J Occup Environ Med. 2025 Jan 22. doi: 10.1097/JOM.0000000000003324. Online ahead of print.

ABSTRACT

OBJECTIVE: This study aimed to share the process of epidemiologic survey and demonstrate the decisive factors for evaluating work-related Idiopathic interstitial pneumonias (IIPs) in Korea.

METHODS: Data and reports of epidemiologic surveys were investigated and completed during 2010-2022.

RESULTS: A total of 228 idiopathic pulmonary fibrosis (IPF) cases requested for an epidemiologic survey, of which 153 (67.1%) were work-related. The level of exposure intensity was more than intermediate and the exposure duration was more than 10 years in the work-related IPF, so the decisive factors for work-related IPF were exposure substances, intensity and duration.The main factors for work-related acute interstitial pneumonia or cryptogenic organising pneumonia were exposure latency and exposure intensity.

CONCLUSIONS: The results of this study indicated the process of the epidemiologic survey and important factors for evaluating work-related IIPs.

PMID:39875335 | DOI:10.1097/JOM.0000000000003324

Virol J. 2025 Jan 28;22(1):19. doi: 10.1186/s12985-025-02636-7.

ABSTRACT

Infection with Influenza A virus (IAV) induces severe inflammatory responses and lung injury, contributing significantly to mortality and morbidity rates. Alterations in the microbial composition of the lungs and intestinal tract resulting from infection could influence disease progression and treatment outcomes. Xiyanping (XYP) injection has demonstrated efficacy in clinical treatment across various viral infections. However, its specific effects and mechanisms against IAV remain unclear. In this study, we established an IAV infection mice model, and utilized 16 S rRNA sequencing, RNA sequencing, protein chips, and molecular docking, to investigate the mechanisms of XYP injection on altering pulmonary and gut microbiota, and identifying its target sites. We revealed that XYP injection significantly reduced mortality, weight loss, lung viral titers, and lung pathology in IAV-infected mice. XYP injection down-regulated the activity of malondialdehyde, and the levels of interleukin (IL)-1β, IL-5, IL-6, tumor necrosis factor-α, IL-18, IL-15, granulocyte colony-stimulating factor, IL-9, chemokine (C-C motif) ligand-5, and C-X-C motif chemokine ligand 5, while up-regulated the activities of glutathione peroxidase reactive and superoxide dismutase, and the level of interferon-γ. The diversity of the pulmonary and gut microbiota was altered slightly after XYP injection. The linear discriminant analysis of the gut microbes revealed a higher proportion of potentially beneficial bacteria, including Akkermansia, Parabacteroides goldsteinii, Defluviitaleaceae, Oscillospirales, and Eubacterium_coprostanoligenes_group characterized the XYP group. Peritoneal macrophage RNA sequencing highlighted Serpinb2 as the most significantly regulated gene by XYP injection, along with consistent changes in multiple downstream Th2 structure genes. KEGG pathway analysis indicated significant modifications in genes associated with influenza A, mitogen-activated protein kinase signaling, nuclear factor kappa-B signaling, and apoptosis following XYP injection. Finally, human protein chips and molecular docking were carried out to confirm the binding of the main component of XYP injection, andrographolide, with SERPINB2/PAI-2 protein. Overall, our study provides valuable insights into the therapeutic potential of XYP injection in treating influenza, highlighting its multifaceted effects on host microbiota and immune responses, and pinpointing SerpinB2/PAI-2 as the target for XYP injection in exerting anti-inflammatory and antiviral therapeutic mechanisms.

PMID:39875956 | DOI:10.1186/s12985-025-02636-7

BMC Mol Cell Biol. 2025 Jan 28;26(1):7. doi: 10.1186/s12860-025-00532-0.

ABSTRACT

This paper describes a method for determining the cytotoxicity of chemical compounds based on the detection of fluorescent proteins-in this case, green fluorescent protein (GFP) and red fluorescent protein (RFP), which are released into the medium from dead cells. This method is similar in principle to the lactate dehydrogenase test (LDH test), but it does not require a reaction with a chromogenic substrate. This method also makes it possible to independently determine the viability of different lines when used in cocultures. Experiments were performed on a classical monolayer, spheroids and 3D cultures in alginate hydrogel. Capecitabine was used as a model cytotoxic agent. We included liver cells (Huh7) in a coculture model and determined changes in the cytotoxicity levels of capecitabine against NCI-H1299 cells. The experimental part also found that there were differences in sensitivity to capecitabine depending on the type of 3D cultures used.

PMID:39875861 | DOI:10.1186/s12860-025-00532-0

EMBO J. 2025 Jan 28. doi: 10.1038/s44318-025-00368-6. Online ahead of print.

ABSTRACT

Na+-K+-Cl- cotransporters functions as an anion importers, regulating trans-epithelial chloride secretion, cell volume, and renal salt reabsorption. Loop diuretics, including furosemide, bumetanide, and torsemide, antagonize both NKCC1 and NKCC2, and are first-line medicines for the treatment of edema and hypertension. NKCC1 activation by the molecular crowding sensing WNK kinases is critical if cells are to combat shrinkage during hypertonic stress; however, how phosphorylation accelerates NKCC1 ion transport remains unclear. Here, we present co-structures of phospho-activated NKCC1 bound with furosemide, bumetanide, or torsemide showing that furosemide and bumetanide utilize a carboxyl group to coordinate and co-occlude a K+, whereas torsemide encroaches and expels the K+ from the site. We also found that an amino-terminal segment of NKCC1, once phosphorylated, interacts with the carboxyl-terminal domain, and together, they engage with intracellular ion exit and appear to be poised to facilitate rapid ion translocation. Together, these findings enhance our understanding of NKCC-mediated epithelial ion transport and the molecular mechanisms of its inhibition by loop diuretics.

PMID:39875725 | DOI:10.1038/s44318-025-00368-6

Nat Biotechnol. 2025 Jan 28. doi: 10.1038/s41587-024-02524-5. Online ahead of print.

ABSTRACT

Understanding a small molecule's mode of action (MoA) is essential to guide the selection, optimization and clinical development of lead compounds. In this study, we used high-throughput non-targeted metabolomics to profile changes in 2,269 putative metabolites induced by 1,520 drugs in A549 lung cancer cells. Although only 26% of the drugs inhibited cell growth, 86% caused intracellular metabolic changes, which were largely conserved in two additional cancer cell lines. By testing more than 3.4 million drug-metabolite dependencies, we generated a lookup table of drug interference with metabolism, enabling high-throughput characterization of compounds across drug therapeutic classes in a single-pass screen. The identified metabolic changes revealed previously unknown effects of drugs, expanding their MoA annotations and potential therapeutic applications. We confirmed metabolome-based predictions for four new glucocorticoid receptor agonists, two unconventional 3-hydroxy-3-methylglutaryl-CoA (HMGCR) inhibitors and two dihydroorotate dehydrogenase (DHODH) inhibitors. Furthermore, we demonstrated that metabolome profiling complements other phenotypic and molecular profiling technologies, opening opportunities to increase the efficiency, scale and accuracy of preclinical drug discovery.

PMID:39875672 | DOI:10.1038/s41587-024-02524-5

World J Microbiol Biotechnol. 2025 Jan 29;41(2):53. doi: 10.1007/s11274-025-04260-7.

ABSTRACT

The photoautotrophic nature of cyanobacteria, coupled with their fast growth and relative ease of genetic manipulation, makes these microorganisms very promising factories for the sustainable production of bio-products from atmospheric carbon dioxide. However, both in nature and in cultivation, cyanobacteria go through different abiotic stresses such as high light (HL) stress, heavy metal stress, nutrient limitation, heat stress, salt stress, oxidative stress, and alcohol stress. In recent years, significant improvement has been made in identifying the stress-responsive genes and the linked pathways in cyanobacteria and developing genome editing tools for their manipulation. Metabolic pathways play an important role in stress tolerance; their modification is also a very promising approach to adapting to stress conditions. Several synthetic as well as systems biology approaches have been developed to identify and manipulate genes regulating cellular responses under different stresses. In this review, we summarize the impact of different stresses on metabolic processes, the small RNAs, genes and heat shock proteins (HSPs) involved, changes in the metabolome and their adaptive mechanisms. The developing knowledge of the adaptive behaviour of cyanobacteria may also be utilised to develop better stress-responsive strains for various applications.

PMID:39875631 | DOI:10.1007/s11274-025-04260-7

Commun Biol. 2025 Jan 28;8(1):131. doi: 10.1038/s42003-025-07586-y.

ABSTRACT

Rapid structural analysis of purified proteins and their complexes has become increasingly common thanks to key methodological advances in cryo-electron microscopy (cryo-EM) and associated data processing software packages. In contrast, analogous structural analysis in cells via cryo-electron tomography (cryo-ET) remains challenging due to critical technical bottlenecks, including low-throughput sample preparation and imaging, and laborious data processing methods. Here, we describe a rapid in situ cryo-ET sample preparation and data analysis workflow that results in the routine determination of sub-nm resolution ribosomal structures. We apply this workflow to E. coli, producing a 5.8 Å structure of the 70S ribosome from cells in less than 10 days and facilitating the discovery of a minor population of 100S-like disomes. We envision our approach to be widely applicable to related bacterial samples.

PMID:39875527 | DOI:10.1038/s42003-025-07586-y

Sci Rep. 2025 Jan 28;15(1):3549. doi: 10.1038/s41598-025-86242-1.

ABSTRACT

SARS-CoV-2 is a viral infection, best studied in the context of epithelial cell infection. Epithelial cells, when infected with SARS-CoV-2 express the viral S-protein, which causes host cells to fuse together into large multi-nucleated cells known as syncytia. Because SARS-CoV-2 infections also frequently present with cardiovascular phenotypes, we sought to understand if S-protein expression would also result in syncytia formation in endothelial cells. S-protein expression in endothelial cells was sufficient to induce the formation of multi-nucleated cells, with an average of 10% of all cells forming syncytia with an average of 6 nuclei per syncytia after 72 h of S-protein expression. Formation of syncytia was associated with the formation of gaps between cells, suggesting the potential for syncytia formation to compromise barrier function. Inhibition of myosin light chain kinase (MLCK), but not Rho-associated protein kinase, inhibited the formation of syncytia, suggesting a role for MLCK in syncytia formation. Further supporting the role of cellular contractility in syncytia formation, we also observed a reduction in the occurrence of syncytia for endothelial cells grown on substrates with reduced stiffness. Because endothelial cells are exposed to physiological forces due to blood flow, we examined the effects of cyclic biaxial stretch and fluid shear stress. While biaxial stretch did not affect syncytia formation, endothelial cells exposed to fluid shear stress were more resistant to syncytia formation. Finally, we observed that endothelial cells are suitable host cells for SARS-CoV-2 viral infection and replication, and that viral infection also causes syncytia formation. Our studies indicate that endothelial cells, in addition to epithelial cells, should also be considered a target for SARS-CoV-2 infection and a driver of COVID-19-associated pathology.

PMID:39875448 | DOI:10.1038/s41598-025-86242-1

NPJ Syst Biol Appl. 2025 Jan 28;11(1):12. doi: 10.1038/s41540-025-00494-1.

ABSTRACT

Cancer-associated fibroblasts (CAFs) play a key role in metabolic reprogramming and are well-established contributors to drug resistance in colorectal cancer (CRC). To exploit this metabolic crosstalk, we integrated a systems biology approach that identified key metabolic targets in a data-driven method and validated them experimentally. This process involved a novel machine learning-based method to computationally screen, in a high-throughput manner, the effects of enzyme perturbations predicted by a computational model of CRC metabolism. This approach reveals the network-wide effects of metabolic perturbations. Our results highlighted hexokinase (HK) as a crucial target, which subsequently became our focus for experimental validation using patient-derived tumor organoids (PDTOs). Through metabolic imaging and viability assays, we found that PDTOs cultured in CAF-conditioned media exhibited increased sensitivity to HK inhibition, confirming the model predictions. Our approach emphasizes the critical role of integrating computational and experimental techniques in exploring and exploiting CRC-CAF crosstalk.

PMID:39875420 | DOI:10.1038/s41540-025-00494-1

Nat Commun. 2025 Jan 28;16(1):1113. doi: 10.1038/s41467-025-56251-9.

ABSTRACT

Complete blood count indices and their ratios are associated with adverse clinical outcomes for many acute illnesses, but the mechanisms generating these associations are not fully understood. Recent identification of a consistent pattern of white blood cell and platelet count co-regulation during acute inflammatory recovery provides a potentially unifying explanation. Here we show that the platelet-to-white-cell ratio, which was selected based on this conserved recovery pattern, is more strongly associated with mortality than other blood count markers and ratios in four important illnesses involving acute inflammation: COVID-19, acute heart failure, myocardial infarction, and stroke. Patients recovering well from these acute illnesses tend to follow a joint white cell and platelet trajectory that can be reduced to this one-dimensional ratio. The platelet-to-white-cell ratio's association with prognosis is consistent with recently identified inflammatory dynamics and may provide a convenient and interpretable summary of patient inflammatory state.

PMID:39875373 | DOI:10.1038/s41467-025-56251-9

Cell Rep Methods. 2025 Jan 27;5(1):100963. doi: 10.1016/j.crmeth.2024.100963.

ABSTRACT

Identifying key regulators of important genes in non-model crop species is challenging due to limited multi-omics resources. To address this, we introduce the network-enabled gene discovery pipeline NEEDLE, a user-friendly tool that systematically generates coexpression gene network modules, measures gene connectivity, and establishes network hierarchy to pinpoint key transcriptional regulators from dynamic transcriptome datasets. After validating its accuracy with two independent datasets, we applied NEEDLE to identify transcription factors (TFs) regulating the expression of cellulose synthase-like F6 (CSLF6), a crucial cell wall biosynthetic gene, in Brachypodium and sorghum. Our analyses uncover regulators of CSLF6 and also shed light on the evolutionary conservation or divergence of gene regulatory elements among grass species. These results highlight NEEDLE's capability to provide biologically relevant TF predictions and demonstrate its value for non-model plant species with dynamic transcriptome datasets.

PMID:39874949 | DOI:10.1016/j.crmeth.2024.100963

JMIR Res Protoc. 2025 Jan 28;14:e62931. doi: 10.2196/62931.

ABSTRACT

BACKGROUND: Diabetes affects half of the patients with cystic fibrosis who are aged 30 years and older. Diabetes progresses asymptomatically over a long period of time. Two treatment options are possible: start insulin as soon as cystic fibrosis diagnosis is made with the additional constraints of cystic fibrosis or wait while monitoring the patient's clinical condition and start insulin when diabetes symptoms develop and therefore later. This situation is particularly well suited to shared decision-making (SDM) between the physician (health care team) and patient/relatives.

OBJECTIVE: The aim of this study was to perform qualitative and quantitative analyses for evaluating the outcomes and experience of SDM implementation between the physician/health care team trained for SDM and patients/their relatives for cystic fibrosis-related diabetes.

METHODS: A quasi-experimental with a comparison study will be developed. Three cystic fibrosis reference centers (CFRCs) will be trained in SDM by using a web-based training, including a validated decision aid and coaching for physicians and the medical team. Two control CFRCs will maintain their usual practices. A qualitative analysis through observation of consultations, individual semistructured interviews with patients, and focus groups in CFRCs will be conducted based on a thematic content analysis. Questionnaires related to decision-making and experience of decision-making with and without SDM implementation will be administered to patients and physicians.

RESULTS: Forty patients will be included (8 patients in each center), that is, 60 consultation observations (2 consultations per patient in the intervention groups given the modalities of the SDM process) will be conducted in 2025. Eight focus groups will be conducted in the 5 centers (2 groups in each intervention CFRC and 1 group in each control CFRC). This qualitative corpus plus responses to the patient and physician questionnaires will make it possible to know whether the practice of SDM in CFRCs is increased by an implementation strategy and to analyze the experience of patients and their relatives regarding decision-making modalities. Analysis of the outcomes and experience of the implementation of SDM are of importance to identify the facilitators and barriers to SDM from patients' and CFRCs' point of views.

CONCLUSIONS: Our study will give us keys to adapt, improve, and disseminate SDM more widely in the context of cystic fibrosis therapy. SDM could thus be used in routine clinical practice in CFRCs at the national level.

TRIAL REGISTRATION: ClinicalTrials.gov NCT04891159; https://clinicaltrials.gov/study/NCT04891159?id=NCT04891159.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): PRR1-10.2196/62931.

PMID:39874570 | DOI:10.2196/62931

Biomed Phys Eng Express. 2025 Jan 28. doi: 10.1088/2057-1976/adaf29. Online ahead of print.

ABSTRACT

Photoacoustic tomography (PAT) is a non-destructive, non-ionizing, and rapidly expanding hybrid biomedical imaging technique, yet it faces challenges in obtaining clear images due to limited data from detectors or angles. As a result, the methodology suffers from significant streak artifacts and low-quality images. The integration of deep learning (DL), specifically convolutional neural networks (CNNs), has recently demonstrated powerful performance in various fields of PAT. This work introduces a post-processing-based CNN architecture named residual-dense UNet (RDUNet) to address the stride artifacts in reconstructed PA images. The framework adopts the benefits of residual and dense blocks to form high-resolution reconstructed images. The network is trained with two different types of datasets to learn the relationship between the reconstructed images and their corresponding ground truths (GTs). In the first protocol, RDUNet (identified as RDUNet I) underwent training on heterogeneous simulated images featuring three distinct phantom types. Subsequently, in the second protocol, RDUNet (referred to as RDUNet II) was trained on a heterogeneous composition of 81% simulated data and 19% experimental data. The motivation behind this is to allow the network to adapt to diverse experimental challenges. The RDUNet algorithm was validated by performing numerical and experimental studies involving single-disk, T-shape, and vasculature phantoms. The performance of this protocol was compared with the famous backprojection (BP) and the traditional UNet algorithms. This study shows that RDUNet can substantially reduce the number of detectors from 100 to 25 for simulated testing images and 30 for experimental scenarios.

PMID:39874604 | DOI:10.1088/2057-1976/adaf29

ACS Sens. 2025 Jan 28. doi: 10.1021/acssensors.4c03209. Online ahead of print.

ABSTRACT

Multiple respiratory viruses can concurrently or sequentially infect the respiratory tract, making their identification crucial for diagnosis, treatment, and disease management. We present a label-free diagnostic platform integrating surface-enhanced Raman scattering (SERS) with deep learning for rapid, quantitative detection of respiratory virus coinfections. Using sensitive silica-coated silver nanorod array substrates, over 1.2 million SERS spectra are collected from 11 viruses, nine two-virus mixtures, and four three-virus mixtures at various concentrations in saliva. A deep learning model, MultiplexCR, is developed to simultaneously predict virus species and concentrations from SERS spectra. It achieves an impressive 98.6% accuracy in classifying virus coinfections and a mean absolute error of 0.028 for concentration regression. In blind tests, the model demonstrates consistent high accuracy and precise concentration predictions. This SERS-MultiplexCR platform completes the entire detection process in just 15 min, offering significant potential for rapid, point-of-care diagnostics in infection detection, as well as applications in food safety and environmental monitoring.

PMID:39874586 | DOI:10.1021/acssensors.4c03209

J Chem Inf Model. 2025 Jan 28. doi: 10.1021/acs.jcim.4c02088. Online ahead of print.

ABSTRACT

Identification of potential drug-target interactions (DTIs) is a crucial step in drug discovery and repurposing. Although deep learning effectively deciphers DTIs, most deep learning-based methods represent drug features from only a single perspective. Moreover, the fusion method of drug and protein features needs further refinement. To address the above two problems, in this study, we develop a novel end-to-end framework named DO-GMA for potential DTI identification by incorporating Depthwise Overparameterized convolutional neural network and the Gated Multihead Attention mechanism with shared-learned queries and bilinear model concatenation. DO-GMA first designs a depthwise overparameterized convolutional neural network to learn drug representations from their SMILES strings and protein representations from their amino acid sequences. Next, it extracts drug representations from their 2D molecular graphs through a graph convolutional network. Subsequently, it fuses drug and protein features by combining the gated attention mechanism and the multihead attention mechanism with shared-learned queries and bilinear model concatenation. Finally, it takes the fused drug-target features as inputs and builds a multilayer perceptron to classify unlabeled drug-target pairs (DTPs). DO-GMA was benchmarked against six newest DTI prediction methods (CPI-GNN, BACPI, CPGL, DrugBAN, BINDTI, and FOTF-CPI) under four different experimental settings on four DTI data sets (i.e., DrugBank, BioSNAP, C.elegans, and BindingDB). The results show that DO-GMA significantly outperformed the above six methods based on AUC, AUPR, accuracy, F1-score, and MCC. An ablation study, robust statistical analysis, sensitivity analysis of parameters, visualization of the fused features, computational cost analysis, and case analysis further validated the powerful DTI identification performance of DO-GMA. In addition, DO-GMA predicted that two drug-protein pairs (i.e., DB00568 and P06276, and DB09118 and Q9UQD0) could be interacting. DO-GMA is freely available at https://github.com/plhhnu/DO-GMA.

PMID:39874533 | DOI:10.1021/acs.jcim.4c02088

Invest Radiol. 2025 Jan 28. doi: 10.1097/RLI.0000000000001151. Online ahead of print.

ABSTRACT

OBJECTIVE: The aim of this study was to intraindividually compare the conspicuity of focal liver lesions (FLLs) between low- and ultra-low-dose computed tomography (CT) with deep learning reconstruction (DLR) and standard-dose CT with model-based iterative reconstruction (MBIR) from a single CT using dual-split scan in patients with suspected liver metastasis via a noninferiority design.

MATERIALS AND METHODS: This prospective study enrolled participants who met the eligibility criteria at 2 tertiary hospitals in South Korea from June 2022 to January 2023. The criteria included (a) being aged between 20 and 85 years and (b) having suspected or known liver metastases. Dual-source CT scans were conducted, with the standard radiation dose divided in a 2:1 ratio between tubes A and B (67% and 33%, respectively). The voltage settings of 100/120 kVp were selected based on the participant's body mass index (<30 vs ≥30 kg/m2). For image reconstruction, MBIR was utilized for standard-dose (100%) images, whereas DLR was employed for both low-dose (67%) and ultra-low-dose (33%) images. Three radiologists independently evaluated FLL conspicuity, the probability of metastasis, and subjective image quality using a 5-point Likert scale, in addition to quantitative signal-to-noise and contrast-to-noise ratios. The noninferiority margins were set at -0.5 for conspicuity and -0.1 for detection.

RESULTS: One hundred thirty-three participants (male = 58, mean body mass index = 23.0 ± 3.4 kg/m2) were included in the analysis. The low- and ultra-low- dose had a lower radiation dose than the standard-dose (median CT dose index volume: 3.75, 1.87 vs 5.62 mGy, respectively, in the arterial phase; 3.89, 1.95 vs 5.84 in the portal venous phase, P < 0.001 for all). Median FLL conspicuity was lower in the low- and ultra-low-dose scans compared with the standard-dose (3.0 [interquartile range, IQR: 2.0, 4.0], 3.0 [IQR: 1.0, 4.0] vs 3.0 [IQR: 2.0, 4.0] in the arterial phase; 4.0 [IQR: 1.0, 5.0], 3.0 [IQR: 1.0, 4.0] vs 4.0 [IQR: 2.0, 5.0] in the portal venous phases), yet within the noninferiority margin (P < 0.001 for all). FLL detection was also lower but remained within the margin (lesion detection rate: 0.772 [95% confidence interval, CI: 0.727, 0.812], 0.754 [0.708, 0.795], respectively) compared with the standard-dose (0.810 [95% CI: 0.770, 0.844]). Sensitivity for liver metastasis differed between the standard- (80.6% [95% CI: 76.0, 84.5]), low-, and ultra-low-doses (75.7% [95% CI: 70.2, 80.5], 73.7 [95% CI: 68.3, 78.5], respectively, P < 0.001 for both), whereas specificity was similar (P > 0.05).

CONCLUSIONS: Low- and ultra-low-dose CT with DLR showed noninferior FLL conspicuity and detection compared with standard-dose CT with MBIR. Caution is needed due to a potential decrease in sensitivity for metastasis (clinicaltrials.gov/ NCT05324046).

PMID:39874436 | DOI:10.1097/RLI.0000000000001151

PLoS One. 2025 Jan 28;20(1):e0318264. doi: 10.1371/journal.pone.0318264. eCollection 2025.

ABSTRACT

Diabetic retinopathy, a retinal disorder resulting from diabetes mellitus, is a prominent cause of visual degradation and loss among the global population. Therefore, the identification and classification of diabetic retinopathy are of utmost importance in the clinical diagnosis and therapy. Currently, these duties are extensively carried out by manual examination utilizing the human visual system. Nevertheless, manual examination is sometimes arduous, time-consuming, and prone to errors. Deep learning-based methods have recently demonstrated encouraging results in several areas, such as image categorization and natural language mining. The majority of deep learning techniques developed for medical image analysis rely on convolutional modules to extract the inherent structure of images within a certain local receptive field. Furthermore, transformer-based models have been utilized to tackle medical image processing problems by capitalizing on global connections among distant pixels in the images. Considering these analyses, this work presents a comprehensive deep learning model that combines convolutional neural network and vision mamba models. This model is designed to accurately identify and classify diabetic retinopathy lesions displayed in fundus images. Furthermore, the vision mamba component incorporates the bidirectional state space method and positional embedding to enable the location sensitivity of visual data samples and meet the conditions for global relationship context. An evaluation of the suggested method was carried out by comparison experiments between state-of-the-art algorithms and the proposed methodology. Empirical findings demonstrate that the suggested methodology surpasses the most advanced algorithms on the datasets that are accessible openly. Hence, the suggested approach may be regarded as a helpful tool for therapeutic processes.

PMID:39874303 | DOI:10.1371/journal.pone.0318264

PLoS One. 2025 Jan 28;20(1):e0317697. doi: 10.1371/journal.pone.0317697. eCollection 2025.

ABSTRACT

Sharing cooking recipes is a great way to exchange culinary ideas and provide instructions for food preparation. However, categorizing raw recipes found online into appropriate food genres can be challenging due to a lack of adequate labeled data. In this study, we present a dataset named the "Assorted, Archetypal, and Annotated Two Million Extended (3A2M+) Cooking Recipe Dataset" that contains two million culinary recipes labeled in respective categories with extended named entities extracted from recipe descriptions. This collection of data includes various features such as title, NER, directions, and extended NER, as well as nine different labels representing genres including bakery, drinks, non-veg, vegetables, fast food, cereals, meals, sides, and fusions. The proposed pipeline named 3A2M+ extends the size of the Named Entity Recognition (NER) list to address missing named entities like heat, time or process from the recipe directions using two NER extraction tools. 3A2M+ dataset provides a comprehensive solution to the various challenging recipe-related tasks, including classification, named entity recognition, and recipe generation. Furthermore, we have demonstrated traditional machine learning, deep learning and pre-trained language models to classify the recipes into their corresponding genre and achieved an overall accuracy of 98.6%. Our investigation indicates that the title feature played a more significant role in classifying the genre.

PMID:39874282 | DOI:10.1371/journal.pone.0317697

Ann Rheum Dis. 2025 Jan;84(1):60-67. doi: 10.1136/ard-2024-225862. Epub 2025 Jan 2.

ABSTRACT

OBJECTIVES: To assess the ability of a previously trained deep-learning algorithm to identify the presence of inflammation on MRI of sacroiliac joints (SIJ) in a large external validation set of patients with axial spondyloarthritis (axSpA).

METHODS: Baseline SIJ MRI scans were collected from two prospective randomised controlled trials in patients with non-radiographic (nr-) and radiographic (r-) axSpA (RAPID-axSpA: NCT01087762 and C-OPTIMISE: NCT02505542) and were centrally evaluated by two expert readers (and adjudicator in case of disagreement) for the presence of inflammation by the 2009 Assessment of SpondyloArthritis International Society (ASAS) definition. Scans were processed by the deep-learning algorithm, blinded to clinical information and central expert readings.

RESULTS: Pooling the patients from RAPID-axSpA (n=152) and C-OPTIMISE (n=579) yielded a validation set of 731 patients (mean age: 34.2 years, SD: 8.6; 505/731 (69.1%) male), of which 326/731 (44.6%) had nr-axSpA and 436/731 (59.6%) had inflammation on MRI per central readings. Scans were obtained from over 30 scanners from 5 manufacturers across over 100 clinical sites. Comparing the trained algorithm with the human central readings yielded a sensitivity of 70% (95% CI 66% to 73%), specificity of 81% (95% CI 78% to 84%), positive predictive value of 84% (95% CI 82% to 87%), negative predictive value of 64% (95% CI 61% to 68%), Cohen's kappa of 0.49 (95% CI 0.43 to 0.55) and absolute agreement of 74% (95% CI 72% to 77%).

CONCLUSION: The algorithm enabled acceptable detection of inflammation according to the 2009 ASAS MRI definition in a large external validation cohort.

PMID:39874235 | DOI:10.1136/ard-2024-225862

Adv Sci (Weinh). 2025 Jan 28:e2409130. doi: 10.1002/advs.202409130. Online ahead of print.

ABSTRACT

Efficient virtual screening methods can expedite drug discovery and facilitate the development of innovative therapeutics. This study presents a novel transfer learning model based on network target theory, integrating deep learning techniques with diverse biological molecular networks to predict drug-disease interactions. By incorporating network techniques that leverage vast existing knowledge, the approach enables the extraction of more precise and informative drug features, resulting in the identification of 88,161 drug-disease interactions involving 7,940 drugs and 2,986 diseases. Furthermore, this model effectively addresses the challenge of balancing large-scale positive and negative samples, leading to improved performance across various evaluation metrics such as an Area under curve (AUC) of 0.9298 and an F1 score of 0.6316. Moreover, the algorithm accurately predicts drug combinations and achieves an F1 score of 0.7746 after fine-tuning. Additionally, it identifies two previously unexplored synergistic drug combinations for distinct cancer types in disease-specific biological network environments. These findings are further validated through in vitro cytotoxicity assays, demonstrating the potential of the model to enhance drug development and identify effective treatment regimens for specific diseases.

PMID:39874191 | DOI:10.1002/advs.202409130