malignant cells. The advent of immune checkpoint inhibitors (ICIs) targeting PD-1/PD-L1 and CTLA-4 has profoundly reshaped the therapeutic landscape for numerous malignancies. Yet durable clinical benefit remains limited to a subset of patients, and both primary and acquired resistance pose enduring challenges. This has spurred an intensive search for novel immune checkpoints and robust predictive biomarkers to cancer patients receiving ICIs and find no significant changes in liver attenuation or liver enzyme levels, offering reassuring evidence that these agents may not exacerbate hepatic steatosis or subclinical liver injury-a finding particularly relevant for managing patients with pre-existing metabolic liver disease. These clinical vignettes and systematic reviews underscore the need for a multidimensional evaluation framework that incorporates baseline patient characteristics, dynamic biomarker monitoring, and vigilant assessment of immune-related adverse events.The articles collected in this Research Topic illustrate the breadth and depth of current investigation into novel tumor immune checkpoints and their applications.Looking ahead, several priorities warrant attention. The functional overlap and contextdependent expression of immune regulatory molecules call for systematic combinatorial strategies grounded in mechanistic understanding, potentially pairing novel checkpoint blockade with epigenetic modulators, metabolic inhibitors, or rational vaccine platforms. Translating emerging targets from preclinical validation to clinical deployment will require coordinated efforts, innovative trial designs, and the integration of integrative multi-omics approaches-from genomics and transcriptomics to single-cell analyses-to delineate the interplay among tumor metabolism, immune regulation, and therapeutic response. The continued incorporation of artificial intelligence into diagnostic and predictive pipelines is poised to accelerate progress across these domains. We extend our sincere appreciation to all the authors, reviewers, and the editorial team for their contributions, and hope this collection will serve as a catalyst for further investigation, interdisciplinary collaboration, and ultimately the advancement of effective, personalized immunotherapy for patients with cancer.
- Open Access
- 14 April 2026
- Published by Frontiers Media SA in Frontiers in Pharmacology
- Vol. 17, 1744180
Post-marketing safety evaluation of paroxetine: a real-world pharmacovigilance analysis based on the FDA adverse event reporting system
Objective: Paroxetine, a selective serotonin reuptake inhibitor widely used in the management of major depressive disorders and other psychiatric disorders, remains insufficiently characterized regarding its adverse event spectrum. Thus, this study wanted to conduct a comprehensive evaluation of its post-marketing safety profile by analyzing reports from the U.S. Food and Drug Administration Adverse Event Reporting System (FAERS). Methods: Adverse event data related to paroxetine were extracted from the FAERS spanning the period from the first quarter of 2004 through the second quarter of 2025. Demographic distributions and reporter characteristics were analyzed to describe the exposed population. Signal detection was performed using four established disproportionality algorithms, including Reporting Odds Ratio (ROR), Proportional Reporting Ratio (PRR), Bayesian Confidence Propagation Neural Network (BCPNN), and Multi-item Gamma Poisson Shrinker (MGPS). Results: A total of 39,404 adverse event reports associated with paroxetine were included in the analysis. Female patients accounted for the majority of cases (57.41%), followed by males (32.84%). The most frequently reported preferred term (PT) was “drug withdrawal syndrome,” categorized under systemic diseases and administration site conditions, with 6,601 cases. This event demonstrated strong positive signals across all disproportionality metrics (ROR = 27.27, PRR = 26.31, χ 2 = 148,641, IC = 4.61, EBGM = 23.47). The highest signal strengths were observed within the SOC of congenital, familial, and genetic disorders. Notably, the PT “carcinogenic effect on offspring,” although reported only three times, exhibited markedly elevated values (ROR = 952.54, PRR = 952.52, χ 2 = 712.89, IC = 7.90, EBGM = 238.88), indicating an extremely disproportionate reporting frequency compared with baseline levels. Most adverse events (12.49%) appeared within the first 30 days of paroxetine therapy prescription. Conclusion: The pharmacovigilance assessment of paroxetine systematically identified significant safety signals. These findings provide important post-marketing evidence to strengthen drug safety surveillance and facilitate more informed, patient-centered therapeutic decisions in clinical settings.
- 14 April 2026
- Published by Frontiers Media SA in Frontiers in Pharmacology
- Vol. 17, 1744180
Open AccessPost-marketing safety evaluation of paroxetine: a real-world pharmacovigilance analysis based on the FDA adverse event reporting system
Objective: Paroxetine, a selective serotonin reuptake inhibitor widely used in the management of major depressive disorders and other psychiatric disorders, remains insufficiently characterized regarding its adverse event spectrum. Thus, this study wanted to conduct a comprehensive evaluation of its post-marketing safety profile by analyzing reports from the U.S. Food and Drug Administration Adverse Event Reporting System (FAERS). Methods: Adverse event data related to paroxetine were extracted from the FAERS spanning the period from the first quarter of 2004 through the second quarter of 2025. Demographic distributions and reporter characteristics were analyzed to describe the exposed population. Signal detection was performed using four established disproportionality algorithms, including Reporting Odds Ratio (ROR), Proportional Reporting Ratio (PRR), Bayesian Confidence Propagation Neural Network (BCPNN), and Multi-item Gamma Poisson Shrinker (MGPS). Results: A total of 39,404 adverse event reports associated with paroxetine were included in the analysis. Female patients accounted for the majority of cases (57.41%), followed by males (32.84%). The most frequently reported preferred term (PT) was “drug withdrawal syndrome,” categorized under systemic diseases and administration site conditions, with 6,601 cases. This event demonstrated strong positive signals across all disproportionality metrics (ROR = 27.27, PRR = 26.31, χ 2 = 148,641, IC = 4.61, EBGM = 23.47). The highest signal strengths were observed within the SOC of congenital, familial, and genetic disorders. Notably, the PT “carcinogenic effect on offspring,” although reported only three times, exhibited markedly elevated values (ROR = 952.54, PRR = 952.52, χ 2 = 712.89, IC = 7.90, EBGM = 238.88), indicating an extremely disproportionate reporting frequency compared with baseline levels. Most adverse events (12.49%) appeared within the first 30 days of paroxetine therapy prescription. Conclusion: The pharmacovigilance assessment of paroxetine systematically identified significant safety signals. These findings provide important post-marketing evidence to strengthen drug safety surveillance and facilitate more informed, patient-centered therapeutic decisions in clinical settings.
- Open Access
- 1 March 2026
- Published by AIP Publishing in APL Bioengineering
- Vol. 10 (1)
ChatGPT in precision medicine
The integration of artificial intelligence, notably the generative language model ChatGPT, into precision medicine precedes a paradigm shift in healthcare. This Review explores the capabilities of ChatGPT to process large data sets, understand contextual details, and generate human-like responses, making it a transformative tool for various applications in precision medicine. By analyzing the genetic sequences, identifying biomarkers, and predicting disease risks, ChatGPT facilitates earlier diagnosis and intervention. Moreover, its natural language processing capabilities enhance patient communication and engagement by elucidating complex medical concepts. This Review further examines ChatGPT's role in streamlining research, augmenting clinical trial recruitment, and supporting healthcare professionals through access to updated information and skill assessment. While acknowledging the complexities and potential concerns regarding reliance on publicly sourced data, this Review highlights ChatGPT's substantial capacity to optimize patient prognoses and enhance the operational efficacy of the healthcare delivery system.
- 1 March 2026
- Published by AIP Publishing in APL Bioengineering
- Vol. 10 (1)
Open AccessChatGPT in precision medicine
The integration of artificial intelligence, notably the generative language model ChatGPT, into precision medicine precedes a paradigm shift in healthcare. This Review explores the capabilities of ChatGPT to process large data sets, understand contextual details, and generate human-like responses, making it a transformative tool for various applications in precision medicine. By analyzing the genetic sequences, identifying biomarkers, and predicting disease risks, ChatGPT facilitates earlier diagnosis and intervention. Moreover, its natural language processing capabilities enhance patient communication and engagement by elucidating complex medical concepts. This Review further examines ChatGPT's role in streamlining research, augmenting clinical trial recruitment, and supporting healthcare professionals through access to updated information and skill assessment. While acknowledging the complexities and potential concerns regarding reliance on publicly sourced data, this Review highlights ChatGPT's substantial capacity to optimize patient prognoses and enhance the operational efficacy of the healthcare delivery system.
- Open Access
- 26 February 2026
- Published by Frontiers Media SA in Frontiers in Immunology
- Vol. 17, 1808165
Editorial: Exploring key pathways in the progression of gastrointestinal diseases based on metabolic reprogramming and developing drugs targeting metabolism
Gastrointestinal neoplasms continue to pose a serious threat to public health due to the high prevalence of highly lethal malignancies [1]. These cancers can affect the esophagus, stomach, colon, liver, and pancreas. Patients often present asymptomatically, which leads to delays in diagnosis.Significant advances have been made in cancer treatment over the past several decades, including the development of chemotherapy, targeted therapies, and immunotherapies.However, resistance remains a significant challenge, with aberrant metabolism being recognized as one of the emerging hallmarks of cancer [2]. As a heterogeneous disease, cancer retains its capacity to develop and progress through genetic and molecular changes. [3].The metabolism of these cells is ingeniously modified so that accelerated proliferation can be facilitated, apoptosis can be evaded, and even drug resistance can be developed. Recently, metabolic alterations have proven to be important in primary tumors, such as the Warburg effect (aerobic glycolysis), glutamine addiction, and reprogramming of lipid metabolism [4,5].Metabolic reprogramming of tumors may also interact with the tumor microenvironment (TME) by secreting metabolites such as lactate and succinate, thereby promoting immune suppression, angiogenesis, and stromal remodeling. These processes play an important role in the metastasis of gastrointestinal tumors, either directly or indirectly [6]. This present editorial introduces the comprehensive collection of articles that have hitherto been published in Frontiers in Immunology. The present volume is part of the Cancer Immunity and Immunotherapy collection. Additionally, the author discovered that GIPC2 has the potential to serve as both a biomarker and a therapeutic target.
- 26 February 2026
- Published by Frontiers Media SA in Frontiers in Immunology
- Vol. 17, 1808165
Open AccessEditorial: Exploring key pathways in the progression of gastrointestinal diseases based on metabolic reprogramming and developing drugs targeting metabolism
Gastrointestinal neoplasms continue to pose a serious threat to public health due to the high prevalence of highly lethal malignancies [1]. These cancers can affect the esophagus, stomach, colon, liver, and pancreas. Patients often present asymptomatically, which leads to delays in diagnosis.Significant advances have been made in cancer treatment over the past several decades, including the development of chemotherapy, targeted therapies, and immunotherapies.However, resistance remains a significant challenge, with aberrant metabolism being recognized as one of the emerging hallmarks of cancer [2]. As a heterogeneous disease, cancer retains its capacity to develop and progress through genetic and molecular changes. [3].The metabolism of these cells is ingeniously modified so that accelerated proliferation can be facilitated, apoptosis can be evaded, and even drug resistance can be developed. Recently, metabolic alterations have proven to be important in primary tumors, such as the Warburg effect (aerobic glycolysis), glutamine addiction, and reprogramming of lipid metabolism [4,5].Metabolic reprogramming of tumors may also interact with the tumor microenvironment (TME) by secreting metabolites such as lactate and succinate, thereby promoting immune suppression, angiogenesis, and stromal remodeling. These processes play an important role in the metastasis of gastrointestinal tumors, either directly or indirectly [6]. This present editorial introduces the comprehensive collection of articles that have hitherto been published in Frontiers in Immunology. The present volume is part of the Cancer Immunity and Immunotherapy collection. Additionally, the author discovered that GIPC2 has the potential to serve as both a biomarker and a therapeutic target.
- Open Access
- 18 February 2026
- Published by Frontiers Media SA in Frontiers in Endocrinology
- Vol. 17, 1801881
Editorial: Exploring Neuroendocrine Mechanisms in Psychiatric and Metabolic Comorbidities
The human body functions as a highly integrated biochemical network governed by intricate regulatory and feedback mechanisms. This biological "chemical plant" operates through tightly coordinated neuroendocrine and metabolic pathways that maintain systemic homeostasis. Throughout scientific history, sustained efforts have been made to unravel the fundamental principles governing these processes; however, their complexity continues to pose challenges long before a comprehensive understanding has been achieved. When any component of this regulatory framework, hormonal, neuronal, or molecular, malfunctions, the effects can propagate across multiple organ systems. Even subtle cellular perturbations may accumulate over time, predisposing individuals to complex pathological states, including psychiatric and metabolic disorders. Among these interconnected systems, neuroendocrine signaling plays a central role in linking brain function with peripheral metabolism. Dysregulation of the hypothalamicpituitary-adrenal (HPA) axis has been consistently implicated in the pathophysiology of numerous mental illnesses (1,2). Contemporary neuropeptidergic models highlight the importance of stress-responsive peptides such as corticotropin-releasing hormone, oxytocin, and vasopressin in modulating emotional regulation, cognitive processing, and social behavior (3). Aberrant activity within these neuroendocrine circuits contributes to the pathogenesis of depression, anxiety, and schizophrenia, partly through their downstream metabolic interactions (4,5). Neuropeptides, secreted in circadian patterns that are dynamically altered by stress, exert both central and peripheral influences, forming key molecular bridges among mood, cognition, and energy homeostasis.The high prevalence of comorbidity between psychiatric disorders and metabolic diseases, particularly obesity and type 2 diabetes, underscores a multidirectional disruption of the brain-body axis (6,7). Meta-analytic evidence indicates that individuals with schizophrenia, bipolar disorder, or major depressive disorder exhibit persistent alterations in glucose regulation, lipid metabolism, and autonomic function, even in the absence of pharmacological confounders (8). Conversely, metabolic dysregulation can amplify neuroinflammatory and oxidative stress pathways, accelerating psychiatric symptom progression and cognitive decline. Despite growing recognition of this bidirectional relationship, clinical management remains suboptimal. Diagnostic biases, social stigma, and the metabolic side effects of psychopharmacological agents continue to obscure accurate assessment and timely intervention.Recent research extends this concept by linking systemic metabolic imbalance to organspecific abnormalities in psychiatric populations. Metabolically-dysfunctionassociated fatty liver disease (MAFLD), for example, has been identified even in lean individuals with bipolar disorder, indicating underlying metabolic vulnerability irrespective of body mass index (9). Similarly, autonomic markers such as heart rate deceleration capacity have shown prognostic value in acute ischemic stroke, a condition influenced by both metabolic and neuropsychiatric factors (10). These observations reinforce the hypothesis that cardiovascular, neuroendocrine, and psychiatric systems form a dynamically regulated continuum rather than discrete entities.Parallel advances in biomarker research provide new diagnostic perspectives. Altered circulating levels of norepinephrine and melatonin have emerged as reproducible neuroendocrine signatures in schizophrenia, reflecting disruptions in circadianmetabolic coupling (11). Meanwhile, advances in cardiac imaging, such as twodimensional speckle tracking echocardiography, have enabled non-invasive assessment of myocardial strain and therapeutic outcomes in patients with combined heart failure and metabolic syndrome, further illustrating the systemic reach of neuroendocrine regulation.Recent developments in computational modeling and biomarker integration demonstrate the translational potential of multi-parametric diagnostics. Predictive nomograms incorporating blood-derived biomarkers have successfully forecasted short-term recovery outcomes in stroke patients, where neuroendocrine stress reactivity serves as a critical determinant. Collectively, such findings highlight a paradigm shift: psychiatric and metabolic disorders must be conceptualized as interconnected manifestations of a shared neuroendocrine framework rather than as isolated conditions.Ultimately, elucidating neuroendocrine mechanisms underlying psychiatric-metabolic comorbidity presents an opportunity to redefine both disease classification and therapeutic intervention. Future strategies aimed at modulating neuropeptide signaling, circadian alignment, and stress reactivity hold promise for enhancing mental and metabolic health concurrently. Progress in this field will depend on interdisciplinary collaboration across neurobiology, psychiatry, and endocrinology, coupled with the integration of multi-omic datasets, longitudinal biomarker profiling, and advanced computational modeling to comprehensively characterize the bidirectional dialogue between the brain and body.
- 18 February 2026
- Published by Frontiers Media SA in Frontiers in Endocrinology
- Vol. 17, 1801881
Open AccessEditorial: Exploring Neuroendocrine Mechanisms in Psychiatric and Metabolic Comorbidities
The human body functions as a highly integrated biochemical network governed by intricate regulatory and feedback mechanisms. This biological "chemical plant" operates through tightly coordinated neuroendocrine and metabolic pathways that maintain systemic homeostasis. Throughout scientific history, sustained efforts have been made to unravel the fundamental principles governing these processes; however, their complexity continues to pose challenges long before a comprehensive understanding has been achieved. When any component of this regulatory framework, hormonal, neuronal, or molecular, malfunctions, the effects can propagate across multiple organ systems. Even subtle cellular perturbations may accumulate over time, predisposing individuals to complex pathological states, including psychiatric and metabolic disorders. Among these interconnected systems, neuroendocrine signaling plays a central role in linking brain function with peripheral metabolism. Dysregulation of the hypothalamicpituitary-adrenal (HPA) axis has been consistently implicated in the pathophysiology of numerous mental illnesses (1,2). Contemporary neuropeptidergic models highlight the importance of stress-responsive peptides such as corticotropin-releasing hormone, oxytocin, and vasopressin in modulating emotional regulation, cognitive processing, and social behavior (3). Aberrant activity within these neuroendocrine circuits contributes to the pathogenesis of depression, anxiety, and schizophrenia, partly through their downstream metabolic interactions (4,5). Neuropeptides, secreted in circadian patterns that are dynamically altered by stress, exert both central and peripheral influences, forming key molecular bridges among mood, cognition, and energy homeostasis.The high prevalence of comorbidity between psychiatric disorders and metabolic diseases, particularly obesity and type 2 diabetes, underscores a multidirectional disruption of the brain-body axis (6,7). Meta-analytic evidence indicates that individuals with schizophrenia, bipolar disorder, or major depressive disorder exhibit persistent alterations in glucose regulation, lipid metabolism, and autonomic function, even in the absence of pharmacological confounders (8). Conversely, metabolic dysregulation can amplify neuroinflammatory and oxidative stress pathways, accelerating psychiatric symptom progression and cognitive decline. Despite growing recognition of this bidirectional relationship, clinical management remains suboptimal. Diagnostic biases, social stigma, and the metabolic side effects of psychopharmacological agents continue to obscure accurate assessment and timely intervention.Recent research extends this concept by linking systemic metabolic imbalance to organspecific abnormalities in psychiatric populations. Metabolically-dysfunctionassociated fatty liver disease (MAFLD), for example, has been identified even in lean individuals with bipolar disorder, indicating underlying metabolic vulnerability irrespective of body mass index (9). Similarly, autonomic markers such as heart rate deceleration capacity have shown prognostic value in acute ischemic stroke, a condition influenced by both metabolic and neuropsychiatric factors (10). These observations reinforce the hypothesis that cardiovascular, neuroendocrine, and psychiatric systems form a dynamically regulated continuum rather than discrete entities.Parallel advances in biomarker research provide new diagnostic perspectives. Altered circulating levels of norepinephrine and melatonin have emerged as reproducible neuroendocrine signatures in schizophrenia, reflecting disruptions in circadianmetabolic coupling (11). Meanwhile, advances in cardiac imaging, such as twodimensional speckle tracking echocardiography, have enabled non-invasive assessment of myocardial strain and therapeutic outcomes in patients with combined heart failure and metabolic syndrome, further illustrating the systemic reach of neuroendocrine regulation.Recent developments in computational modeling and biomarker integration demonstrate the translational potential of multi-parametric diagnostics. Predictive nomograms incorporating blood-derived biomarkers have successfully forecasted short-term recovery outcomes in stroke patients, where neuroendocrine stress reactivity serves as a critical determinant. Collectively, such findings highlight a paradigm shift: psychiatric and metabolic disorders must be conceptualized as interconnected manifestations of a shared neuroendocrine framework rather than as isolated conditions.Ultimately, elucidating neuroendocrine mechanisms underlying psychiatric-metabolic comorbidity presents an opportunity to redefine both disease classification and therapeutic intervention. Future strategies aimed at modulating neuropeptide signaling, circadian alignment, and stress reactivity hold promise for enhancing mental and metabolic health concurrently. Progress in this field will depend on interdisciplinary collaboration across neurobiology, psychiatry, and endocrinology, coupled with the integration of multi-omic datasets, longitudinal biomarker profiling, and advanced computational modeling to comprehensively characterize the bidirectional dialogue between the brain and body.
- Top Cited PapersOpen Access
- 21 January 2026
- Published by Frontiers Media SA in Frontiers in Pharmacology
- Vol. 17, 1737113
The post-marketing safety of venlafaxine: a real-world two-decade pharmacovigilance study using the FAERS database
Objective: Venlafaxine, a serotonin-norepinephrine reuptake inhibitor widely prescribed for major depressive disorder and related conditions, remains insufficiently characterized regarding its adverse event profile. This study aimed to comprehensively evaluate the post-marketing safety of venlafaxine using data from the U.S. Food and Drug Administration Adverse Event Reporting System (FAERS). Methods: The data were extracted from the FAERS database from the first quarter (Q1) 2004 to Q2 2025. Adverse drug events were analyzed using the Reporting Odds Ratio (ROR), Bayesian Confidence Propagation Neural Network (BCPNN), Proportional Reporting Ratio (PRR), and Multi-Item Gamma Poisson Shrinker (MGPS) methods. Results: A total of 47,325 venlafaxine-related adverse event reports were analyzed. Disproportionality analysis revealed significant signals for drug ineffectiveness (n = 2,108, ROR = 1.6, PRR = 1.6, χ 2 = 483.4, IC = 0.7, EBGM = 1.6), drug hypersensitivity (n = 1,014; ROR = 4.9, PRR = 4.8, χ 2 = 2998.7, IC = 2.2, EBGM = 4.7), and anxiety (n = 198, ROR = 4.3, PRR = 4.2, χ 2 = 2202.0, IC = 2.0, EBGM = 4.1). Strongest signals (EBGM = 9.9) were observed for suicide attempt, agitated depression, and renal dysplasia. Tachycardia (EBGM = 8.3) was the most frequently reported adverse event among patients under 18 years, while emotional disorder (EBGM = 9.5) predominated in those aged 65 years and older. Most adverse events (39.3%) occurred within the first 30 days of venlafaxine therapy initiation. Conclusion: This pharmacovigilance analysis systematically identified significant safety signals associated with venlafaxine. The findings provide important evidence to support safer clinical use of venlafaxine and may assist in optimizing individualized therapeutic decisions in practice.
- 21 January 2026
- Published by Frontiers Media SA in Frontiers in Pharmacology
- Vol. 17, 1737113
Top Cited PapersOpen AccessThe post-marketing safety of venlafaxine: a real-world two-decade pharmacovigilance study using the FAERS database
Objective: Venlafaxine, a serotonin-norepinephrine reuptake inhibitor widely prescribed for major depressive disorder and related conditions, remains insufficiently characterized regarding its adverse event profile. This study aimed to comprehensively evaluate the post-marketing safety of venlafaxine using data from the U.S. Food and Drug Administration Adverse Event Reporting System (FAERS). Methods: The data were extracted from the FAERS database from the first quarter (Q1) 2004 to Q2 2025. Adverse drug events were analyzed using the Reporting Odds Ratio (ROR), Bayesian Confidence Propagation Neural Network (BCPNN), Proportional Reporting Ratio (PRR), and Multi-Item Gamma Poisson Shrinker (MGPS) methods. Results: A total of 47,325 venlafaxine-related adverse event reports were analyzed. Disproportionality analysis revealed significant signals for drug ineffectiveness (n = 2,108, ROR = 1.6, PRR = 1.6, χ 2 = 483.4, IC = 0.7, EBGM = 1.6), drug hypersensitivity (n = 1,014; ROR = 4.9, PRR = 4.8, χ 2 = 2998.7, IC = 2.2, EBGM = 4.7), and anxiety (n = 198, ROR = 4.3, PRR = 4.2, χ 2 = 2202.0, IC = 2.0, EBGM = 4.1). Strongest signals (EBGM = 9.9) were observed for suicide attempt, agitated depression, and renal dysplasia. Tachycardia (EBGM = 8.3) was the most frequently reported adverse event among patients under 18 years, while emotional disorder (EBGM = 9.5) predominated in those aged 65 years and older. Most adverse events (39.3%) occurred within the first 30 days of venlafaxine therapy initiation. Conclusion: This pharmacovigilance analysis systematically identified significant safety signals associated with venlafaxine. The findings provide important evidence to support safer clinical use of venlafaxine and may assist in optimizing individualized therapeutic decisions in practice.
- Open Access
- 12 January 2026
- Published by Frontiers Media SA in Frontiers in Medicine
- Vol. 12, 1728061
Hybrid intelligent systems for liver disease prediction: a demographic-aware machine learning framework
Background: Liver disease remains a major global health burden, often progressing undetected until advanced stages. Traditional diagnostic approaches, while accurate, are invasive, costly, and limited in accessibility. Objective: To address these challenges, we propose a hybrid intelligent framework that integrates demographic segmentation with advanced machine learning for the early detection of liver disease. Results: Two datasets were employed, including the Indian Liver Patient Dataset (ILPD, n = 583) and a large-scale dataset ( n = 29,787). Patients were stratified by age and gender into six groups, enabling segment-specific model development. Sixteen algorithms, including Random Forest, Support Vector Machines, XGBoost, and LightGBM, were evaluated using recursive feature elimination, resampling techniques, and Bayesian hyperparameter optimization. Segment-specific best models were integrated into a hybrid system through dynamic selection and ensemble strategies. The framework achieved 94.2% accuracy on ILPD and 99.8% on the large dataset, with consistent improvements across demographic groups. Feature analysis revealed distinct biomarker importance by age and gender, underscoring the need for tailored diagnostic approaches.. Conclusion: By combining demographic awareness, hybrid learning, and interpretability, this study offers a scalable, non-invasive, and clinically relevant tool for early detection of liver diseases, advancing personalized and accessible healthcare.
- 12 January 2026
- Published by Frontiers Media SA in Frontiers in Medicine
- Vol. 12, 1728061
Open AccessHybrid intelligent systems for liver disease prediction: a demographic-aware machine learning framework
Background: Liver disease remains a major global health burden, often progressing undetected until advanced stages. Traditional diagnostic approaches, while accurate, are invasive, costly, and limited in accessibility. Objective: To address these challenges, we propose a hybrid intelligent framework that integrates demographic segmentation with advanced machine learning for the early detection of liver disease. Results: Two datasets were employed, including the Indian Liver Patient Dataset (ILPD, n = 583) and a large-scale dataset ( n = 29,787). Patients were stratified by age and gender into six groups, enabling segment-specific model development. Sixteen algorithms, including Random Forest, Support Vector Machines, XGBoost, and LightGBM, were evaluated using recursive feature elimination, resampling techniques, and Bayesian hyperparameter optimization. Segment-specific best models were integrated into a hybrid system through dynamic selection and ensemble strategies. The framework achieved 94.2% accuracy on ILPD and 99.8% on the large dataset, with consistent improvements across demographic groups. Feature analysis revealed distinct biomarker importance by age and gender, underscoring the need for tailored diagnostic approaches.. Conclusion: By combining demographic awareness, hybrid learning, and interpretability, this study offers a scalable, non-invasive, and clinically relevant tool for early detection of liver diseases, advancing personalized and accessible healthcare.