SEARCH · PubMed Health

Results for “data mining”

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 55 records · Page 3Linked to original sources

PubMed2025

Systematic mining and quantification reveal the dominant contribution of non-HLA variations to acute graft-versus-host disease.

Human leukocyte antigen (HLA) disparity between donors and recipients is a key determinant triggering intense alloreactivity, leading to a lethal complication, namely, acute graft-versus-host disease (aGVHD), after allogeneic transplantation. Moreover, aGVHD remains a cause of mortality after HLA-matched allogeneic transplantation. Protocols for HLA-haploidentical hematopoietic cell transplantation (haploHCT) have been established successfully and widely applied, further highlighting the urgency of performing panoramic screening of non-HLA variations correlated with aGVHD. On the basis of our time-consecutive large haploHCT cohort (with a homogenous discovery set and an extended confirmatory set), we first delineated the genetic landscape of 1366 samples to quantitatively model aGVHD risk by assessing the contributions of HLA and non-HLA genes together with clinical factors. In addition to identifying multiple loss-of-function (LoF) risk variations in non-HLA coding genes, our data-driven study revealed that non-HLA genetic variations, independent of HLA disparity, contributed the most to the occurrence of aGVHD. This unexpected major effect was verified in an independent cohort that received HLA-identical sibling HCT. Subsequent functional experiments further revealed the roles of a representative non-HLA LoF gene and LoF gene pair in regulating the alloreactivity of primary human T cells. Our findings highlight the importance of non-HLA genetic risk in the new era of transplantation and propose a new direction to explore the immunogenetic mechanism of alloreactivity and to optimize donor selection strategies for allogeneic transplantation.

Humans↗

PubMed2025

CAGEcleaner: reducing genomic redundancy in gene cluster mining.

SUMMARY: Mining homologous biosynthetic gene clusters (BGCs) typically involves searching colocalised genes against large genomic databases. However, the high degree of genomic redundancy in these databases often propagates into the resulting hit sets, complicating downstream analyses and visualization. To address this challenge, we present CAGEcleaner, a Python-based pipeline with auxiliary bash scripts designed to reduce redundancy in gene cluster hit sets by dereplicating the genomes that host these hits. CAGEcleaner integrates seamlessly with widely used gene cluster mining tools, such as cblaster and CAGECAT, enabling efficient filtering and streamlining BGC discovery workflows. AVAILABILITY AND IMPLEMENTATION: Source code and documentation is hosted at GitHub (https://github.com/LucoDevro/CAGEcleaner) and Zenodo (https://doi.org/10.5281/zenodo.14726119) under an MIT license. For accessibility, CAGEcleaner is installable from Bioconda (https://anaconda.org/bioconda/cagecleaner) and PyPi (https://pypi.org/project/cagecleaner/), and is also available as a Docker image from DockerHub (https://hub.docker.com/r/lucodevro/cagecleaner).

Software↗

PubMed2026

GAMMA: gap-aware motif mining under incomplete labeling with applications to MHC motifs.

MOTIVATION: Sequence motif identification is crucial for understanding molecular recognition, particularly in immune responses involving peptide binding to major histocompatibility complex (MHC) Class I molecules for antigen presentation to T cells. Traditionally, MHC Class I binding motifs are assumed to be contiguous and span nine amino acids. However, structural evidence suggests that binding may involve nonadjacent residues, challenging the assumptions of existing methods. RESULTS: In this study, we propose Gap-Aware Motif Mining Algorithm (GAMMA), a probabilistic framework designed to identify noncontiguous motifs under conditions of incomplete labeling. GAMMA employs Bayesian inference with Markov chain Monte Carlo sampling to jointly estimate motif parameters, binding locations, and the relative spacing between binding positions. Through extensive simulations and real-world applications to MHC Class I peptide datasets, GAMMA outperforms existing motif discovery tools such as GLAM2 in accurately localizing binding residues and identifying the underlying motifs. Notably, our results suggest that the true number of binding residues may be eight, fewer than the commonly assumed nine. In addition, for longer peptides, the model captures increased flexibility in the central region, consistent with structural observations that peptides may bulge in the middle. AVAILABILITY AND IMPLEMENTATION: The raw data and the source codes are available on GitHub (https://github.com/RanLIUaca/GAMMAmotif).

Amino Acid Motifs↗

PubMed2026

Mining Stored-Specimen Studies for Information about Cancer Natural History.

The advent of new multicancer early detection tests and publication of early diagnostic results have generated expectations of clinical benefit from multicancer screening. The clinical benefit of a cancer screening test depends critically on disease natural history, which is typically learned from prospective screening studies. Retrospective studies of stored blood specimens are important in learning about a test's preclinical diagnostic performance but have rarely been used to infer natural history. The extent to which these studies might be harnessed to also learn natural history is discussed in the context of an article in this issue that infers the combined natural history of a range of cancers targeted by a multicancer early detection test using a case-control subsample of specimens from a large cohort study. The critical question concerns the identifiability of key transition rates in multistate models of natural history alongside state-specific sensitivities. The article suggests that these parameters are estimable within a Bayesian framework that leverages prior information about test sensitivity from diagnostic studies. We offer a heuristic discussion of identifiability in this setting and encourage formal study to determine the extent to which models with varying degrees of complexity may be learned from stored-specimen studies. See related article by Dai et al., p. 1535.

Humans↗

PubMed2026

Integration of genome mining and HiTES reveals secondary metabolic potential in marine-derived Aspergillus sp. WHUF0304.

AIMS: Marine-derived Aspergillus species are prolific producers of bioactive secondary metabolites, yet the majority of their biosynthetic gene clusters (BGCs) remain silent. This study aimed to integrate genome mining with high-throughput elicitor screening (HiTES) to unlock the metabolic potential of Aspergillus sp. WHUF0304 and identify elicitors that promote the accumulation of previously undetected metabolites. METHODS AND RESULTS: A high-quality genome of Aspergillus sp. WHUF0304 was assembled and annotated using multiple functional databases, revealing substantial secondary metabolic potential. antiSMASH analysis identified diverse BGCs, including NRPS/indole-related clusters potentially associated with indole diketopiperazine biosynthesis. A HiTES-inspired elicitor screening strategy was then applied to evaluate 42 small molecules for their ability to alter the metabolite profile of this strain. Among the tested elicitors, fluconazole was identified as the optimal inducer, triggering the production of several indole diketopiperazine-related differential metabolites. Subsequent activity-guided isolation led to the identification of a bioactive indole diketopiperazine dimer, cristatumin E, which exhibited antibacterial activity against Escherichia coli and Bacillus subtilis with minimum inhibitory concentrations (MICs) of 32 µg mL-1 and 256 µg mL-1, respectively. CONCLUSIONS: These findings demonstrate that integrating genomic and functional approaches effectively activates silent BGCs in marine fungi. The fluconazole-associated accumulation and subsequent isolation of cristatumin E, a bioactive indole diketopiperazine dimer, highlight the potential of elicitor-mediated activation to expand the detectable metabolite profile of Aspergillus sp. WHUF0304.

Aspergillus↗

PubMed1986

An approach to the characterization of silica exposure in U.S. industry.

Quantitative evaluation of worker exposure to silica in nine Standard Industrial Classification (SIC) codes was conducted, using data derived from OSHA compliance inspections, in order to assess the silica exposure problem in the U.S. The nine SICs studied were those in which OSHA inspections were concentrated. They include: construction; chemical manufacture; stone, glass, and clay manufacturing; primary metal industries; metal fabrication; machinery; transportation; and miscellaneous manufacturing industries. High exposures to silica were documented in each industry, with the number of test samples over the permissible exposure limit ranging from 14% (aluminum foundries) to 73% (pottery). An estimation is made that 24,889 workers employed in ferrous and nonferrous foundries are at risk of silica-related pulmonary effects. The data developed in this analysis also indicate the need to investigate certain industries that had high exposures but few inspections. The limitations of the data base for estimating the scope of the silica problem, including lack of data on mining and milling, are discussed. We conclude that exposure to silica represents a continuing and significant problem in a number of U.S. industries.

Air Pollutants, Occupational↗

PubMed2026

PubMind: literature-based genetic variant extraction and functional annotation using large language models.

Biomedical literature contains extensive functional knowledge on genetic variants, but much remains inaccessible in unstructured text. Existing resources such as ClinVar and HGMD remain limited by coverage, submission bias, update frequency, and sparse annotation. We develop PubMind, an artificial intelligence (AI) framework that uses large language models (LLMs) to triage and extract variant-function-disease associations and supporting evidence from biomedical text. PubMind captures single-nucleotide, copy-number, structural, and gene-fusion variants, and normalizes records to genomic and transcriptomic coordinates. Benchmarking shows >90% accuracy for variant recognition and 99% precision for disease extraction. Applied to >41 million PubMed abstracts and >5 million full-text articles, PubMind generates PubMind-DB, a database of ~1.3 million unique variants with contextual annotations, accessible via web interface and API. Only ~10% of PubMind variants overlap with ClinVar, and >80% of them show concordant pathogenicity labels. PubMind transforms unstructured biomedical text into structured genomic knowledge, advancing variant interpretation for precision medicine.

Large Language Models↗

PubMed2025

Lit-OTAR framework for extracting biological evidences from literature.

SUMMARY: The lit-OTAR framework, developed through a collaboration between Europe PMC and Open Targets, leverages deep learning to revolutionize drug discovery by extracting evidence from scientific literature for drug target identification and validation. This novel framework combines named entity recognition for identifying gene/protein (target), disease, organism, and chemical/drug within scientific texts, and entity normalization to map these entities to databases like Ensembl, Experimental Factor Ontology, and ChEMBL. Continuously operational, it has processed over 39 million abstracts and 4.5 million full-text articles and preprints to date, identifying more than 48.5 million unique associations that significantly help accelerate the drug discovery process and scientific research >29.9 m distinct target-disease, 11.8 m distinct target-drug, and 8.3 m distinct disease-drug relationships. AVAILABILITY AND IMPLEMENTATION: The results are accessible through Europe PMC's SciLite web app (https://europepmc.org/) and its annotations API (https://europepmc.org/annotationsapi), as well as via the Open Targets Platform (https://platform.opentargets.org/). The daily pipeline is available at https://github.com/ML4LitS/otar-maintenance, and the Open Targets ETL processes are available at https://github.com/opentargets.

Drug Discovery↗

PubMed2025

AutoPM3: enhancing variant interpretation via LLM-driven PM3 evidence extraction from scientific literature.

MOTIVATION: Rare diseases affect over 300 million people worldwide and are often caused by genetic variants. While variant detection has become cost-effective, interpreting these variants-particularly collecting literature-based evidence like ACMG/AMP PM3-remains complex and time-consuming. RESULTS: We present AutoPM3, a method that automates PM3 evidence extraction from literatures using open-source large language models (LLMs). AutoPM3 combines a Text2SQL-based variant extractor and a retrieval-augmented generation (RAG) module, enhanced by a variant-specific retriever and fine-tuned LLM, to separately process tables and text. We curated PM3-Bench, a dataset of 1027 variant-publication evidence pairs from ClinGen. On openly accessible pairs, AutoPM3 achieved 86.1% accuracy for variant hits and 72.5% recall for in trans variants-outperforming other methods, including those using larger models. We uncovered the effectiveness of AutoPM3's key modules, especially for variant-specific retriever and Text2SQL, through the sequential ablation study. AutoPM3 located evidence in 76 s, demonstrating that open-source LLMs can offer an efficient, cost-effective solution for rare disease diagnosis. AVAILABILITY AND IMPLEMENTATION: AutoPM3 is implemented and freely available under the MIT license at https://github.com/HKU-BAL/AutoPM3.

Genetic Variation↗

PubMed2026

Analysis of the molecular mechanism underlying di(2-ethylhexyl) phthalate-induced bladder carcinogenesis via network toxicology and molecular docking approaches: An observational study.

This study aims to investigate the toxicity of di(2-ethylhexyl) phthalate (DEHP) and the potential molecular mechanisms of DEHP-induced bladder cancer (BLCA) using network toxicology and molecular docking strategies. The toxicity of DEHP was assessed using Prox-II software, and potential targets for DEHP-induced BLCA were identified by integrating data from ChEMBL database, Search Tool for Interactions of Chemicals, SwissTargetPrediction, GeneCards, Therapeutic Target Database, Online Mendelian Inheritance in Man, and The Cancer Genome Atlas. STRING database and Cytoscape were employed to construct target networks and determine core targets. The expression levels of core targets were analyzed using R. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analyses were performed on potential and core targets. Molecular docking was carried out using CB-Dock 2 to verify the interactions between DEHP and core targets. A total of 105 potential targets related to DEHP-induced BLCA were identified, from which 7 core targets were selected: cyclin-dependent kinase 1, interleukin 6, cyclin-dependent kinase 2, cyclin B1, Erb-B2 receptor tyrosine kinase 2, cyclin B2, and B-cell lymphoma 2. IL-6 and B-cell lymphoma 2 showed downregulated expression in tumor tissues, while cyclin-dependent kinase 1, cyclin-dependent kinase 2, cyclin B1, Erb-B2 receptor tyrosine kinase 2, and cyclin B2 were upregulated. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes enrichment analyses indicated that these targets were enriched in cell signaling and cancer-related pathways. Molecular docking confirmed that DEHP interacts with these core targets. DEHP may promote the development of BLCA by interacting with key proteins and signaling pathways. This study provides a theoretical basis for understanding the molecular mechanisms of DEHP-induced BLCA and offers references for future prevention and treatment strategies.

Diethylhexyl Phthalate↗

PubMed2024

Shared etiology of Mendelian and complex disease supports drug discovery.

BACKGROUND: Drugs targeting disease causal genes are more likely to succeed for that disease. However, complex disease causal genes are not always clear. In contrast, Mendelian disease causal genes are well-known and druggable. Here, we seek an approach to exploit the well characterized biology of Mendelian diseases for complex disease drug discovery, by exploiting evidence of pathogenic processes shared between monogenic and complex disease. One way to find shared disease etiology is clinical association: some Mendelian diseases are known to predispose patients to specific complex diseases (comorbidity). Previous studies link this comorbidity to pleiotropic effects of the Mendelian disease causal genes on the complex disease. METHODS: In previous work studying incidence of 90 Mendelian and 65 complex diseases, we found 2,908 pairs of clinically associated (comorbid) diseases. Using this clinical signal, we can match each complex disease to a set of Mendelian disease causal genes. We hypothesize that the drugs targeting these genes are potential candidate drugs for the complex disease. We evaluate our candidate drugs using information of current drug indications or investigations. RESULTS: Our analysis shows that the candidate drugs are enriched among currently investigated or indicated drugs for the relevant complex diseases (odds ratio = 1.84, p = 5.98e-22). Additionally, the candidate drugs are more likely to be in advanced stages of the drug development pipeline. We also present an approach to prioritize Mendelian diseases with particular promise for drug repurposing. Finally, we find that the combination of comorbidity and genetic similarity for a Mendelian disease and cancer pair leads to recommendation of candidate drugs that are enriched for those investigated or indicated. CONCLUSIONS: Our findings suggest a novel way to take advantage of the rich knowledge about Mendelian disease biology to improve treatment of complex diseases.

Humans↗

PubMed2026

Benchmarking large language models for extracting biobank-derived insights into health and disease.

Biobank-scale datasets such as the UK Biobank have become foundational resources for advancing biomedical discovery. Yet the complexity and heterogeneity of these resources, spanning genomics, imaging, clinical records, and metadata, pose substantial barriers to access and interpretation. Large Language Models (LLMs) offer a promising avenue for making such datasets more navigable through natural language interfaces. However, the extent to which current general-purpose LLMs can retrieve and synthesize biobank-specific insights has not yet been systematically evaluated. In this study, we present a reproducible, multi-metric evaluation framework to benchmark the capabilities of leading LLMs. We evaluated six leading large language models: Gemini 3 Pro, Claude Opus 4.5, Claude Sonnet 4.5, GPT-5.2, Mistral Large 2, and DeepSeek V3, on four benchmark tasks designed to assess biobank-related knowledge retrieval. We evaluate model performance across six dimensions (semantic accuracy, factual correctness, domain knowledge, reasoning quality, response depth, and biobank specificity) and assessed output consistency using curated UK Biobank references and a robust random baseline. All models outperformed the baseline by 2× to 3× , with strong statistical separation (p < 0.001), confirming meaningful biobank-specific knowledge retrieval. Gemini 3 Pro achieved the highest overall accuracy across tasks such as keyword synthesis, institution recognition, and topic inference, while Claude Sonnet 4.5 demonstrated the most uniform performance across evaluation dimensions. Our benchmark provides a rigorous framework for evaluating LLMs in biomedical settings. Using the UK Biobank as a real-world testbed, we highlight both the capabilities and limitations of current models, measuring their capacity to recall structured biomedical knowledge consistent with authoritative biobank metadata.

Large Language Models↗

PubMed2024

Diverse RNA viruses discovered in multiple seagrass species.

Seagrasses are marine angiosperms that form highly productive and diverse ecosystems. These ecosystems, however, are declining worldwide. Plant-associated microbes affect critical functions like nutrient uptake and pathogen resistance, which has led to an interest in the seagrass microbiome. However, despite their significant role in plant ecology, viruses have only recently garnered attention in seagrass species. In this study, we produced original data and mined publicly available transcriptomes to advance our understanding of RNA viral diversity in Zostera marina, Zostera muelleri, Zostera japonica, and Cymodocea nodosa. In Z. marina, we present evidence for additional Zostera marina amalgavirus 1 and 2 genotypes, and a complete genome for an alphaendornavirus previously evidenced by an RNA-dependent RNA polymerase gene fragment. In Z. muelleri, we present evidence for a second complete alphaendornavirus and near complete furovirus. Both are novel, and, to the best of our knowledge, this marks the first report of a furovirus infection naturally occurring outside of cereal grasses. In Z. japonica, we discovered genome fragments that belong to a novel strain of cucumber mosaic virus, a prolific pathogen that depends largely on aphid vectoring for host-to-host transmission. Lastly, in C. nodosa, we discovered two contigs that belong to a novel virus in the family Betaflexiviridae. These findings expand our knowledge of viral diversity in seagrasses and provide insight into seagrass viral ecology.

RNA Viruses↗

PubMed1987

Pathological findings in mine workers: II. Quality of the PATHAUT data.

To assess the feasibility of using the pathology automation system (PATHAUT) for research, the quality of the data was explored by examining who comes to autopsy, the quality of the autopsy material, interobserver variability, and repeatability of diagnoses. The data indicated that autopsy rates in the gold mining industry, especially for whites, are high and that even among blacks, gold miners are represented in proportions exceeding the relative size of the working population. Because of the perception of the autopsy service as a means of obtaining compensation, miners with occupational diseases fully compensated in life are probably underrepresented. The autopsy material submitted for full autopsy is generally better preserved than cardiorespiratory organs that are sent for examination. The gold mining industry has a high proportion of full autopsies as does the Iron and Steel Corporation of South Africa. Full autopsies are more commonly performed on older deceased miners. This was true for both blacks and whites. The allocation of material to pathologists for full autopsies and examinations of the cardiorespiratory organs were clearly not random, and this may affect comparisons among pathologists. Active tuberculosis, silicosis, and emphysema prevalences appeared fairly comparable across pathologists; however, there was wide variability in the prevalence of bronchiolitis as determined by the pathologists. Agreement between the diagnoses on PATHAUT and reclassifications by single pathologists was very good for the severity of emphysema and the histological type of bronchogenic carcinoma.

Autopsy↗

PubMed1994

Radon update: facts concerning environmental radon: levels, mitigation strategies, dosimetry, effects and guidelines. SNM Committee on Radiobiological Effects of Ionizing Radiation.

The risk from environmental radon levels is not higher now than in the past, when residential exposures were not considered to be a significant health hazard. The majority of the radon dose is not from radon itself, but from short-lived alpha-emitting radon daughters, most notably 218Po(T1/2 3 min) and 214Po (T1/2 0.164 msec) along with beta particles from 214Bi (T1/2 19.7 min). Radon gas can penetrate homes from many sources and in various fashions. Measuring radon in homes is simple and relatively inexpensive and may be accomplished in a variety of ways. Although it is not possible to radon-proof a house, it is possible to reduce the level. In high radon areas, if the average level is higher than 4-8 pCi/liter (NCRP recommended level is 8 pCi/liter; EPA recommended level is 4 pCi/liter), appropriate action is advised. The shape of the dose response curves for miners exposed to alpha-emitting particles in the workplace is consistent with current biologic knowledge. It is linear in the low dose range and saturates in the high dose range. No detectable increase in lung cancer frequency is seen in the lowest exposed miners (those with exposures < 120 WLM, the relevant dose interval for most homes). Evidence for a health effect from radon exposure is based on data from animal studies and epidemiologic studies of mines. Extensive radiobiologic data predict a linear dose-response curve in the low dose region due to poor biological repair mechanisms for the high density of ionizing events that alpha particles create. However, no compelling evidence for increased cancer risks has yet been demonstrated from "acceptable" levels (< 4-8 pCi/liter).

Air Pollutants↗

PubMed1985

Underground mining, smoking, and lung cancer: a case-control study in the iron ore municipalities in northern Sweden.

A case-control study of lung cancer in males was performed in two municipalities in northern Sweden with large iron ore mines. Previous studies had revealed an increased lung cancer risk for underground workers in these mines, with all probability related to radon daughter exposure. Data concerning underground mining and smoking were obtained from questionnaires. All analyses suggested an interaction of a multiplicative type between underground mining and smoking in the causation of lung cancer in this population. The calculated population etiologic fraction was about 45% for underground mining and about 80% for smoking.

Aged↗

PubMed1991

[Physiological and hygienic evaluation of controllers' work in coal mines].

The article contains data on the results of the on-the-spot studies of coal mine controllers' labour conditions and relating hygienic factors. It was established that the work at the dispatcher's control point in coal mines was considerably affected by unfavourable factors (low degree of illumination, constructional shortcomings of the control point and seat) with concomitant neuropsychic stress conditions. Revealed were specific functional shifts in CVS and CNS, neuromuscular disorders and the analyzers' malfunctioning. A set of measures was proposed for labour conditions improvement, ergometric perfection of the working place and reduction of neuroemotional tension.

Administrative Personnel↗

PubMed1980

Baseline levels of selected trace elements in Colorado oil shale region animals.

Baseline levels of boron, fluorine, molybdenum, and copper are described for 18 mule deer (Odocoileus hemionus) and for 45 composite samples of deer mice (Peromyscus maniculatus) from the Piceance Creek Basin, Rio Blanco County, Colorado. These data were collected before oil shale mining took place, and can be used to compare with levels found after mining is initiated. The data can thus be used to monitor changes in levels in animal tissues and as a basis for mitigating possible harmful effects due to the mining. Mean ppm (+/- S.D.) dry basis of each element is presented for selected tissues of each species. Results are also presented by habitat type for deer mice and by age for mule deer. Significant differences (P < 0.05) in molybdenum levels in deer mice were found between habitats. Significant differences (P < 0.05) were found between fawns and adult mule deer for boron levels, but not for the other elements. A need to standardize bone selection for analysis of fluorine was indicated. Kidneys appeared to be the organ of choice for baseline sampling of molybdenum and copper, and livers may be the organ of choice when toxic levels are suspected.

Animals↗