Front Oncol. 2025 Jan 10;14:1469427. doi: 10.3389/fonc.2024.1469427. eCollection 2024.

ABSTRACT

PURPOSE: To create a system to enable the identification of histological variants of bladder cancer in a simple, efficient, and noninvasive manner.

MATERIAL AND METHODS: In this multicenter diagnostic study, we retrospectively collected basic information and CT images about the patients concerned from three hospitals. An interactive deep learning-based bladder cancer image segmentation framework was constructed using the Swin UNETR algorithm for further features extraction. Radiomic features and deep learning features were extracted for further stacking ensemble system construction. The segmentation model' performance was assessed by using Dice Similarity (Dice) metrics, Intersection Over Union (IOU), Sensitivity (SEN) and Specificity (SPE). To evaluate the system's performance, we used the Receiver Operating Characteristics (ROC) curve, the Accuracy Score (ACC) and Decision Curve Analysis (DCA).

RESULTS: 410 patients from one hospital were included in the training set, while 60 patients from two other hospitals were included in the test set. A total of 50 features comprising 46 radiomic features and 4 deep learning features were finally retained for further stacking ensemble model building. The interactive segmentation model and system exhibited excellent performance in both training (Dice = 0.78, IOU = 0.65, SEN = 0.83, SPE = 1.00, AUC = 0.940, ACC = 0.868) and testing datasets (Dice = 0.80, IOU = 0.67, SEN = 0.89, SPE = 1.00, AUC = 0.905, ACC = 0.900).

CONCLUSION: We successfully constructed a stacking ensemble machine learning model for early, non-invasive identification of histological variants in bladder cancer which will help urologists make clinical decisions.

PMID:39868365 | PMC:PMC11757263 | DOI:10.3389/fonc.2024.1469427

J Virol. 2025 Jan 27:e0171724. doi: 10.1128/jvi.01717-24. Online ahead of print.

ABSTRACT

The imperative for developing robust tools to detect, analyze, and characterize viruses has become increasingly evident as they continue to threaten human health. In this review, we focus on recent advancements in studying human viruses with flow virometry (FV), an emerging technique that has gained considerable momentum over the past 5 years. These advancements include the application of FV in viral surface phenotyping, viral protein functionality, virus sorting, vaccine development, and diagnostics. With examples illustrated using primary data from our recent studies, we demonstrate that FV is a powerful yet underutilized methodology that, when employed with best practices and experimental rigor, can be highly valuable for studying individual virion heterogeneity, virus phenotypes, and virus-antibody interactions. In this review, we also address the current challenges when performing FV studies, propose strategies to overcome these obstacles, and outline best practices for both new and experienced researchers. Finally, we discuss the promising future prospects of FV within the broader context of virology research.

PMID:39868829 | DOI:10.1128/jvi.01717-24

Biol Aujourdhui. 2024;218(3-4):91-98. doi: 10.1051/jbio/2024009. Epub 2025 Jan 27.

ABSTRACT

The advent of high-throughput omics data and the generation of new algorithms provide the biologists with the opportunity to explore living processes in the context of systems biology aiming at revealing the gene interactions, the networks underlying complex cellular functions. In this article, we discuss two methods for gene network reconstruction, WGCNA (Weighted Gene Correlation Network Analysis) developed by Steve Horvath and collaborators in 2008, and MIIC (Multivariate Information-based Inductive Causation) developed by Hervé Isambert and his team in 2017 and 2024. These two methods are complementary, WGCNA generating undirected networks in which most gene-to-gene interactions are indirect, while MIIC reveals direct interactions and some causal links. We illustrate these aspects according to our own work aiming at identifying the gene interactions underlying the hematopoietic stem cell supportive activity of mesenchymal stromal cells at an early developmental stage.

PMID:39868708 | DOI:10.1051/jbio/2024009

Plant J. 2025 Jan;121(2):e17220. doi: 10.1111/tpj.17220.

ABSTRACT

The traditional Chinese medicinal plant Prunella vulgaris contains numerous triterpene saponin metabolites, notably ursolic and oleanolic acid saponins, which have significant pharmacological values. Despite their importance, the genes responsible for synthesizing these triterpene saponins in P. vulgaris remain unidentified. This study used a comprehensive screening methodology, combining phylogenetic analysis, gene expression assessment, metabolome-transcriptome correlation and co-expression analysis, to identify candidate genes involved in triterpene saponins biosynthesis. Nine candidate genes - two OSCs, three CYP716s and four UGT73s - were precisely identified from large gene families comprising hundreds of members. These genes were subjected to heterologous expression and functional characterization, with enzymatic activity assays confirming their roles in the biosynthetic pathway, aligning with bioinformatics predictions. Analysis revealed that these genes originated from a whole-genome duplication (WGD) event in P. vulgaris, highlighting the potential importance of WGD for plant metabolism. This study addresses the knowledge gap in the biosynthesis of triterpene saponins in P. vulgaris, establishing a theoretical foundation for industrial production via synthetic biology. Additionally, we present an efficient methodological protocol that integrates evolutionary principles and bioinformatics techniques in metabolite biosynthesis research. This approach holds significant value for studies focused on unraveling various biosynthetic pathways.

PMID:39868644 | DOI:10.1111/tpj.17220

Nucleic Acids Res. 2025 Jan 24;53(3):gkaf028. doi: 10.1093/nar/gkaf028.

ABSTRACT

Genome graphs, including the recently released draft human pangenome graph, can represent the breadth of genetic diversity and thus transcend the limits of traditional linear reference genomes. However, there are no genome-graph-compatible tools for analyzing whole genome bisulfite sequencing (WGBS) data. To close this gap, we introduce methylGrapher, a tool tailored for accurate DNA methylation analysis by mapping WGBS data to a genome graph. Notably, methylGrapher can reconstruct methylation patterns along haplotype paths precisely and efficiently. To demonstrate the utility of methylGrapher, we analyzed the WGBS data derived from five individuals whose genomes were included in the first Human Pangenome draft as well as WGBS data from ENCODE (EN-TEx). Along with standard performance benchmarking, we show that methylGrapher fully recapitulates DNA methylation patterns defined by classic linear genome analysis approaches. Importantly, methylGrapher captures a substantial number of CpG sites that are missed by linear methods, and improves overall genome coverage while reducing alignment reference bias. Thus, methylGrapher is a first step toward unlocking the full potential of Human Pangenome graphs in genomic DNA methylation analysis.

PMID:39868538 | DOI:10.1093/nar/gkaf028

Toxicol Rep. 2024 Dec 28;14:101879. doi: 10.1016/j.toxrep.2024.101879. eCollection 2025 Jun.

ABSTRACT

Alzheimer's Disease (AD) is one of the leading neurodegenerative diseases that affect the human population. Several hypotheses are in the pipeline to establish the commencement of this disease; however, the amyloid hypothesis is one of the most widely accepted ones. Amyloid plaques are rich in Amyloid Beta (Aβ) proteins, which are found in the brains of Alzheimer's patients. They are the spliced product of a transmembrane protein called Amyloid Precursor Protein (APP); when they enter into the amylogenic pathway, they get cleaved simultaneously by Beta and Gamma Secretase and produce Aβ protein. Appearances of Amyloid plaques are the significant clinical hallmarks of this disease. AD is mainly present in two genetically distinct forms; sporadic and familial AD. Sporadic Alzheimer's Disease (sAD) is marked by a later clinical onset of the disease, whereas, familial Alzheimer's Disease (fAD) is an early onset of the disease with mendelian inheritance. Several mutations have been clinically reported in the last decades that have shown a direct link with fAD. Many of those mutations are reported to be present in the APP. In this study, we selected a few significant mutations present in the Aβ stretch of the APP and tried to differentiate the wild-type Aβ dimers formed in sAD and the mutant dimers formed in fAD through molecular modelling as there are no structures available from wet-lab studies till date. We analysed the binding interactions leading to formations of the dimers. Our next aim was to come up with a solution to treat AD using the method of drug repurposing. For that we used virtual screening and molecular docking simulations of the already existing anti-inflammatory drugs and studied their potency in resisting the formation of Aβ dimers. This is the first such report of drug repurposing for the treatment of AD, which might pave new pathways in therapy.

PMID:39867516 | PMC:PMC11762949 | DOI:10.1016/j.toxrep.2024.101879

medRxiv [Preprint]. 2025 Jan 14:2025.01.10.25320348. doi: 10.1101/2025.01.10.25320348.

ABSTRACT

BACKGROUND: The development and approval of novel drugs are typically time-intensive and expensive. Leveraging a computational drug repurposing framework that integrates disease-relevant genetically regulated gene expression (GReX) and large longitudinal electronic medical record (EMR) databases can expedite the repositioning of existing medications. However, validating computational predictions of the drug repurposing framework remains a challenge.

METHODS: To benchmark the drug repurposing framework, we first performed a 5-method-rank-based computational drug prioritization pipeline by integrating multi-tissue GReX associated with COVID-19-related hospitalization, with drug transcriptional signature libraries from the Library of Integrated Network-Based Cellular Signatures. We prioritized FDA-approved medications from the 10 top-ranked compounds, and assessed their association with COVID-19 incidence within the Veterans Health Administration (VHA) cohort (~9 million individuals). In parallel, we evaluated in vitro SARS-CoV-2 replication inhibition in human lung epithelial cells for the selected candidates.

RESULTS: Our in silico pipeline identified seven FDA-approved drugs among the top ten candidates. Six (imiquimod, nelfinavir and saquinavir, everolimus, azathioprine, and retinol) had sufficient prescribing rates or feasibility for further testing. In the VHA cohort, azathioprine (odds ratio [OR]=0.69, 95% CI 0.62-0.77) and retinol (OR=0.81, 95% CI 0.72-0.92) were significantly associated with reduced COVID-19 incidence. Conversely, nelfinavir and saquinavir demonstrated potent SARS-CoV-2 inhibition in vitro (~95% and ~65% viral load reduction, respectively). No single compound showed robust protection in both in vivo and in vitro settings.

CONCLUSIONS: These findings underscore the power of GReX-based drug repurposing in rapidly identifying existing therapies with potential clinical relevance; four out of six compounds showed a protective effect in one of the two validation approaches. Crucially, our results highlight how a complementary evaluation-combining epidemiological data and in vitro assays-helps refine the most promising candidates for subsequent mechanistic studies and clinical trials. This integrated validation approach may prove vital for accelerating therapeutic development against current and future health challenges.

PMID:39867394 | PMC:PMC11759241 | DOI:10.1101/2025.01.10.25320348

medRxiv [Preprint]. 2025 Jan 13:2025.01.11.25320392. doi: 10.1101/2025.01.11.25320392.

ABSTRACT

RATIONALE: Individuals homozygous for the Alpha-1 Antitrypsin (AAT) Z allele (Pi*ZZ) exhibit heterogeneity in COPD risk. COPD occurrence in non-smokers with AAT deficiency (AATD) suggests inflammatory processes may contribute to COPD risk independently of smoking. We hypothesized that inflammatory protein biomarkers in non-AATD COPD are associated with moderate-to-severe COPD in AATD individuals, after accounting for clinical factors.

METHODS: Participants from the COPDGene (Pi*MM) and AAT Genetic Modifier Study (Pi*ZZ) were included. Proteins associated with FEV 1 /FVC were identified, adjusting for confounders and familial relatedness. Lung-specific protein-protein interaction (PPI) networks were constructed. Proteins associated with AAT augmentation therapy were identified, and drug repurposing analyses performed. A protein risk score (protRS) was developed in COPDGene and validated in AAT GMS using AUC analysis. Machine learning ranked proteomic predictors, adjusting for age, sex, and smoking history.

RESULTS: Among 4,446 Pi*MM and 352 Pi*ZZ individuals, sixteen blood proteins were associated with airflow obstruction, fourteen of which were highly expressed in lung. PPI networks implicated regulation of immune system function, cytokine and interleukin signaling, and matrix metalloproteinases. Eleven proteins, including IL4R, were linked to augmentation therapy. Drug repurposing identified antibiotics, thyroid medications, hormone therapies, and antihistamines as potential AATD treatments. Adding protRS improved COPD prediction in AAT GMS (AUC 0.86 vs. 0.80, p = 0.0001). AGER was the top-ranked protein predictor of COPD.

CONCLUSIONS: Sixteen proteins are associated with COPD and inflammatory processes that predict airflow obstruction in AATD after accounting for age and smoking. Immune activation and inflammation are modulators of COPD risk in AATD.

PMID:39867385 | PMC:PMC11759610 | DOI:10.1101/2025.01.11.25320392

RSC Adv. 2025 Jan 23;15(4):2298-2316. doi: 10.1039/d4ra08245a. eCollection 2025 Jan 23.

ABSTRACT

Despite significant strides in improving cancer survival rates, the global cancer burden remains substantial, with an anticipated rise in new cases. Immune checkpoints, key regulators of immune responses, play a crucial role in cancer evasion mechanisms. The discovery of immune checkpoint inhibitors (ICIs) targeting PD-1/PD-L1 has revolutionized cancer treatment, with monoclonal antibodies (mAbs) becoming widely prescribed. However, challenges with current mAb ICIs, such as limited oral bioavailability, adverse effects, and high costs, underscore the need to explore alternative small-molecule inhibitors. In this work, we aimed to identify new potential ICI among all FDA-approved drugs. We employed QSAR models to predict PD-1/PD-L1 inhibition, utilizing a diverse dataset of 29 197 molecules sourced from ChEMBL, PubChem, and recent literature. Machine learning techniques, including Random Forest, Support Vector Machine, and Convolutional Neural Network, were employed for benchmarking to assess model performance. Additionally, we undertook a drug repurposing strategy, leveraging the best in silico model for a virtual screening campaign involving 1576 off-patent approved drugs. Only two virtual screening hits were proposed based on the criteria established for this approach, including: (1) QSAR probability of being active against PD-L1; (2) QSAR applicability domain; (3) prediction of the affinity between the PD-L1 and ligands through molecular docking. One of the proposed hits was sonidegib, an anticancer drug, featuring a biphenyl system. Sonidegib was subsequently validated for in vitro PD-1/PD-L1 binding modulation using ELISA and flow cytometry. This integrated approach, which combines computer-aided drug design (CADD) tools, QSAR modelling, drug repurposing, and molecular docking, offers a pioneering strategy to expedite drug discovery for PD-1/PD-L1 axis inhibition. The findings underscore the potential to identify a wider range small molecules to contribute to the ongoing efforts to advancing cancer immunotherapy.

PMID:39867321 | PMC:PMC11755407 | DOI:10.1039/d4ra08245a

ACS Omega. 2025 Jan 8;10(2):1864-1892. doi: 10.1021/acsomega.4c09114. eCollection 2025 Jan 21.

ABSTRACT

Parkinson's disease (PD) is a progressive neurodegenerative disorder that primarily affects movement. It occurs due to a gradual deficit of dopamine-producing brain cells, particularly in the substantia nigra. The precise etiology of PD is not fully understood, but it likely involves a combination of genetic and environmental factors. The therapies available at present alleviate symptoms but do not stop the disease's advancement. Research endeavors are currently directed at inventing disease-controlling therapies that aim at the inherent mechanisms of PD. PD biomarker breakthroughs hold enormous potential: earlier diagnosis, better monitoring, and targeted treatment based on individual response could significantly improve patient outcomes and ease the burden of this disease. PD research is an active and evolving field, focusing on understanding disease mechanisms, identifying biomarkers, developing new treatments, and improving care. In this report, we explore data from the CAS Content Collection to outline the research progress in PD. We analyze the publication landscape to offer perspective into the latest expertise advancements. Key emerging concepts are reviewed and strategies to fight disease evaluated. Pharmacological targets, genetic risk factors, as well as comorbid diseases are explored, and clinical usage of products against PD with their production pipelines and trials for drug repurposing are examined. This review aims to offer a comprehensive overview of the advancing landscape of the current understanding about PD, to define challenges, and to assess growth prospects to stimulate efforts in battling the disease.

PMID:39866628 | PMC:PMC11755173 | DOI:10.1021/acsomega.4c09114

Heliyon. 2025 Jan 3;11(1):e41688. doi: 10.1016/j.heliyon.2025.e41688. eCollection 2025 Jan 15.

ABSTRACT

Basal Cell Carcinoma (BCC) and Actinic Keratosis (AK) are prevalent skin conditions with significant health complications. The molecular mechanisms underlying these conditions and their potential shared pathways remain ambiguous despite their prevalence. Therefore, this study aims to elucidate the common molecular pathways and potential therapeutic targets for BCC and AK through comprehensive computational network analysis. Linkage analysis was performed to identify common liable genes between BCC and AK. Protein-protein interactions (PPIs), Topological properties, GO enrichment, pathway enrichment, and gene regulatory network analyses were also performed to reveal potential molecular mechanisms and pathways. Furthermore, we evaluated protein-drug interactions (PDIs) to identify potential therapeutic targets. Our analysis revealed 22 common genes between BCC and AK: TP53, EGFR, CDKN2A, MMP9, PTGS2, VDR, BCL2, MMP2, EZH2, TP63, FOXP3, MSH2, MMP14, FLG, MC1R, CDKN2B, TIMP3, TYR, SOX10, IRF4, KRT17, and NID1. PPI network analysis highlighted TP53 and EGFR as central hubs, validated using RNA-seq data. Co-expression and physical interaction analysis revealed a strong interplay between the common genes at the transcriptional and functional levels. GO analysis identified skin cancer-relevant terms: "skin development", "immune system development", and "response to radiation" as significantly enriched biological processes, while pathway enrichment analysis highlighted several cancer-related pathways enrichment. Gene regulatory network analysis revealed complex interactions between genes, miRNAs, and transcription factors, with TP53, BCL2, and EGFR playing central roles. PDI network analysis identified ibuprofen as a potential therapeutic agent targeting PTGS2 and BCL2, while other proteins VDR, MMP2, MMP9, and TYR showed interactions with multiple drugs. This computational analysis provides valuable insights into the shared molecular mechanisms of BCC and AK, revealing common pathways and potential therapeutic targets for developing novel treatment strategies and repurposing existing drugs for these prevalent skin cancers. Therefore, these findings may guide future research in understanding and developing targeted therapies for both conditions.

PMID:39866430 | PMC:PMC11761328 | DOI:10.1016/j.heliyon.2025.e41688

Front Mol Neurosci. 2025 Jan 10;17:1503396. doi: 10.3389/fnmol.2024.1503396. eCollection 2024.

ABSTRACT

Multiple sclerosis (MS) affects 2.8 million people worldwide. Although the cause is unknown, various risk factors might be involved. MS involves the immune system attacking the central nervous system's myelin sheath, leading to neuron damage. This study used a cuprizone (CPZ)-intoxicated mouse model to simulate MS's demyelination/remyelination process. It evaluated the molecular, histological, and behavioral effects of vanillic acid (VA), a natural phenolic acid, alone and with Ibudilast (IBD), a clinically tested MS medication. Mice were divided into a control group (regular chow) and a CPZ group (0.3% cuprizone chow for 5 consecutive weeks). During remyelination, the CPZ group was split into four groups: no therapy, 10 mg/kg of IBD, 30 mg/kg of VA, and combined, each treated for 4 weeks. Behavioral, biochemical, molecular, and histopathological tests occurred in the 5th week (demyelination), 7th (early remyelination), and 9th (late remyelination). Cognitive assessments were at weeks 5 and 9. VA enhanced motor, coordination, and cognitive impairments in CPZ-intoxicated mice and improved histopathological, molecular, and biochemical features during early remyelination. IBD improved behavioral abnormalities across all tests, but combined therapy showed no significant difference from single therapies. Further investigations are necessary to understand VA's mechanisms and potential as an MS treatment.

PMID:39866908 | PMC:PMC11760597 | DOI:10.3389/fnmol.2024.1503396

Curr Neuropharmacol. 2025 Jan 24. doi: 10.2174/011570159X349579241231080602. Online ahead of print.

ABSTRACT

INTRODUCTION: Glucagon-Like Peptide-1 Receptor (GLP1R) agonists have become widespread anti-obesity/diabetes pharmaceuticals in the United States.

AIM: This article aimed to provide our current knowledge on the plausible mechanisms linked to the role of Ozempic (Semaglutide), which is generalized as one of the anti-addiction compounds.

METHODS: The effects of GLP1R agonists in Alcohol Use Disorder (AUD) and substance use disorder (SUD) are mediated, in part, through the downregulation of dopamine signaling. We posit that while GLP1R agonism could offer therapeutic advantages in hyperdopaminergia, it may be detrimental in patients with hypodopaminergia, potentially leading to long-term induction of Suicidal Ideation (SI). The alleged posit of GLP1 agonists to induce dopamine homeostasis is incorrect. This study refined 31 genes based on the targets of Ozempic, GLP1R, and related enzymes for SI and 10 genes of the Genetic Addiction Risk Score (GARS) test. STRING-MODEL refined 29 genes, and further primary analyses indicated associations of GLP1R with DRD3, BDNF, CREB1, CRH, IL6, and DPP4.

RESULTS: In-depth silico enrichment analysis revealed an association between candidate genes and depressive phenotypes linked with dopaminergic signaling. Finally, through primary and in-depth silico analyses, we demonstrated multiple findings supporting that GLP1R agonists can induce depression phenotypes.

CONCLUSION: Our findings suggest that associated polymorphisms seem to have overlapping effects with addictive behaviors of Reward Deficiency Syndrome (RDS) and dopamine regulation. Consequently, GLP1R agonists may represent a double-edged sword, potentially triggering both antiaddictive effects and SI by exacerbating depressive phenotypes. Thus, we encourage the scientific community to perform further empirical clinical studies to confirm this proposed pathway.

PMID:39865816 | DOI:10.2174/011570159X349579241231080602

Biofilm. 2024 Dec 30;9:100252. doi: 10.1016/j.bioflm.2024.100252. eCollection 2025 Jun.

ABSTRACT

Antibiotics are central to managing airway infections in cystic fibrosis (CF), yet current treatments often fail due to the presence of Pseudomonas aeruginosa biofilms, settling down the need for seeking therapies targeting biofilms. This study aimed to investigate the antibiofilm activity of aspartic acid and its potential as an adjuvant to tobramycin against P. aeruginosa biofilms formed by mucoid and small colony variant (SCV) tobramycin tolerant strain. We assessed the effect of aspartic acid on both surface-attached and suspended P. aeruginosa biofilms within CF artificial mucus and investigated the synergistic impact of combining it with non-lethal tobramycin concentrations. Our findings showed that aspartic acid inhibited planktonic P. aeruginosa without affecting its viability and prevented biofilm formation by hindering bacterial adhesion or interfering with EPS production, depending on the experimental conditions. In CF mucus, aspartic acid significantly reduced bacterial growth, with the highest inhibition observed when combined with tobramycin, showing notable effects against the mucoid and tolerant SCV strain. Despite these reductions, P. aeruginosa repopulated the mucus within 24 h of stress withdrawal. Additional strategies, including delayed tobramycin application and a second dose of co-application of aspartic acid and tobramycin were explored to address bacterial survival and recovery. Although none of the strategies eradicated P. aeruginosa, the second co-application resulted in slower bacterial recovery rates. In conclusion, this study highlighted aspartic acid as an effective antibiofilm agent and demonstrated for the first time its potential as an adjuvant to tobramycin. The combined use of aspartic acid and tobramycin offers a promising advancement in CF therapeutics, particularly against P. aeruginosa biofilms formed by mucoid and SCV strains, mitigating their antibiotic resistance.

PMID:39866543 | PMC:PMC11759549 | DOI:10.1016/j.bioflm.2024.100252

Heliyon. 2025 Jan 7;11(1):e41716. doi: 10.1016/j.heliyon.2025.e41716. eCollection 2025 Jan 15.

ABSTRACT

BACKGROUND: Cystic fibrosis is a heterogeneous disease whose severity and symptoms largely depend on the functional impact of mutations in the cystic fibrosis transmembrane conductance regulator gene. Other genes may also modulate the clinical manifestations and complications associated with cystic fibrosis. Genetic variants of the bitter taste receptor TAS2R38 have been shown to contribute to the susceptibility and severity of chronic rhinosinusitis. This study aims to elucidate the role of TAS2R38 as a novel modifier gene influencing sinonasal disease severity and pulmonary Pseudomonas Aeruginosa colonization in children with cystic fibrosis.

METHODS: This retrospective observational case-control study evaluated sinus clinical features, quality of life, and the occurrence of Pseudomonas Aeruginosa pulmonary colonization in 69 children with cystic fibrosis. Propylthiouracil testing and TAS2R38 genotyping were performed to characterize patients based on receptor functionality.

RESULTS: The non-taster genetic variant of bitter taste receptor TAS2R38 was associated with greater severity of chronic rhinosinusitis, as measured by endoscopic and radiological scores, compared to the taster variant (p = 0.031 and p = 0.03, respectively). Furthermore, an inverse correlation was observed between the age at first Pseudomonas Aeruginosa infection and chronic rhinosinusitis severity assessed by endoscopic score (r = -0.3388, p = 0.0302).

CONCLUSIONS: The findings highlight the role of TAS2R38 as a potential genetic modifier influencing the severity of chronic rhinosinusitis in children with cystic fibrosis. The clinical implications include the potential development of T2R38-targeted topical therapies and the use of taste testing or genotyping to predict susceptibility to infection. In addition, these results may pave the way for novel, tailored therapeutic approaches in the era of precision medicine.

PMID:39866409 | PMC:PMC11761314 | DOI:10.1016/j.heliyon.2025.e41716

iScience. 2024 Dec 20;28(1):111658. doi: 10.1016/j.isci.2024.111658. eCollection 2025 Jan 17.

ABSTRACT

The regulation of gene expression relies on the coordinated action of transcription factors (TFs) at enhancers, including both activator and repressor TFs. We employed deep learning (DL) to dissect HepG2 enhancers into positive (PAR), negative (NAR), and neutral activity regions. Sharpr-MPRA and STARR-seq highlight the dichotomy impact of NARs and PARs on modulating and catalyzing the activity of enhancers, respectively. Approximately 22% of HepG2 enhancers, termed "repressive impact enhancers" (RIEs), are predominantly populated by NARs and transcriptional repression motifs. Genes flanking RIEs exhibit a stage-specific decline in expression during late development, suggesting RIEs' role in trimming enhancer activities. About 16.7% of human NARs emerge from neutral rhesus macaque DNA. This gain of repressor binding sites in RIEs is associated with a 30% decrease in the average expression of flanking genes in humans compared to rhesus macaque. Our work reveals modulated enhancer activity and adaptable gene regulation through the evolutionary dynamics of TF binding sites.

PMID:39868043 | PMC:PMC11761325 | DOI:10.1016/j.isci.2024.111658

iScience. 2024 Dec 20;28(1):111638. doi: 10.1016/j.isci.2024.111638. eCollection 2025 Jan 17.

ABSTRACT

Over the last decade, Hippo signaling has emerged as a major tumor-suppressing pathway. Its dysregulation is associated with abnormal expression of YAP1 and TEAD-family genes. Recent works have highlighted the role of YAP1/TEAD activity in several cancers and its potential therapeutic implications. Therefore, identifying patients with a dysregulated Hippo pathway is key to enhancing treatment impact. Although recent studies have derived RNA-seq-based signatures, there remains a need for a reproducible and cost-effective method to measure the pathway activation. In recent years, deep learning applied to histology slides have emerged as an effective way to predict molecular information from a data modality available in clinical routine. Here, we trained models to predict YAP1/TEAD activity from H&E-stained histology slides in multiple cancers. The robustness of our approach was assessed in seven independent validation cohorts. Finally, we showed that histological markers of disease aggressiveness were associated with dysfunctional Hippo signaling.

PMID:39868035 | PMC:PMC11758823 | DOI:10.1016/j.isci.2024.111638

Pak J Med Sci. 2025 Jan;41(1):225-229. doi: 10.12669/pjms.41.1.9171.

ABSTRACT

OBJECTIVE: To design a deep learning-based model for early screening of diabetic retinopathy, predict the condition, and provide interpretable justifications.

METHODS: The experiment's model structure is designed based on the Vision Transformer architecture which was initiated in March 2023 and the first version was produced in July 2023 at Affiliated Hospital of Hangzhou Normal University. We use the publicly available EyePACS dataset as input to train the model. Using the trained model, we predict whether a given patient's fundus images indicate diabetic retinopathy and provide the relevant affected areas as the basis for the judgement.

RESULTS: The model was validated using two subsets of the IDRiD dataset. Our model not only achieved good results in terms of detection accuracy, reaching around 0.88, but also performed comparably to similar models annotated for affected areas in predicting the affected regions.

CONCLUSION: Utilizing image-level annotations, we implemented a method for detecting diabetic retinopathy through deep learning and provided interpretable justifications to assist clinicians in diagnosis.

PMID:39867796 | PMC:PMC11755306 | DOI:10.12669/pjms.41.1.9171

J Pathol Inform. 2024 Dec 11;16:100416. doi: 10.1016/j.jpi.2024.100416. eCollection 2025 Jan.

ABSTRACT

BACKGROUND: Traditional liver fibrosis staging via percutaneous biopsy suffers from sampling bias and variable inter-pathologist agreement, highlighting the need for more objective techniques. Deep learning models for disease staging from medical images have shown potential to decrease diagnostic variability, with recent weakly supervised learning strategies showing promising results even with limited manual annotation.

PURPOSE: To study the clustering-constrained attention multiple instance learning (CLAM) approach for staging liver fibrosis on trichrome whole slide images (WSIs) of children and young adults.

METHODS: This is an ethics board approved retrospective study utilizing 217 trichrome WSI from pediatric liver biopsies for model development and testing. Two pediatric pathologists scored WSI using two liver fibrosis staging systems, METAVIR and Ishak. Cases were then secondarily categorized into either high- or low-stage liver fibrosis and used for model development. The CLAM pipeline was used to develop binary classification models for histological liver fibrosis. Model performance was evaluated using area under the curve (AUC), accuracy, sensitivity, specificity, and Cohen's Kappa.

RESULTS: The CLAM models showed strong diagnostic performance, with sensitivities up to 0.76 and AUCs up to 0.92 for distinguishing low- and high-stage fibrosis. The agreement between model predictions and average pathologist scores was moderate to substantial (Kappa: 0.57-0.69), whereas pathologist agreement on the METAVIR and Ishak scoring systems was only fair (Kappa: 0.39-0.46).

CONCLUSIONS: CLAM pipeline showed promise in detecting features important for differentiating low- and high-stage fibrosis from trichrome WSI based on the results, offering a promising objective method for liver fibrosis detection in children and young adults.

PMID:39867463 | PMC:PMC11760786 | DOI:10.1016/j.jpi.2024.100416

Front Neurosci. 2025 Jan 10;18:1434444. doi: 10.3389/fnins.2024.1434444. eCollection 2024.

ABSTRACT

INTRODUCTION: The study of attention has been pivotal in advancing our comprehension of cognition. The goal of this study is to investigate which EEG data representations or features are most closely linked to attention, and to what extent they can handle the cross-subject variability.

METHODS: We explore the features obtained from the univariate time series from a single EEG channel, such as time domain features and recurrence plots, as well as representations obtained directly from the multivariate time series, such as global field power or functional brain networks. To address the cross-subject variability in EEG data, we also investigate persistent homology features that are robust to different types of noise. The performance of the different EEG representations is evaluated with the Support Vector Machine (SVM) accuracy on the WithMe data derived from a modified digit span experiment, and is benchmarked against baseline EEG-specific models, including a deep learning architecture known for effectively learning task-specific features.

RESULTS: The raw EEG time series outperform each of the considered data representations, but can fall short in comparison with the black-box deep learning approach that learns the best features.

DISCUSSION: The findings are limited to the WithMe experimental paradigm, highlighting the need for further studies on diverse tasks to provide a more comprehensive understanding of their utility in the analysis of EEG data.

PMID:39867449 | PMC:PMC11758281 | DOI:10.3389/fnins.2024.1434444