Pharmacogenomics-based subtype decoded implications for risk stratification and immunotherapy in pancreatic adenocarcinoma

Molecular Medicine, Journal Year: 2025, Volume and Issue: 31(1)

Published: Feb. 19, 2025

Abstract Background With fatal malignant peculiarities and poor survival rate, outcomes of pancreatic adenocarcinoma (PAAD) were frustrated by non-response even resistance to therapy due heterogeneity across clinical patients. Nevertheless, pharmacogenomics has been developed for individualized-treatment still maintains obscure in PAAD. Methods A total 964 samples from 10 independent multi-center cohorts enrolled our study. drug response data the profiling relative inhibition simultaneously mixtures (PRISM) genomics sensitivity cancer (GDSC) databases, we established validated multidimensionally three pharmacogenomics-classified subtypes using non-negative matrix factorization (NMF) nearest template prediction (NTP) algorithms, separately. The heterogenous biological characteristics precision medicine strategies among further investigated. Results Three after stable reproducible validation, distinguished six aspects prognosis, peculiarities, immune landscapes, genomic variations, immunotherapy individualized management strategies. Subtype 2 was close immunocompetent phenotype projected immunotherapy; 3 held most favorable metabolic pathways distinctively, promising be treated with first-line agents. 1 worst anticipated chromosome instability (CIN) resistant chemotherapeutic In addition, ITGB6 contributed subtype 5-fluorouracil, knockdown enhanced 5-fluorouracil vitro experiments. Ultimately, appropriate stratified treatments assigned corresponding according pharmacogenomic transcripts. Some limitations not taken into account, thus needs supported more research. Conclusion span-new molecular exploited PAAD uncovered an insight precise medication on ground pharmacogenomics, highly refined multiple specific

Language: Английский

---

Development of a tertiary lymphoid structure-based prognostic model for breast cancer: integrating single-cell sequencing and machine learning to enhance patient outcomes

Frontiers in Immunology, Journal Year: 2025, Volume and Issue: 16

Published: Feb. 26, 2025

Breast cancer, a highly prevalent global poses significant challenges, especially in advanced stages. Prognostic models are crucial to enhance patient outcomes. Tertiary lymphoid structures (TLS) within the tumor microenvironment have been associated with better prognostic We analyzed data from 13 independent breast cancer cohorts, totaling over 9,551 patients. Using single-cell RNA sequencing and machine learning algorithms, we identified critical TLS-associated genes developed TLS-based predictive model. This model stratified patients into high low-risk groups. Genomic alterations, immune infiltration, cellular interactions were assessed. The demonstrated superior accuracy compared traditional models, predicting overall survival. High TLS had higher mutation burden more chromosomal correlating poorer prognosis. High-risk exhibited depletion of CD4+ T cells, CD8+ B as evidenced by bulk transcriptomic analyses. In contrast, checkpoint inhibitors greater efficacy patients, whereas chemotherapy proved effective for high-risk individuals. is robust tool outcomes, highlighting microenvironment's role progression. It enhances our understanding biology supports personalized therapeutic strategies.

Language: Английский

---

Application of artificial intelligence in the diagnosis of malignant digestive tract tumors: focusing on opportunities and challenges in endoscopy and pathology

Journal of Translational Medicine, Journal Year: 2025, Volume and Issue: 23(1)

Published: April 9, 2025

Language: Английский

---

Integrating single-cell and bulk RNA-seq to reveal cholesterol homeostasis-related genes via machine learning to predict prognosis and therapeutic response in hepatocellular carcinoma

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: April 21, 2025

Background: Liver cancer, particularly hepatocellular carcinoma (HCC), has emerged as a significant global health challenge. Recent studies have highlighted cholesterol homeostasis (CH) new research frontier, providing insights into its involvement in diverse biological functions and diseases. This study seeks to investigate the significance of CH context HCC. Methods: This explores CH's role HCC using single-cell RNA sequencing data (GSE140228) from TISCH database, analyzed via "Seurat" R package. Genes associated with were sourced MsigDB database. Utilizing these CH-related genes, we performed unsupervised hierarchical clustering analysis stratify (HCC) molecular subtypes. A comprehensive was conducted on differences among identified clusters, focusing clinical characteristics, pathways, infiltration immune cells. By leveraging score computed various machine learning techniques predict overall survival patients Results: We began by investigating subsequently identifying three distinct risk model developed classify high-score low-score groups. Evaluation tumor microenvironment (TIME) demonstrated that individuals categorized high-risk subgroup showed significantly reduced rates diminished therapeutic efficacy response checkpoint inhibitor treatment regimens. ANXA5, ADH4, ATXN2, ACTG1, MVD, S100A11 essential genes Conclusion: We signature derived offers strong prediction outcomes responses immunotherapy

Language: Английский

---

A Novel Deep Learning–based Pathomics Score for Prognostic Stratification in Pancreatic Ductal Adenocarcinoma

Pancreas, Journal Year: 2025, Volume and Issue: 54(5), P. e430 - e441

Published: Jan. 15, 2025

Background and Objectives: Accurate survival prediction for pancreatic ductal adenocarcinoma (PDAC) is crucial personalized treatment strategies. This study aims to construct a novel pathomics indicator using hematoxylin eosin–stained whole slide images deep learning enhance PDAC prognosis prediction. Methods: A retrospective, 2-center analyzed 864 patients diagnosed between January 2015 March 2022. Using weakly supervised multiple instance learning, pathologic features predicting 2-year were extracted. Pathomics features, including probability histograms TF-IDF, selected through random forests. Survival analysis was conducted Kaplan-Meier curves, log-rank tests, Cox regression, with AUROC C-index used assess model discrimination. Results: The cohort comprised 489 training, 211 validation, 164 in the neoadjuvant therapy (NAT) group. score developed 7 dividing into high-risk low-risk groups based on median of 131.11. Significant differences observed ( P <0.0001). robust independent prognostic factor [Training: hazard ratio (HR)=3.90; Validation: HR=3.49; NAT: HR=4.82; all <0.001]. Subgroup analyses revealed higher rates early-stage NAT responders compared counterparts (both <0.05 surpassed clinical models 1-, 2-, 3-year survival. Conclusions: serves as cost-effective precise tool, functioning an that enables stratification enhances when combined traditional features. advancement has potential significantly impact planning improve patient outcomes.

Language: Английский

---

Machine learning-based neddylation landscape indicates different prognosis and immune microenvironment in endometrial cancer

Frontiers in Oncology, Journal Year: 2023, Volume and Issue: 13

Published: Feb. 22, 2023

Endometrial cancer (EC) is women’s fourth most common malignant tumor. Neddylation plays a significant role in many diseases; however, the effect of neddylation and neddylation-related genes (NRGs) on EC rarely reported. In this study, we first used MLN4924 to affect activation different cell lines (Ishikawa HEC-1-A) determined critical pathways for progression. Subsequently, screened 17 prognostic NRGs based expression files TCGA-UCEC cohort. Based unsupervised consensus clustering analysis, patients with were classified into two patterns (C1 C2). terms prognosis, substantial differences observed between patterns. Compared C2, C1 exhibited low levels immune infiltration promoted tumor More importantly, NRGs, transformed nine machine-learning algorithms 89 combinations. The random forest (RSF) was selected construct risk score according average C-index cohorts. Notably, our had important clinical implications EC. Patients high scores have poor prognoses cold state. conclusion, can distinguish microenvironment (TME) prognosis guide personalized treatment

Language: Английский

---

The resistance to anoikis, mediated by Spp1, and the evasion of immune surveillance facilitate the invasion and metastasis of hepatocellular carcinoma

APOPTOSIS, Journal Year: 2024, Volume and Issue: 29(9-10), P. 1564 - 1583

Published: July 27, 2024

Anoikis-Related Genes (ARGs) lead to the organism manifesting resistance anoikis and are associated with unfavorable prognostic outcomes across various malignancies.Therefore, it is crucial identify pivotal target genes related in HCC .We found that ARGs were significantly correlated prognosis immune responses HCC. The core gene, SPP1, notably promoted metastasis through both vivo vitro studies. PI3K-Akt-mTOR pathway played a critical role suppression within contexts. Our research unveiled SPP1's enhancing PKCα phosphorylation, which turn activated cascade. Additionally, SPP1 was identified as key regulator of MDSCs Tregs migration, directly affecting their immunosuppressive capabilities.These findings indicate HCC, facilitated evasion by modulating Tregs.

Language: Английский

---

Identification and isolation of BZR transcription factor and screening of cell wall degradation marker genes based on machine learning in ripening kiwifruit

Postharvest Biology and Technology, Journal Year: 2024, Volume and Issue: 211, P. 112798 - 112798

Published: Jan. 31, 2024

Language: Английский

---

Machine learning-based nomogram: integrating MRI radiomics and clinical indicators for prognostic assessment in acute ischemic stroke

Frontiers in Neurology, Journal Year: 2024, Volume and Issue: 15

Published: June 12, 2024

Background Acute Ischemic Stroke (AIS) remains a leading cause of mortality and disability worldwide. Rapid precise prognostication AIS is crucial for optimizing treatment strategies improving patient outcomes. This study explores the integration machine learning-derived radiomics signatures from multi-parametric MRI with clinical factors to forecast prognosis. Objective To develop validate nomogram that combines multi-MRI signature predicting prognosis AIS. Methods retrospective involved 506 patients two centers, divided into training (n = 277) validation ( n 229) cohorts. 4,682 radiomic features were extracted T1-weighted, T2-weighted, diffusion-weighted imaging. Logistic regression analysis identified significant risk factors, which, alongside features, used construct predictive clinical-radiomics nomogram. The model’s accuracy was evaluated using calibration ROC curves, focusing on distinguishing between favorable (mRS ≤ 2) unfavorable &gt; Results Key findings highlight coronary heart disease, platelet-to-lymphocyte ratio, uric acid, glucose levels, homocysteine, as independent predictors model achieved ROC-AUC 0.940 (95% CI: 0.912–0.969) in set 0.854 0.781–0.926) set, underscoring its reliability utility. Conclusion underscores efficacy forecasting prognosis, showcasing pivotal role artificial intelligence fostering personalized plans enhancing care. innovative approach promises revolutionize management, offering leap toward more individualized effective healthcare solutions.

Language: Английский

---

Integrative transcriptome-proteome approach reveals key hypoxia-related features involved in the neuroprotective effects of Yang Xue oral liquid on Alzheimer’s and Parkinson’s disease

Frontiers in Pharmacology, Journal Year: 2024, Volume and Issue: 15

Published: July 9, 2024

Introduction: This study investigates the role of hypoxia-related genes in neuroprotective efficacy Yang Xue oral liquid (YXKFY) Alzheimer’s disease (AD) and Parkinson’s (PD). Methods results: Using differential expression weighted gene co-expression network analysis (WGCNA), we identified 106 9 hypoxia-associated AD PD, respectively, that are implicated transcriptomic proteomic profiles. An artificial intelligence-driven hypoxia signature (AIDHS), comprising 17 3 for was developed validated across nine independent cohorts ( n = 1713), integrating 10 machine learning algorithms 113 algorithmic combinations. Significant associations were observed between AIDHS markers immune cells including naive CD4 + T cells, macrophages, neutrophils. Interactions with miRNAs (hsa-miR-1, hsa-miR-124) transcription factors (USF1) also identified. Single-cell RNA sequencing (scRNA-seq) data highlighted distinct patterns various cell types, such as high TGM2 endothelial PDGFRB mesenchymal SYK microglia. YXKFY treatment shown to repair cellular damage decrease reactive oxygen species (ROS) levels. Notably, previously dysfunctional expression, FKBPL, TGM2, PPIL1, BLVRB, PDGFRB, exhibited significant recovery after treatment, associated riboflavin lysicamine. Conclusion: The above suggested be central neuroinflammation responses potential key mediators YXKFY’s action.

Language: Английский

---