Respir Med. 2025 Apr 22:108110. doi: 10.1016/j.rmed.2025.108110. Online ahead of print.

ABSTRACT

BACKGROUND: Chronic respiratory diseases are associated with significant disability and death. Pulmonary rehabilitation (PR) is recommended in the management of chronic respiratory diseases. There is limited population level data comparing PR utilization and completion among patients with chronic respiratory diseases.

METHODS: A retrospective, cross sectional analysis concerning PR use in adults residing in the U.S. with chronic obstructive pulmonary disease (COPD), interstitial lung disease (ILD), idiopathic pulmonary fibrosis (IPF), pulmonary hypertension, and bronchiectasis was conducted using the Merative™ MarketScan® Research Databases. PR use was identified using current procedural terminology (CPT) and healthcare common procedure coding system (HCPCS) codes. Demographics, comorbidities, oxygen use, medications, initiation and participation of PR by disease state were collected. Analysis involved chi-square tests and generalized estimating equations.

RESULTS: From 2014 to 2019, we identified 892,741 adults with chronic respiratory diseases and COPD was the most prevalent. PR initiation occurred in 2.3% and annual participation ranged from 1.5 % to 1.7 %. The IPF group had the largest proportion of patients that initiated PR compared to other groups. Completion of ≥ 8 sessions was greatest for the group with IPF (60.8 %), followed by non IPF ILD (56.2 %), bronchiectasis (55.3 %), pulmonary hypertension (55.1 %) and COPD (53.9 %). Completion of ≥ 8 sessions was significantly greater for the IPF group compared to the COPD group, (p <0.0001).

CONCLUSION: PR was underutilized among individuals with chronic respiratory disease, however the group with IPF demonstrated the greatest proportion of PR initiation and completion compared with other groups.

PMID:40273996 | DOI:10.1016/j.rmed.2025.108110

Genome Biol. 2025 Apr 24;26(1):105. doi: 10.1186/s13059-025-03554-1.

ABSTRACT

In spatial transcriptomics data, spatially correlated genes promise to reveal high-interest phenomena like cell-cell interactions and latent variables. But in practice, most spatial correlations arise from the spatial arrangement of cell types, obscuring the more interesting relationships we hope to discover. We introduce InSituCor, a toolkit for discovering modules of spatially correlated genes. InSituCor returns only correlations not explainable by already-known factors like the cell type landscape; this spares precious analyst effort. InSituCor supports both unbiased discovery of whole-dataset correlations and knowledge-driven exploration of genes of interest. As a special case, it evaluates ligand-receptor pairs for spatial co-regulation.

PMID:40275395 | DOI:10.1186/s13059-025-03554-1

BMC Microbiol. 2025 Apr 24;25(1):243. doi: 10.1186/s12866-025-03933-5.

ABSTRACT

BACKGROUND: Inflammatory events followed by bacterial infections are related to the progression of the atherosclerosis process. The study investigated the signaling and metabolic pathways of endothelial cells (ECs), macrophages (MQs), vascular smooth muscle cells (VSMCs), and dendritic cells (DCs) after exposure to Gram-negative bacterial infections. Moreover, it aimed at cross-talking and enriching the pathways on the cellular and plaque networks.

METHODS AND MATERIALS: High-throughput expression data series (n = 9) were selected through GEO and MAT data repositories. Upregulated differential expression genes (DEGs) were determined using R software and applied to identify the cellular signaling pathways using Enricher/Reactome tools. Then, the cell networks were visualized using the Cytoscape software and enriched by the pathways of secretory proteins identified using Gene ontology (GO).

RESULTS: The important pathways of the Cytokines (Degree 4, p < 6 × 10-26), and INF (Degree 4, p < 8.6 × 10-31) in ECs, Cytokines (Degree 4, p < 9.35 × 10-8), and GPCR (Degree 3, p < 1.45 × 10-4) in MQs, NOTCH (Degree 6, p < 0.027) in VSMCs, and Cytokines (Degree 4, p < 1.45 × 10-17) in DCs were found to be activated and enriched after exposure to Gram-negative bacterial infections on the cell networks. Furthermore, the Netrin- 1 (Degree 6, p < 0.028), and EGFR (Degree 5, p < 0.036) pathways were activated in the intimal thick/xanthoma plaque network while the innate (Degree 9, p < 8.9 × 10-20) and adaptive (Degree 7, p < 4.1 × 10-12) immune systems pathways were activated in the fibrous cap atheroma plaque network.

CONCLUSION: The study revealed the signaling pathways after exposure to Gram-negative bacterial infections on the cell networks in the vessel microenvironment. Furthermore, the cell cross-talks exacerbated these pathways in cells and unstable plaques.

CLINICAL TRIAL NUMBER: Not applicable.

PMID:40275124 | DOI:10.1186/s12866-025-03933-5

BJC Rep. 2025 Apr 24;3(1):27. doi: 10.1038/s44276-025-00134-5.

ABSTRACT

BACKGROUND: Glioblastoma (GB), an aggressive brain malignancy with a poor prognosis of 1.5-2 years, rarely exhibits extracranial metastasis (ECM). However, metabolic reprogramming has emerged as a key driver of GB progression and invasiveness. This study presents a rare case of recurrent GB with scalp metastasis, exploring how metabolic shifts enable GB cells to evade treatment and adapt to hostile environments, offering insights for developing innovative therapies.

METHODS: Tandem mass spectrometry (MS/MS) was employed to analyze amino acid profiles in both the recurrent and metastatic stages of GB. Systems biology approaches were used to uncover genetic alterations and metabolic reprogramming associated with the progression from recurrence to metastasis.

RESULTS: Our analysis revealed distinct amino acid utilization patterns in a patient with a molecular phenotype of wild-type IDH-1&2, TERT mutation, non-mutated BRAF and EGFR, and non-methylated MGMT. During recurrence and metastasis, significant differences in amino acid profiles were observed between blood and cerebrospinal fluid (CSF) samples. Additionally, protein-protein interaction (PPI) analysis identified key genomic drivers potentially responsible for the transition from recurrent to metastatic GB.

CONCLUSIONS: Beyond established risk factors such as craniotomy, biopsies, ventricular shunting, and radiation therapy, our findings suggest that metabolic reprogramming plays a crucial role in the transition from recurrent to metastatic GB. Targeting these metabolic shifts could provide new avenues for managing and preventing extracranial metastasis in GB, making this an important focus for future research.

PMID:40274950 | DOI:10.1038/s44276-025-00134-5

Anal Chim Acta. 2025 Jun 15;1355:343999. doi: 10.1016/j.aca.2025.343999. Epub 2025 Apr 1.

ABSTRACT

BACKGROUND: Quantitative bioimaging of trace elements at the single-cell level is crucial for understanding cellular processes, including metal uptake and distribution. Laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) has emerged as a gold standard for elemental bioimaging due to its high sensitivity and spatial resolution. However, calibration remains challenging due to the lack of homogeneous biological standards. This study addresses these challenges by introducing a gelatin-based calibration strategy optimized for Zn mapping in human parietal cells. By minimizing heterogeneity in gelatin standards and optimizing laser ablation conditions, the approach ensures accurate and reproducible results for cellular bioimaging.

RESULTS: A gelatin-based calibration strategy for LA-ICP-MS was developed to quantify intracellular Zn at a single-cell level in human parietal cells. Preparation conditions for gelatin standards were optimized to minimize heterogeneity, eliminating the need for entire droplet ablation and significantly reducing analysis time. Atomic force microscopy (AFM) was employed to optimize laser ablation conditions and determine ablated volumes, ensuring quantitative Zn detection. The method demonstrated high linearity (R2 > 0.99) and reproducibility. Application of the calibration strategy to ZnCl2-treated parietal cells revealed Zn distribution at a cellular level, visualized using a 5 μm laser beam. Integration with bright field imaging enabled the exclusion of apoptotic cells and debris, ensuring robust analysis. Validation with bulk ICP-MS showed excellent agreement, confirming the method's reliability and potential for high-resolution bioimaging.

SIGNIFICANCE: This work introduces a robust and reproducible calibration strategy for quantitative elemental bioimaging using LA-ICP-MS. It details the preparation of a gelatin matrix with a homogeneous element distribution, serving as an alternative to using biological material and significantly reducing analysis time. Laser ablation parameters were optimized using AFM to ensure quantitative ablation, which is necessary for calibration through LA-ICP-MS imaging. This approach provides a powerful tool for studying trace element dynamics in single cells and holds potential for diverse biological and biomedical applications.

PMID:40274329 | DOI:10.1016/j.aca.2025.343999

Lupus Sci Med. 2025 Apr 23;12(1):e001446. doi: 10.1136/lupus-2024-001446.

ABSTRACT

OBJECTIVES: This study aimed to identify novel, non-invasive biomarkers for lupus nephritis (LN) through serum proteomics.

METHODS: Serum proteins were detected in patients with LN and healthy control (HC) groups through liquid chromatography-tandem mass spectrometry. The key networks associated with LN were screened out using Cytoscape software, followed by pathway enrichment analysis. The best candidate biomarkers were selected by machine learning models, further validated in a larger independent cohort. Finally, the expression of these candidate markers was verified in kidney tissue samples, and the mechanism was explored by knocking down the expression of intercellular adhesion molecule 2 (ICAM2) through in vitro cell transfection with siRNA.

RESULTS: Following the serum proteomic screening of LN, a key network of 20 proteins was identified. Machine learning models were used to select ICAM2 (CD102), metalloproteinase inhibitor 1 (TIMP1) and thrombospondin 1 (THSB1) for validation in independent cohorts. ICAM2 exhibited the highest area under the curve (AUC) value in distinguishing LN from HC (AUC=0.92) and was significantly correlated with activity index, proteinuria, albumin and anti-dsDNA antibody levels. Particularly, ICAM2 was significantly elevated in proliferative LN and was associated with specific pathological attributes, outperforming conventional parameters in distinguishing proliferative LN from non-proliferative LN. ICAM2 expression was also elevated in renal tissue samples from patients with proliferative LN. In vitro, knockdown of ICAM2 expression can inhibit the activation of the PI3K/Akt pathway and alleviate the injury of glomerular endothelial cells.

CONCLUSION: ICAM2 (CD102) may serve as a potential serum biomarker for proliferative LN that reflects renal pathology activity, potentially contributing to the progression of LN through the PI3K/Akt pathway.

PMID:40274316 | DOI:10.1136/lupus-2024-001446

Pract Radiat Oncol. 2025 Apr 22:S1879-8500(25)00097-9. doi: 10.1016/j.prro.2025.03.010. Online ahead of print.

ABSTRACT

177Lu-PSMA-617 (LuPSMA) is an effective radiopharmaceutical therapy for patients with metastatic castration-resistant prostate cancer (mCRPC). While LuPSMA can treat disseminated disease, additional localized control of metastatic disease may be required. Metastasis-targeted external beam radiation therapy (M-EBRT) can be an effective adjunct. However, the indications, efficacy, and safety/toxicity of combining M-EBRT with LuPSMA are unclear. Here, we report our experience with M-EBRT in patients receiving LuPSMA and assess M-EBRT's ability for local disease control and palliation.

METHODS: This retrospective IRB-exempted study reviewed patients treated with LuPSMA at a multi-institutional academic cancer center within the first two years post-FDA approval receiving contemporaneous M-EBRT. Clinical factors driving the use of M-EBRT were analyzed.

RESULTS: Treatment courses of 261 patients receiving LuPSMA were reviewed; 52 patients received M-EBRT contemporaneously. M-EBRT was administered for intracranial/epidural disease (n=22/52; 42%), bone pain palliation (n=17/52; 33%), prevention of pathological fractures (n=12/52; 23%), 12% (n=6/52) for various other indications. M-EBRT timing varied among patients, with 54% (n=28/52) receiving M-EBRT before, 27% (n=14/52) after, and 13% (n=7/52) during LuPSMA therapy. EBRT was mostly well-tolerated, although lymphopenia was commonly experienced. Most patients (n=32/52, 62%) had symptom relief following M-EBRT. Symptom relief post-M-EBRT was 68%, 85% vs. 50%, and mortality rates were 32%, 29% vs. 57% for patients receiving EBRT before, during, and after LuPSMA treatment, respectively, albeit not statistically significant (p>0.23). PSA50 (decrease in prostate-specific antigen by 50% during treatment) response in this patient population was 41% compared to 50% in the general LuPSMA population, but the magnitude of PSA response was heterogeneous (p=0.27).

CONCLUSION: In our experience, M-EBRT was used effectively with LuPSMA therapy for local tumor control and symptom management, especially for localized osseous and central nervous system lesions, and with good tolerability. M-EBRT may be an important adjunct treatment modality that facilitates the initiation and/or continuation of LuPSMA.

PMID:40274261 | DOI:10.1016/j.prro.2025.03.010

Life Sci. 2025 Apr 22:123635. doi: 10.1016/j.lfs.2025.123635. Online ahead of print.

ABSTRACT

AIMS: Aralia continentalis Kitag roots (ACKRs) have been regarded as a nutritional natural resource for treating different diseases, including type 2 diabetes mellitus (T2DM), and its complications (heart attack; HA, diabetic nephropathy; DN). Nonetheless, an extensive investigation of T2DM-derived complications has yet to be performed.

MAIN METHODS: Accordingly, we adopted gas chromatography-mass spectrometry (GC-MS) to identify the molecules of ACKRs, followed by the use of cheminformatics (Similarity Ensemble Approach; SEA, SwissTargetPrediction; STP), bioinformatics (STRING, DisGeNET, and OMIM), and computer screening tools to investigate its corresponding targets, in T2DM diseases and its complications.

KEY FINDINGS: The primary targets (PPARG, and IL6) were confirmed via a protein-protein interaction (PPI) network, suggesting that IL6- Andrographolide, PPARA-Germacrene D, PPARD- Kaurenoic acid, PPARG- Kaurenoic acid, NR1H3- 1-Naphthalenepropanol, α-ethenyldecahydro-5-(hydroxymethyl)-α,2,5,5,8a-pentamethyl-, and FABP4- Kaurenoic acid conformers on PPAR signaling pathway might exert agonistic mode.

SIGNIFICANCE: These findings underline that ACKRs' bioactives filtered by the devised platform could prevent T2DM-derived complications through multiple-target.

PMID:40274257 | DOI:10.1016/j.lfs.2025.123635

Cell. 2025 Apr 21:S0092-8674(25)00398-8. doi: 10.1016/j.cell.2025.03.046. Online ahead of print.

ABSTRACT

In humans, selective and promiscuous interactions between 46 secreted chemokine ligands and 23 cell surface chemokine receptors of the G-protein-coupled receptor (GPCR) family form a complex network to coordinate cell migration. While chemokines and their GPCRs each share common structural scaffolds, the molecular principles driving selectivity and promiscuity remain elusive. Here, we identify conserved, semi-conserved, and variable determinants (i.e., recognition elements) that are encoded and decoded by chemokines and their receptors to mediate interactions. Selectivity and promiscuity emerge from an ensemble of generalized ("public/conserved") and specific ("private/variable") determinants distributed among structured and unstructured protein regions, with ligands and receptors recognizing these determinants combinatorially. We employ these principles to engineer a viral chemokine with altered GPCR coupling preferences and provide a web resource to facilitate sequence-structure-function studies and protein design efforts for developing immuno-therapeutics and cell therapies.

PMID:40273912 | DOI:10.1016/j.cell.2025.03.046

Cell. 2025 Apr 17:S0092-8674(25)00397-6. doi: 10.1016/j.cell.2025.03.045. Online ahead of print.

ABSTRACT

Virtually all individuals aged 65 or older develop at least early pathology of Alzheimer's disease (AD), yet most lack disease-causing mutations in APP, PSEN, or MAPT, and many do not carry the APOE4 risk allele. This raises questions about AD development in the general population. Although transcriptional dysregulation has not traditionally been a hallmark of AD, recent studies reveal significant epigenomic changes in late-onset AD (LOAD) patients. We show that altered expression of the LOAD biomarker phosphoglycerate dehydrogenase (PHGDH) modulates AD pathology in mice and human brain organoids independent of its enzymatic activity. PHGDH has an uncharacterized role in transcriptional regulation, promoting the transcription of inhibitor of nuclear factor kappa-B kinase subunit alpha (IKKa) and high-mobility group box 1 (HMGB1) in astrocytes, which suppress autophagy and accelerate amyloid pathology. A blood-brain-barrier-permeable small-molecule inhibitor targeting PHGDH's transcriptional function reduces amyloid pathology and improves AD-related behavioral deficits. These findings highlight transcriptional regulation in LOAD and suggest therapeutic strategies beyond targeting familial mutations.

PMID:40273909 | DOI:10.1016/j.cell.2025.03.045

J Hazard Mater. 2025 Apr 22;493:138383. doi: 10.1016/j.jhazmat.2025.138383. Online ahead of print.

ABSTRACT

Uranium (U) can impact microbially driven soil phosphorus (P) and carbon (C) cycling. However, the response of microbial P and C turnover to U in different soils is not well understood. Through the quantitative assay of P pools and soil organic C (SOC) quantitative assay and sequencing of 16S rRNA gene amplicons and metagenomes, we investigated the effect of U on P and C biotransformation in grassland (GL), paddy soil (PY), forest soil (FT). U (60 mg kg-1) impacted the diversity, interaction and stability of soil bacterial communities, leading to a decrease in available P (AP). Under U stress, organophosphate mineralization substantially contributed to the AP in GL and FT, whereas intracellular P metabolism dominated the AP in PY. Also, the reductive citrate cycle (rTCA cycle) promoted the content of SOC in GL, while the rTCA cycle and complex organic C degradation pathways enhanced the SOC in PY and FT, respectively. Notably, functional bacteria carrying organic C degradation genes could decompose SOC to enhance soil AP. Bacteria developed various resistance strategies to cope with U stress. This study reveals soil-dependent response of microbial P and C cycling and its ecological functions under the influence of radioactive contaminants in different soil systems.

PMID:40273857 | DOI:10.1016/j.jhazmat.2025.138383

Eur J Hosp Pharm. 2025 Apr 24:ejhpharm-2025-004492. doi: 10.1136/ejhpharm-2025-004492. Online ahead of print.

ABSTRACT

INTRODUCTION: Fluid overload is associated with increased morbidity in patients in paediatric intensive care units (PICUs). This study aimed to evaluate pharmaceutical infusion management as a quality assurance measure to reduce fluid overload in routine paediatric intensive care.

METHODS: This was a retrospective observational study in a PICU with two periods: a control period and a period after the implementation of pharmaceutical infusion management (PharmInfuManagement period). Pharmaceutical infusion management consisted of two components carried out simultaneously: the creation of flushing schedules to reduce incompatibilities and flushing volume and the reduction of dilution volume for six non-continuous intravenous (IV) drugs to reduce fluid intake because of IV drugs. The primary outcome was the number of patients with ≥5% fluid overload. In addition, daily furosemide dose (mg/kg/day), non-continuous IV drug volume (mL/kg/day), flushing volume (mL/kg/day) and number of incompatibilities were evaluated.

RESULTS: Sixty-six patients were included in each period. Fluid overload of ≥5% occurred in 52% of patients in the control period and in 29% of patients in the PharmInfuManagement period (p=0.01). Flushing volume decreased from 0.7 mL/kg/day (median Q25/Q75 0.4/1.4) to 0.3 mL/kg/day (median Q25/Q75 0.1/0.7) (p<0.001) after implementation. During the PharmInfuManagement period, potentially incompatible drug combinations were reduced from 17.1% (86/504) to 8.2% (43/523) (p<0.001). The volume required for reconstitution and dilution of non-continuously administered IV drugs was reduced from 8.8 mL/kg/day (median Q25/Q75 7.1/12.6) to 6.8 mL/kg/day (median Q25/Q75 5.5/8.0) (p=0.02).

CONCLUSION: Pharmaceutical infusion management reduces incompatibilities and fluid overload in PICU patients.

PMID:40274396 | DOI:10.1136/ejhpharm-2025-004492

BMJ Glob Health. 2025 Apr 24;10(4):e016720. doi: 10.1136/bmjgh-2024-016720.

ABSTRACT

INTRODUCTION: Plasmodium vivax recurrence prevention using tafenoquine or primaquine is critical for achieving Thailand's malaria elimination targets. Both drugs may cause haemolysis in glucose-6-phosphate dehydrogenase (G6PD) deficient individuals. This study evaluated the operational feasibility and safety of administering tafenoquine or primaquine after quantitative G6PD point-of-care testing in Thailand.

METHODS: This prospective, observational, multicentre, longitudinal study was conducted between 23 May 2022 and 14 September 2023 during pilot implementation at seven sites in Yala and Mae Hong Son provinces. Eligible patients were ≥16 years old with uncomplicated P. vivax malaria. G6PD enzyme activity was quantified using a point-of-care device. All patients received 3-day chloroquine plus (based on G6PD enzyme activity): single-dose tafenoquine 300 mg (≥6.1 U/g Hb), or primaquine 15 mg/day for 14 days (≥4.1 U/g Hb), or primaquine 45 mg/week for 8 weeks (≤4.0 U/g Hb), with follow-up on days 5 and 14. Hospital admissions were reviewed to confirm acute haemolytic anaemia cases. The primary endpoint was the percentage of P. vivax patients ≥16 years old treated or not treated with tafenoquine in accordance with G6PD enzyme activity.

RESULTS: Of 316 P. vivax patients screened, 187 were enrolled. All patients completed quantitative G6PD testing. According to G6PD status, appropriate use or non-use was 100% (95% CI 97.2, 100 (132/132)) with tafenoquine, 100% (95% CI 96.5, 100 (104/104)) with daily primaquine and 99.5% (97.1, 100 (186/187)) with weekly primaquine. At day 5, adverse events possibly related to haemolysis occurred in 46.3% (37/80) of patients with tafenoquine, 56.8% (46/81) with daily primaquine and 77.8% (14/18) with weekly primaquine, with no confirmed drug-induced acute haemolytic anaemia cases.

CONCLUSION: Point-of-care quantitative G6PD testing prior to appropriate tafenoquine or primaquine administration was operationally feasible within the Thailand health system, with no concerning adverse events, supporting implementation of this treatment algorithm in areas of active P. vivax transmission.

PMID:40274286 | DOI:10.1136/bmjgh-2024-016720

J Med Internet Res. 2025 Apr 24;27:e65937. doi: 10.2196/65937.

ABSTRACT

BACKGROUND: The incidence of adrenal incidentalomas is increasing annually, and most types of adrenal masses require surgical intervention. Accurate classification of common adrenal masses based on tumor computed tomography (CT) images by radiologists or clinicians requires extensive experience and is often challenging, which increases the workload of radiologists and leads to unnecessary adrenal surgeries. There is an urgent need for a fully automated, noninvasive, and precise approach for the identification and accurate classification of common adrenal masses.

OBJECTIVE: This study aims to enhance diagnostic efficiency and transform the current clinical practice of preoperative diagnosis of adrenal masses.

METHODS: This study is a retrospective analysis that includes patients with adrenal masses who underwent adrenalectomy from January 1, 2021, to May 31, 2023, at Center 1 (internal dataset), and from January 1, 2016, to May 31, 2023, at Center 2 (external dataset). The images include unenhanced, arterial, and venous phases, with 21,649 images used for the training set, 2406 images used for the validation set, and 12,857 images used for the external test set. We invited 3 experienced radiologists to precisely annotate the images, and these annotations served as references. We developed a deep learning-based adrenal mass detection model, Multi-Attention YOLO (MA-YOLO), which can automatically localize and classify 6 common types of adrenal masses. In order to scientifically evaluate the model performance, we used a variety of evaluation metrics, in addition, we compared the improvement in diagnostic efficacy of 6 doctors after incorporating model assistance.

RESULTS: A total of 516 patients were included. In the external test set, the MA-YOLO model achieved an intersection over union of 0.838, 0.885, and 0.890 for the localization of 6 types of adrenal masses in unenhanced, arterial, and venous phase CT images, respectively. The corresponding mean average precision for classification was 0.885, 0.913, and 0.915, respectively. Additionally, with the assistance of this model, the classification diagnostic performance of 6 radiologists and clinicians for adrenal masses improved. Except for adrenal cysts, at least 1 physician significantly improved diagnostic performance for the other 5 types of tumors. Notably, in the categories of adrenal adenoma (for senior clinician: P=.04, junior radiologist: P=.01, and senior radiologist: P=.01) and adrenal cortical carcinoma (junior clinician: P=.02, junior radiologist: P=.01, and intermediate radiologist: P=.001), half of the physicians showed significant improvements after using the model for assistance.

CONCLUSIONS: The MA-YOLO model demonstrates the ability to achieve efficient, accurate, and noninvasive preoperative localization and classification of common adrenal masses in CT examinations, showing promising potential for future applications.

PMID:40273442 | DOI:10.2196/65937

Brief Bioinform. 2025 Mar 4;26(2):bbaf196. doi: 10.1093/bib/bbaf196.

ABSTRACT

In post-genome-wide association study era, interpretation of noncoding variants remains a significant challenge due to their complexity and the limited understanding of their functions. Here, we developed MIRACN, a novel residual convolutional neural network designed to predict cell line-specific functional regulatory variants. By utilizing a substantial dataset from massively parallel reporter assays (MPRAs) and employing a multitask learning strategy, MIRACN was trained across seven distinct cell lines, attaining superior performance compared to existing methods, especially in predicting cell type specificity. Comparative evaluations on an independent MPRA test dataset demonstrated that MIRACN not only outperformed in identifying regulatory variants but also provided valuable insights into their cellular context-specific regulatory mechanisms. MIRACN is capable of not only providing scores for functional variants but also pinpointing the specific cell line in which these variants display their function. This enhancement has improved the resolution of current research on the functionality of noncoding variants and has paved the way for more precise diagnostic and therapeutic strategies.

PMID:40273430 | DOI:10.1093/bib/bbaf196

Brief Bioinform. 2025 Mar 4;26(2):bbaf192. doi: 10.1093/bib/bbaf192.

ABSTRACT

Cancer is a major public health problem while liver cancer is the main cause of global cancer-related deaths. The previous study demonstrates that the 5-year survival rate for advanced liver cancer is only 30%. Few of the first-line targeted drugs including sorafenib and lenvatinib are available, which often develop resistance. Drug combination therapy is crucial for improving the efficacy of cancer therapy and overcoming resistance. However, traditional methods for discovering drug synergy are costly and time consuming. In this study, we developed a novel predicting model PathSynergy by integrating drug feature data, cell line data, drug-target interactions, and signaling pathways. PathSynergy combined the advantages of graph neural networks and pathway map mapping. Comparing with other baseline models, PathSynergy showed better performance in model classification, accuracy, and precision. Excitingly, six Food and Drug Administration (FDA)-approved drugs including pimecrolimus, topiramate, nandrolone_decanoate, fluticasone propionate, zanubrutinib, and levonorgestrel were predicted and validated to show synergistic effects with sorafenib or lenvatinib against liver cancer for the first time. In general, the PathSynergy model provides a new perspective to discover synergistic combinations of drugs and has broad application potential in the fields of drug discovery and personalized medicine.

PMID:40273429 | DOI:10.1093/bib/bbaf192

Brief Bioinform. 2025 Mar 4;26(2):bbaf190. doi: 10.1093/bib/bbaf190.

ABSTRACT

Research on protein stability changes is vital for understanding disease mechanisms and optimizing industrial enzymes. Protein thermal stability can be modified by variants leading to changes in ΔΔG values between wild-type and mutant proteins. Despite advances, most models focus on single-point mutations, overlooking multipoint and indel mutations. Typically, the single-point mutation is expected to have a relatively limited impact on the function of a protein, necessitating more drastic modifications to meet new challenges. Current methods for multipoint mutations yield poor results, and no method exists for any length of indel mutations. To address this, we introduce UniMutStab, a shared-graph convolutional network leveraging protein language models and residue interaction networks to access any type of mutation. An embedded edge weight module enhances the integration of residue node features and interactions, improving prediction accuracy. Trained on the "Mega-scale" dataset with ~780 000 mutations, UniMutStab surpasses existing methods in predicting protein stability changes. It is a purely sequence-based approach to predict arbitrary mutation types, demonstrating robust generalization across multiple tasks and potentially contributing significantly to protein engineering, personalized therapeutics, and diagnostic methodologies.

PMID:40273428 | DOI:10.1093/bib/bbaf190

Brief Bioinform. 2025 Mar 4;26(2):bbaf187. doi: 10.1093/bib/bbaf187.

ABSTRACT

The prediction of enzyme kinetic parameters is crucial for screening enzymes with high catalytic efficiency and desired characteristics to catalyze natural or non-natural reactions. Data-driven machine learning models have been explored to reduce experimental cost and speed up the enzyme design process. However, the prediction performance is still subject to significant limitations due to the variance in sequence similarity between training and testing datasets. In this work, we introduce DEKP, an integrated deep learning approach enzyme kinetic parameter prediction. It leverages pretrained models of protein sequences and incorporates enhanced graph neural networks that provide comprehensive representation of protein structural features. This novel approach can effectively alleviate the performance degradation caused by sequence similarity variation. Moreover, it provides sensitive detection of changes in catalytic efficiency due to enzyme mutations. Experiments validate that DEKP outperforms existing models in predicting enzyme kinetic parameters. This work is expected to significantly improve the performance of the enzyme screening process and provide a robust tool for enzyme-directed evolution research.

PMID:40273427 | DOI:10.1093/bib/bbaf187

PLoS One. 2025 Apr 24;20(4):e0321754. doi: 10.1371/journal.pone.0321754. eCollection 2025.

ABSTRACT

Recognition of Human Actions (HAR) Portrays a crucial significance in various applications due to its ability for analyzing behaviour of humans within videos. This research investigates HAR in Red, Green, and Blue, or RGB videos using frameworks for deep learning. The model's ensemble method integrates the forecasts from two models, 3D-AlexNet-RF and InceptionV3 Google-Net, to improve accuracy in recognizing human activities. Each model independently predicts the activity, and the ensembles method merges these predictions, often using voting or averaging, to produce a more accurate and reliable final classification. This approach leverages the advantages of each design, leading to enhanced performance recognition for activities. The performance of our ensemble framework is evaluated on our contesting HMDB51 dataset, known for its diverse human actions. Training the Inflated-3D (I3D) video classifiers using HMDB51 dataset, our system aims to improve patient care, enhance security, surveillances, Interaction between Humans and Computers, or HCI, and advance human-robot interaction. The ensemble model achieves exceptional results in every class, with an astounding aggregate accuracy of 99.54% accuracy, 97.94% precision, 97.94% recall, 99.56% specificity, 97.88% F1-Score, 95.43% IoU,97. 36% MCC and Cohen's Kappa 97.17%. These findings suggest that the ensemble model is highly effective & a powerful tool for HAR tasks. Multi-tiered ensembles boost wearable recognition, setting a new gold standard for healthcare, surveillance, and robotics.

PMID:40273193 | DOI:10.1371/journal.pone.0321754

PLoS One. 2025 Apr 24;20(4):e0320077. doi: 10.1371/journal.pone.0320077. eCollection 2025.

ABSTRACT

Pseudouridine is an important modification site, which is widely present in a variety of non-coding RNAs and is involved in a variety of important biological processes. Studies have shown that pseudouridine is important in many biological functions such as gene expression, RNA structural stability, and various diseases. Therefore, accurate identification of pseudouridine sites can effectively explain the functional mechanism of this modification site. Due to the rapid increase of genomics data, traditional biological experimental methods to identify RNA modification sites can no longer meet the practical needs, and it is necessary to accurately identify pseudouridine sites from high-throughput RNA sequence data by computational methods. In this study, we propose a deep learning-based computational method, Definer, to accurately identify RNA pseudouridine loci in three species, Homo sapiens, Saccharomyces cerevisiae and Mus musculus. The method incorporates two sequence coding schemes, including NCP and One-hot, and then feeds the extracted RNA sequence features into a deep learning model constructed from CNN, GRU and Attention. The benchmark dataset contains data from three species, H. sapiens, S. cerevisiae and M. musculus, and the results using 10-fold cross-validation show that Definer significantly outperforms other existing methods. Meanwhile, the data sets of two species, H. sapiens and S. cerevisiae, were tested independently to further demonstrate the predictive ability of the model. In summary, our method, Definer, can accurately identify pseudouridine modification sites in RNA.

PMID:40273178 | DOI:10.1371/journal.pone.0320077