In-hospital cardiac arrest (IHCA) from non-cardiac causes is a life-threatening condition characterized by high mortality and complex etiology. Despite advances in critical care monitoring, effective early warning systems remain lacking due to challenges in balancing sensitivity and specificity within imbalanced clinical data and the interpretability of machine learning models remains poorly defined.
We developed and evaluated six machine learning algorithms using data from 43,618 ICU patients in the MIMIC-IV database. Model performance was assessed through multiple metrics, and SHapley Additive exPlanations (SHAP) analysis was employed for model interpretability. Decision curve analysis was performed to evaluate clinical utility.
Distinct performance trade-offs were identified across models. XGBoost achieved optimal discriminative ability (AUC = 0.730, 95% CI: 0.695–0.763), while CatBoost favored sensitivity (0.907) to meet clinical guidelines. SHAP analysis identified anion gap (AG), white blood cell count (WBC), red cell distribution width (RDW), Glasgow Coma Scale (GCS), platelet, partial thromboplastin time (PTT), and red blood cell count (RBC) as key predictors, with elevated inflammatory markers and reduced neurological parameters consistently associated with increased risk.
Gradient boosting algorithms demonstrate crucial utility in predicting non-cardiac IHCA. The identified interpretable biomarkers align with established pathophysiological mechanisms and may reflect downstream manifestations of genome instability pathways, representing potential targets for early clinical intervention.
Erectile dysfunction (ED) is increasingly recognized as a vascular disorder in which endothelial nitric oxide synthase (eNOS)-dependent signaling is disrupted early and may precede overt cardiovascular disease. In vasculogenic and endothelial dysfunction-dominant forms of ED, accumulating evidence suggests that epigenetic regulation and genome-maintenance pathways contribute to cavernosal endothelial vulnerability. In this review, we synthesize data linking locus-specific DNA methylation—particularly at NOS3/eNOS-related regions—to reduced nitric oxide bioavailability and impaired cavernosal relaxation, while also examining how DNA repair pathways and the DNA damage response may shape endothelial phenotype under oxidative, hypoxic, metabolic, and inflammatory stress. Available evidence supports roles for base excision repair, homologous recombination, PARP-associated signaling, and broader genome-maintenance mechanisms in modulating endothelial resilience, although the strength of support is not uniform across pathways. Direct ED-related evidence is strongest for methylation-associated dysregulation of nitric oxide signaling and for DNA damage-associated impairment in diabetic or radiation-related cavernosal injury models, whereas substantial parts of the proposed methylation–DNA repair–genome stability axis are still inferred from broader vascular and endothelial biology rather than from matched human cavernosal datasets. Recent human single-cell studies of the corpus cavernosum have strengthened the biological rationale for this framework by demonstrating marked cellular and microenvironmental remodeling in diseased tissue, but they do not directly quantify DNA methylation status or pathway-specific DNA repair capacity. Clinically, current practice can phenotype vascular ED using hemodynamic tools, yet standardized approaches for directly assessing methylation-associated repression, narrowed DNA repair reserve, or genome-instability-related stress in human cavernosal tissue remain lacking. Addressing this gap may improve biological stratification and support more mechanism-informed translational studies in ED.
TRIM33α, a bromodomain-containing member of the TRIM protein family involved in protein ubiquitination and transcriptional regulation, plays a critical role in tumor progression, and its dysregulation has been linked to genomic instability and aberrant transcriptional activity in multiple cancers, highlighting its potential as a therapeutic target in oncology. However, selective inhibitors against TRIM33α have not been developed yet.
This study aims to identify novel TRIM33α inhibitors using in silico strategy integrating structure-based pharmacophore modeling, high-throughput virtual screening, molecular docking, and molecular dynamics (MD) simulations.
A pharmacophore model was derived from the crystal structure of TRIM33α and validated using a benchmark set of known ligands. The optimized model, comprising 12 essential features, was applied to screen over 2.7 million compounds from the ChEMBL and ZINC databases. After pharmacophore-based filtering and drug-likeness selection, 1180 candidates were subjected to LibDock-based molecular docking. The top 15 compounds, selected based on LibDock scores, underwent 100 ns MD simulations and MM/GBSA binding free energy analysis for top 4 molecules. Among these, CHEMBL1795351 exhibited higher binding affinity than the reference ligand IACS-9571.
This study identifies promising lead compounds and establishes a robust computational framework for developing TRIM33α-targeted therapies. The identified compounds may serve as valuable starting points for further experimental validation and structural optimization toward the development of novel anticancer agents targeting TRIM33.
Computational identification of TRIM33α inhibitors through virtual screening, molecular docking and MDsimulations.
Colorectal cancer (CRC) continues to represent one of the leading causes of cancer-related morbidity and mortality, globally. Natural phytochemicals from Commiphora wightii (C. wightii) has recently garnered attention as possible anticancer agents due to their multi-targeting mechanisms. The aim of this study was to investigate the molecular mechanisms of inhibition for these phytochemicals against CRC, via an integrated in silico approach.
Potential targets of C. wightii phytochemicals and CRC-associated genes were retrieved from publicly available databases. Overlapping targets were identified using Venn analysis, which were subjected to protein-protein interaction (PPI) network construction and hub gene selection using Cytoscape and CytoHubba. Functional enrichment was performed through Gene Ontology (GO) and KEGG pathway analyses. The expression and prognostic significance of hub genes were evaluated using GEPIA2 and Kaplan–Meier survival analysis. Immune infiltration correlations were assessed using TIMER 3.0, and molecular docking was performed using CB-Dock2 to explore binding affinities of phytochemicals with hub proteins.
A total of 85 overlapping targets were identified between C. wightii phytochemicals and CRC-associated genes. Network analysis revealed ten hub genes, among which MET, CDK1, MMP9, PLAU, and CCND1 were significantly upregulated in CRC tissues. Immune infiltration analysis demonstrated significant association of MMP9 and PLAU with macrophage infiltration in both COAD and READ, suggesting their potential role in modulating the tumor immune microenvironment. Molecular docking revealed strong binding affinities of E-Guggulsterol toward MMP9 (-10.6 kcal/mol) and MET (-9.4 kcal/mol), while Quercetin showed favorable interactions with PLAU (-8.5 kcal/mol), MMP9 (-8.7 kcal/mol), and MET (-8.3 kcal/mol). These findings highlight E-Guggulsterol-MMP9, E-Guggulsterol-MET, Quercetin-PLAU, and Quercetin-MET as promising compound-target pairs for further investigations.
This study identified key CRC-associated targets potentially modulated by C. wightii phytochemicals. Integrated network pharmacology, immune infiltration, and molecular docking analyses highlighted MMP9, PLAU, and MET as biologically relevant targets. E-Guggulsterol and Quercetin demonstrated favorable interactions with these proteins, particularly E-Guggulsterol-MMP9, E-Guggulsterol-MET, Quercetin-PLAU, and Quercetin-MET, suggesting their potential as lead compounds for CRC therapy. Since this work relies on in silico analysis, we call for further in vitro and in vivo studies to support our findings and check if these effects could have therapeutic implications.
The Cleavage and Polyadenylation Specificity Factor (CPSF) complex, a core regulator of mRNA 3′-end processing, plays an increasingly recognized role in maintaining genomic stability beyond its canonical function in pre-mRNA cleavage and polyadenylation. By facilitating transcription termination, CPSF ensures timely dissociation of RNA Polymerase II (Pol II) from gene termini, thereby preventing transcriptional readthrough, R-loop (three-stranded RNA: DNA hybrids with displaced single-stranded DNA) accumulation, and transcription-replication conflicts (TRCs)—all of which are potent drivers of DNA double-strand breaks and genomic instability. Dysregulation of CPSF subunits (e.g., CPSF3, CPSF4, CPSF6) has been observed in various cancers, including acute myeloid leukemia, glioblastoma, and hepatocellular carcinoma, where aberrant CPSF function disrupts transcriptional fidelity and exacerbates replication stress. Protein interaction analyses further suggest that CPSF associates with DNA replication and repair factors, such as RPA, BARD1, and PARP1, positioning it at the interface of RNA processing and genome maintenance. Given the transcriptional dependency of cancer cells, CPSF—particularly its catalytic subunit CPSF3—has emerged as a promising therapeutic target, with preclinical studies demonstrating that CPSF3 inhibitors can induce transcriptional stress and synthetic lethality. This review comprehensively summarizes the structural and functional roles of the CPSF complex in cancer biology and targeted therapy, and explores its emerging function as a guardian of genomic stability.
The physical constraints of lung parenchyma impose solid stress, elevated interstitial fluid pressure, and spatial confinement on expanding non-small cell lung cancer (NSCLC) tumors. This mechanical landscape directly loads the cell nucleus, the primary load-bearing organelle, and compromises nuclear envelope integrity. Here we review how tumor microenvironmental forces propagate through the linker of the nucleoskeleton and cytoskeleton complex to the nuclear lamina, and how downregulation or aberrant splicing of lamin A/C generates localized mechanical vulnerabilities. When mechanical insults exceed the capacity of endogenous repair machinery such as ESCRT-III, transient nuclear envelope rupture occurs. This structural failure exposes chromatin to cytoplasmic nucleases and actomyosin-mediated shearing, serving as an upstream, non-genetic driver of DNA double-strand breaks, micronuclei formation, and chromothripsis. The resulting genomic instability fuels clonal heterogeneity and tyrosine kinase inhibitor resistance. Cytoplasmic release of double-stranded DNA chronically activates cGAS-STING signaling, paradoxically remodeling the tumor microenvironment toward immune evasion. We discuss emerging mechanotherapies that exploit these structural dependencies, including VPS4/ESCRT-III inhibitors, lamin stabilizers, and LINC complex disruptors. Integrating mechanobiological readouts through histology-based deep learning and circulating tumor cell profiling could enable multimodal strategies to target the physical forces that drive genomic instability and therapeutic failure in NSCLC.
Alzheimer’s disease (AD) imposes a substantial public health burden, and socioeconomic status, particularly household income, has been implicated as a key determinant of disease risk. However, the causal nature of this relationship and the mediating mechanisms remain incompletely characterized.
A bidirectional, two‑sample, two‑stage Mendelian randomization (MR) study was conducted using summary-level data from large-scale genome-wide association studies. Genetic instruments for household income (n = 397,751), cognitive performance (n = 257,841), and AD (21,982 cases and 41,944 controls) were extracted. Inverse variance weighted (IVW), MR‑Egger, weighted median, and MR‑PRESSO methods were applied. Mediation analysis was employed to decompose the total effect into direct and indirect (via cognitive performance) components. Robustness was assessed through sensitivity analyses and reverse-direction MR.
Genetically predicted higher household income was associated with a protective effect on AD risk (IVW OR = 0.79, 95% CI: 0.68–0.92). Cognitive performance mediated 65.7% of this total effect. No reverse causal effect of AD on household income was observed. Sensitivity analyses confirmed the robustness of these findings.
Household income exerts a protective causal effect on AD, which is predominantly mediated by cognitive performance. These results elucidate a key mechanistic pathway linking socioeconomic factors to neurodegenerative disease and highlight cognitive function as a potential interventional target, supporting the integration of socioeconomic considerations into AD prevention strategies.
Tumorigenesis occurs due to changes in both coding and non-coding regions of the genome. While researchers have studied protein-coding mutations in detail, the role of non-coding variants is less understood. In this study, we employed a computational framework to investigate changes in the dark DNA, including intronic single nucleotide variants (SNVs) and copy number variations (CNVs), in genes on human chromosome 1 (T2T-CHM13v2.0 assembly). We conducted expression profiling, followed by enrichment analysis, mutational mapping, CNV annotation, and drug-gene interaction studies. To fine-tune the Notch pathway, we docked NOTCH2 harboring an intronic single nucleotide variant (SNV) with six curcuma molecules and recorded the binding affinity. Protein-protein interaction (PPI) networks, along with enrichment analysis, identified 9 genes involved in various oncogenic pathways. The hubs identified were pivotal for cancer progression. CNVs that affect regulatory elements, especially promoters, enhancers, and epigenetically modified accessible regions, were strongly linked to changes in the expression of NOTCH2 and NRAS. Drug-gene interaction analysis highlighted Tazarotene (TP73) and Odevixibat (NOTCH2) as high-priority candidates for therapeutic repurposing. Docking studies revealed promising binding conformations and strong interactions between the modeled NOTCH2 protein and selected ligands, suggesting their potential as modulators of oncogenic pathways. Collectively, these findings underscore that non-coding structural alterations play a critical role in regulating cancer pathways and represent promising biomarkers and therapeutic targets in solid tumors.