PubMed Health⌕ Search

SEARCH · PubMed Health

Results for “Generative artificial intelligence”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 19 recordsLinked to original sources
PubMed2026

Generative artificial intelligence for enzyme design and biocatalysis.

Sparked by innovations in generative artificial intelligence (AI), the field of protein design has undergone a paradigm shift with an explosion of new models for optimizing existing enzymes or creating them from scratch. After more than one decade of low success rates for computationally designed enzymes, generative AI models are now frequently used for designing proficient enzymes. Here, we provide a comprehensive overview and classification of generative AI models for enzyme design, highlighting models with experimental validation relevant to real-world settings and outlining their respective limitations. We argue that generative AI models now have the maturity to create and optimize enzymes for industrial applications. Wider adoption of generative AI models with experimental feedback loops can speed up the development of biocatalysts and serve as a community assessment to inform the next generation of models.

Biocatalysis↗
PubMed2005

Prospects of second generation artificial intelligence tools in calibration of chemical sensors.

Multivariate data driven calibration models with neural networks (NNs) are developed for binary (Cu++ and Ca++) and quaternary (K+, Ca++, NO3- and Cl-) ion-selective electrode (ISE) data. The response profiles of ISEs with concentrations are non-linear and sub-Nernstian. This task represents function approximation of multi-variate, multi-response, correlated, non-linear data with unknown noise structure i.e. multi-component calibration/prediction in chemometric parlance. Radial distribution function (RBF) and Fuzzy-ARTMAP-NN models implemented in the software packages, TRAJAN and Professional II, are employed for the calibration. The optimum NN models reported are based on residuals in concentration space. Being a data driven information technology, NN does not require a model, prior- or posterior- distribution of data or noise structure. Missing information, spikes or newer trends in different concentration ranges can be modeled through novelty detection. Two simulated data sets generated from mathematical functions are modeled as a function of number of data points and network parameters like number of neurons and nearest neighbors. The success of RBF and Fuzzy-ARTMAP-NNs to develop adequate calibration models for experimental data and function approximation models for more complex simulated data sets ensures AI2 (artificial intelligence, 2nd generation) as a promising technology in quantitation.

Artificial Intelligence↗
PubMed2026

ELISA (Embedding-Linked Interactive Single-cell Agent): an interpretable hybrid generative Artificial Intelligence agent for expression-grounded discovery in single-cell genomics.

Translating single-cell RNA sequencing (scRNA-seq) data into mechanistic biological hypotheses remains a critical bottleneck, as agentic AI systems lack direct access to transcriptomic representations while expression foundation models remain opaque to natural language. Here, we introduce ELISA (Embedding-Linked Interactive Single-cell Agent), an interpretable framework that unifies single-cell generative pretrained transformer expression embeddings with biomedical bidirectional encoder representations from transformers-based semantic retrieval and large-language model (LLM)-mediated interpretation for interactive single-cell discovery. An automatic query classifier routes inputs to gene marker scoring, semantic matching, or reciprocal rank fusion pipelines depending on whether the query is a gene signature, natural language concept, or mixture of both. Integrated analytical modules perform pathway activity scoring across 60+ gene sets, ligand-receptor interaction prediction using 280+ curated pairs, condition-aware comparative analysis, and cell-type proportion estimation, all operating directly on embedded data without access to the original count matrix. Benchmarked across six diverse scRNA-seq datasets spanning inflammatory lung disease, pediatric and adult cancers, organoid models, healthy tissue, and neurodevelopment, ELISA significantly outperforms CellWhisperer, a classical lexical retriever (BM25), and a random baseline in cell type retrieval (combined permutation test, $p < 2\times 10^{-5}$ for each), with particularly large gains on gene-signature queries (Cohen's $d = 5.98$ for mean reciprocal rank). ELISA replicates published biological findings (mean composite score 0.88), and generates candidate hypotheses through grounded LLM reasoning, bridging the gap between transcriptomic data exploration and biological discovery.

Generative Artificial Intelligence↗
PubMed2026

Toward life with a 19-amino acid alphabet through generative artificial intelligence design.

Because all known living organisms are made from at least 20 canonical amino acids, the feasibility of life using a more simplified alphabet remains unclear. In this work, we leveraged computational design and synthetic biology to explore building a cell from a 19-amino acid alphabet. Initial analyses suggested that isoleucine (Ile) may be dispensable, which we confirmed by directly replacing Ile residues in essential proteins in Escherichia coli. Critically, protein language models and structure-based models were necessary to redesign functional Ile-less proteins in most cases. We systematically replaced all 382 Ile residues from the ribosome and combined 21 redesigned subunits at a native genomic locus to produce a viable, evolutionarily stable cell. This work provides a roadmap to create the first 19-amino acid organism since early evolution.

Escherichia coli↗
PubMed2026

A safety-centric perspective on innovation and risk in the use of artificial intelligence in genomics.

Adopting a safety-centric approach, this article explores how generative artificial intelligence (AI), and more specifically, foundation models for biological sequences, can exacerbate data quality issues, technical biases, and dual-use potential, particularly in critical applications such as clinical genetics, precision medicine, and pathogen engineering. This work centres on how misuse risks emerge throughout the innovation pipeline and how these intersect with the growing accessibility of generative genomic models. Particular attention is given to dual-use governance and infrastructure hardening in sequence analysis workflows. The work aims to provide scientists, regulators, and policymakers with a toolkit to discuss beneficial innovation in genomic AI while maintaining robust safeguards against harm and misuse.

Genomics↗
PubMed2026

Emulated trial of artificial intelligence use and subsequent depressive outcomes in a survey of US adults.

BACKGROUND: Generative artificial intelligence (AI) use has been suggested to have adverse mental health consequences but a causal relationship has not been examined. OBJECTIVE: To simulate a randomised controlled trial of AI use in a work, school or personal context by applying target trial emulation to multiple waves of data from a nationally representative survey. METHODS: We conducted a target trial emulation using non-probability survey data from three waves of a nationally representative survey conducted between 18 June 2024 and 8 January 2025. Participants aged &#x2265;18 years reported generative AI use frequency at baseline. High-frequency use was defined as multiple times per week or more. The primary outcome was depressive symptom severity measured using the Patient Health Questionnaire 9-item (PHQ-9) at follow-up. Generalised causal forests assessed heterogeneity of treatment effects. FINDINGS: Among 19&#x2009;099 participants assessed at baseline, 2862 (15.0%) reported AI use at least multiple times per week. A subset of 3109 (16.3%) returned for follow-up. In the primary weighted analysis, high-frequency use was not significantly associated with change in PHQ-9 score at follow-up (mean difference -0.18, 95% CI -0.94 to 0.59; p=0.65). Multiple sensitivity analyses using alternate outcome definitions also did not identify significant causal effects. Generalised causal forests yielded no significant evidence of heterogeneity of effect (p=0.81). CONCLUSIONS: In an emulated randomised trial among US adults, generative AI use was not associated with subsequent depressive symptoms. This result does not support the premise that AI use causes greater depressive symptoms, although adverse outcomes among vulnerable individuals cannot be excluded. CLINICAL IMPLICATIONS: AI use is unlikely to cause increased depressive symptoms among most US adults. Continued monitoring should clarify potential risks among vulnerable populations.

Humans↗
PubMed2025

Generative AI Models in Time-Varying Biomedical Data: Scoping Review.

BACKGROUND: Trajectory modeling is a long-standing challenge in the application of computational methods to health care. In the age of big data, traditional statistical and machine learning methods do not achieve satisfactory results as they often fail to capture the complex underlying distributions of multimodal health data and long-term dependencies throughout medical histories. Recent advances in generative artificial intelligence (AI) have provided powerful tools to represent complex distributions and patterns with minimal underlying assumptions, with major impact in fields such as finance and environmental sciences, prompting researchers to apply these methods for disease modeling in health care. OBJECTIVE: While AI methods have proven powerful, their application in clinical practice remains limited due to their highly complex nature. The proliferation of AI algorithms also poses a significant challenge for nondevelopers to track and incorporate these advances into clinical research and application. In this paper, we introduce basic concepts in generative AI and discuss current algorithms and how they can be applied to health care for practitioners with little background in computer science. METHODS: We surveyed peer-reviewed papers on generative AI models with specific applications to time-series health data. Our search included single- and multimodal generative AI models that operated over structured and unstructured data, physiological waveforms, medical imaging, and multi-omics data. We introduce current generative AI methods, review their applications, and discuss their limitations and future directions in each data modality. RESULTS: We followed the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines and reviewed 155 articles on generative AI applications to time-series health care data across modalities. Furthermore, we offer a systematic framework for clinicians to easily identify suitable AI methods for their data and task at hand. CONCLUSIONS: We reviewed and critiqued existing applications of generative AI to time-series health data with the aim of bridging the gap between computational methods and clinical application. We also identified the shortcomings of existing approaches and highlighted recent advances in generative AI that represent promising directions for health care modeling.

Artificial Intelligence↗
PubMed2026

Antibody-drug conjugates in selected solid tumours: a position statement update based on findings from the third workshop held by the ETOP IBCSG Partners Foundation.

The European Thoracic Oncology Platform (ETOP) International Breast Cancer Study Group (IBCSG) Partners Foundation initiated a series of workshops for experts to review current evidence and offer recommendations to guide future antibody-drug conjugate (ADC) research. Here, we summarise key findings from the third workshop, which included experts in various solid tumours, basic/translational research scientists and pharmaceutical industry representatives. Recent positive phase III trial data have further incorporated ADCs into the standard of care [e.g. lung: sacituzumab tirumotecan; breast: trastuzumab deruxtecan (T-DXd), sacituzumab govitecan, datopotamab deruxtecan; muscle-invasive bladder cancer: enfortumab vedotin; ovarian cancer: mirvetuximab soravastine; cervical cancer: tisotumab vedotin]. Thus, research priorities must be tailored according to tumour type, potentially focussing initially on settings where ADCs could replace chemotherapy. Many phase III ADC trials have been initiated based on positive phase I data and although these trials are larger than those conducted historically, prespecified criteria (e.g. patient numbers and magnitude of efficacy) should be met to justify proceeding directly to phase III. Importantly, although several ADCs have been successfully developed without mandatory biomarker selection, biomarker-driven ADC development enables rational patient selection, as illustrated by multiple ADCs (e.g. T-DXd, mirvetuximab soravtansine and telisotuzumab vedotin). The identification, development and validation of predictive biomarkers are therefore essential, particularly given several critical nuances, including the algorithms used to assess biomarker status and the type of specimen analysed, all of which may be influenced by temporal and spatial heterogeneity. Additional ADC research priorities include the optimisation of ADC constructs to enhance efficacy/tolerability and the identification of reliable ADC targets, including work to elucidate attributes of already-identified targets. Finally, considering the vast amount of ADC-related data being generated, artificial intelligence could be leveraged to analyse combined datasets and generate composite biomarkers, including tumour histology, optimal target expression thresholds, molecular alterations and activated pathways affecting payload activity and target function, to accelerate research.

Humans↗
PubMed2026

Effectiveness of artificial intelligence in nursing simulation education: A systematic review, meta-analysis and bibliometric visualization analysis.

OBJECTIVES: To synthesize the roles and core functions of AI in nursing simulation education for nursing students via systematic review, quantitatively evaluate its effects on students' knowledge and skill outcomes through meta-analysis, and map the research landscape and development trends of this field through bibliometric visualization analysis. DESIGN: Systematic review, meta-analysis and bibliometric visualization analysis. DATA SOURCES: Eight electronic databases: PubMed, Web of Science, MEDLINE, ERIC, Academic Search Complete, China National Knowledge Infrastructure (CNKI), Wanfang Database, VIP Chinese Science and Technology Journal Database (VIP) were employed to search studies from the time of construction to 16 December 2025. REVIEW METHODS: Studies meeting the inclusion criteria were screened. The revised Cochrane Risk of Bias tool (ROB 2) and Joanna Briggs Institute (JBI) critical appraisal checklists were used for quality assessment. Meta-analysis was performed with Review Manager 5.4, and bibliometric visualization analysis was conducted using VOSviewer 1.6.20 and Bibliometrix (based on R4.4.3). RESULTS: A total of 61 studies were included. AI primarily played two roles in nursing simulation education: peer-type new subject (n&#xa0;=&#xa0;24) and direct mediator (n&#xa0;=&#xa0;22). Meta-analysis showed that AI interventions significantly improved nursing students' knowledge (SMD&#xa0;=&#xa0;1.49, 95% CI [0.55,2.43], p&#xa0;=&#xa0;0.002) and skills (SMD&#xa0;=&#xa0;0.66, 95% CI [0.02,1.31], p&#xa0;=&#xa0;0.04). Bibliometric analysis identified that the United States of America and China were the two main contributing countries in this field, and the key motor themes included generative artificial intelligence, virtual patients, and geriatric care. CONCLUSIONS: AI exerts positive effects on nursing students' knowledge acquisition and skill enhancement in simulation education, with peer-type new subject and direct mediator as the dominant roles. Future research should focus on expanding AI applications in multi-specialty simulation scenarios, activating the data-driven value of machine learning, and strengthening international collaboration and standardization construction, so as to promote the sustainable development of AI-integrated nursing simulation education.

Humans↗
PubMed2025

Integrating expert knowledge into large language models improves performance for psychiatric reasoning and diagnosis.

BACKGROUND AND METHODS: The authors sought to evaluate the performance of common large language models (LLMs) in psychiatric diagnosis, and the impact of integrating expert-derived reasoning on their performance. Clinical case vignettes and associated diagnoses were retrieved from the DSM-5-TR Clinical Cases book. Diagnostic decision trees were retrieved from the DSM-5-TR Handbook of Differential Diagnosis and refined for LLM use. Three LLMs were prompted to provide diagnosis candidates for the vignettes either by directly prompting or using the decision trees. These candidates and diagnostic categories were compared against the correct diagnoses. The positive predictive value (PPV), sensitivity, and F1 statistic were used to measure performance. RESULTS: When directly prompted to predict diagnoses, the best LLM by F1 statistic (gpt-4o) had sensitivity of 76.7 % and PPV of 40.4 %. When making use of the refined decision trees, PPV was significantly increased (65.3 %) without a significant reduction in sensitivity (70.9 %). Across all experiments, the use of the decision trees statistically significantly increased the PPV, significantly increased the F1 statistic in 5/6 experiments, and significantly reduced sensitivity in 4/6 experiments. DISCUSSION: When used to predict psychiatric diagnoses from case vignettes, direct prompting of the LLMs yielded most true positive diagnoses but had significant overdiagnosis. Integrating expert-derived reasoning into the process using decision trees improved LLM performance (as measured by F1 statistic), primarily by suppressing overdiagnosis with a lower-magnitude negative impact on sensitivity. This suggests that the integration of clinical expert-derived reasoning could improve the performance of LLM-based tools in the behavioral health setting.

Humans↗
PubMed2026

AI-generated familiarity estimates are a useful new source of information about word knowledge in Simplified Chinese.

This study evaluated the usefulness of AI-generated estimates of word familiarity for predicting word difficulty in Simplified Chinese, building on previous research in alphabetic languages. We found that familiarity estimates produced using large language models (LLMs) showed moderate-to-strong correlations with human familiarity ratings. These LLM estimates were the most effective predictors of both word naming and lexical decision times, surpassing traditional metrics such as word frequency and human familiarity ratings, while the latter still provided modest, non-overlapping variance. GPT-4o with English instructions produced superior results compared to the Chinese-centered models currently available. The results imply that LLM familiarity estimates are a valuable resource for Chinese psycholinguistics, supporting work across experimental design, modeling, and norming. We release familiarity estimates for 27,624 words for unrestricted research and educational use.

Humans↗
PubMed2026

Upcycling Vegetable Waste Into Functional Food Ingredients via Synergistic Microbial Engineering and Artificial Intelligence.

The escalating generation of global vegetable waste represents a critical loss of bioactive resources, necessitating a paradigm shift from passive disposal to active nutrient upcycling. However, the industrial conversion of this heterogeneous biomass into standardized functional food ingredients is currently impeded by significant techno-economic barriers, primarily structural recalcitrance, compositional inconsistency, and the presence of toxic fermentation inhibitors. This review provides a comprehensive analysis of the synergistic application of microbial engineering and artificial intelligence (AI) to resolve these bioprocessing bottlenecks within a food-to-food closed-loop framework (as shown in the graphical abstract). We evaluate recent advances in engineering food-grade microbial chassis (e.g., Saccharomyces cerevisiae and Escherichia coli) to enhance lignocellulose degradation and stress tolerance. Concurrently, we examine the integration of AI across the entire value chain, covering deep learning-based rational enzyme design, genome-scale metabolic modeling, and intelligent process control for precision fermentation. Current evidence demonstrates that the hardware-software coupling of engineered strains and AI algorithms significantly enhances conversion efficiency and process robustness. Key findings highlight that AI-driven Design-Build-Test-Learn cycles facilitate the de novo creation of enzymes with superior kinetics and strains with adaptive stress response capabilities against toxins. Moreover, dynamic digital twin models effectively mitigate the impact of substrate variability, ensuring the batch-to-batch consistency required for food applications. We conclude that this data-driven synergistic paradigm is pivotal for establishing a resilient circular bioeconomy, enabling the reliable bioconversion of waste into high-value single-cell proteins, natural flavor additives, and sustainable packaging materials.

Artificial Intelligence↗
PubMed2026

Evaluation of a cornea-specialized large language model for diagnostic and management accuracy in complex corneal cases.

PURPOSE: To evaluate whether a cornea-specialized large language model (LLM) enhanced with retrieval-augmented generation (RAG) improves clinicians' diagnostic and management accuracy in complex corneal cases compared to a general-purpose GPT-4o model and unaided clinician performance. METHODS: This prospective, randomized, masked evaluation study involved three cornea trainees who each independently reviewed 39 real-world corneal cases under three experimental conditions: unaided, GPT-4o-assisted, and assisted by a cornea-specialized GPT-4o model. The cornea-specialized model was constructed by embedding over 200 publicly available Wikipedia articles into GPT-4o's RAG framework. Participants provided open-ended diagnoses and selected the next-step management options (multiple choice). They were allowed up to three GPT-4o queries per case, and the AI-assisted arms were randomized to minimize bias. Accuracy for both tasks was compared against expert reference standards using McNemar's test. RESULTS: Diagnostic accuracy was 48.7%, 20.5%, and 38.5% unaided, improving to 69.2%, 46.2%, and 59.0% with general GPT-4o (p<0.04). The cornea-specialized GPT-4o further improved accuracy to 71.8%, 48.7%, and 74.4%, with improvements over unaided performance for all clinicians (p<0.01). For next-step decisions, unaided accuracy was 76.9%, 87.2%, and 59.0%. With the specialized model, Ophthalmologist 3 improved to 71.8% (p<0.05), Ophthalmologist 1 remained high at 82.1%, and Ophthalmologist 2 declined to 64.1% (p<0.05). CONCLUSIONS: A cornea-specialized LLM enhanced with RAG improved diagnostic accuracy in complex corneal cases, particularly among clinicians with lower baseline performance. Effects on management accuracy were inconsistent. Future studies should explore the use of open-ended management tasks and examine whether smaller, curated retrieval corpora yield better model performance.

Humans↗
PubMed2026

Structure and evolution-guided design of minimal RNA-guided nucleases.

The design of RNA-guided nucleases with properties not limited by evolution can expand programmable genome-editing capabilities. However, generating diverse multidomain proteins with robust enzymatic properties remains challenging. Here, we use a protein design strategy that couples a structure-guided inverse-folding model with evolution-informed residue constraints to generate active, divergent variants of TnpB, a minimal CRISPR-Cas12-like nuclease, termed SynTnpBs. High-throughput screening of artificial intelligence-generated variants yielded editors that retained or exceeded wild-type activity in bacterial, plant, and human cells. Cryo-electron microscopy-based structure determination of the most divergent variant revealed stabilizing contacts in the RNA-DNA interfaces across conformations, demonstrating the design potential of this approach. Together, these results establish a strategy for creating non-natural RNA-guided nucleases and conformationally active nucleic acid binders, enlarging the designable protein space.

Humans↗
PubMed1998

Development of a health promotion system for the elderly: Committee of Health Evaluation for Elderly Persons Council of Japan AMHTS Institutions.

A new health promotion approach for elderly persons is required which maintains not only their physical health but also their quality of life. We are developing a health promotion system which makes use of questionnaires dealing with physical conditions and quality of life, and makes health reports. Health evaluation is carried out in three steps. First, detailed information about the physical health of each client is collected. Second, quality of life is evaluated according to five health indicators. Last, health recommendations are generated. An artificial intelligence (AI) program produces detailed questions to collect necessary information for the evaluation of a client's health. The information related to quality of life is converted into five health indicators and presented as a radar-chart in documents and displays. The knowledge-based AI program automatically generates the health recommendation documents. This information is available for physicians and nurses for health counseling.

Aged↗
PubMed1997

Basic Science in Medical Reasoning: An Artificial Intelligence Approach.

A new generation of intelligent systems is growing up in the community of Artificial Intelligence in Medicine. The main goal of these systems is the representation and use of real theory of diseases, as they are represented in medical textbooks or in scientific articles, rather than the heuristic shortcuts of human experts. In this paper, we will argue that the difficulties in the integration of basic science and clinical knowledge in intelligent systems arise from ontological differences between these kinds of knowledge and that the solution can be found in their dynamic integration during the reasoning process. In order to illustrate this point, we will first describe an epistemological analysis of the interplay between basic science knowledge and clinical knowledge, and then we will provide the example of a computational architecture implementing this view.

Journal Article↗