Cancer Informatics最新文献

Tumors consist of various cell types, including a small population of cancer stem cells (CSCs), which are linked to metastasis, drug resistance, and recurrence. Maintaining the CSC phenotype requires regulation of molecules, including miRNAs; they are small non-coding RNAs involved in processes such as proliferation, differentiation, invasion, and apoptosis. However, miRNAs involved in CSC generation and maintenance remain largely unidentified. In this study, we aimed to identify miRNAs associated with CSC populations by studying the miRNA profiles in Spheroids-Derived cancer stem cells (SDCSCs) from different cancer cell lines. Firstly, we used small RNA sequencing by Illumina to identify differentially expressed miRNAs from SDCSCs compared with adherent cells from lung cancer cell line. MiRNAs with P < .05 and fold change >1.0 were considered significant. Next, we conducted a meta-analysis to integrate expression data from studies performed under same conditions from several tumor cell lines such as ovarium, breast, colorectal cancer cell lines. We reanalyzed microarrays and RNA sequencing data. For integration we employed the Robust Rank Aggregation approach. We identified only one upregulated miRNA, the hsa-miR-1246 with a P-value 1.6356^-5. hsa-miR-1246 showed consistent overexpression in CSC-enriched spheroids, with fold changes ranging from 2.4 to 3.8 (P < .05) across studies. Bioinformatics analysis revealed that hsa-miR-1246 interacts with cyclins, GSK3B, and other experimentally validated targets, which were related to the cell cycle (FDR 8,40E-03) and the regulation of transcription from RNA polymerase II (FDR 8,30E-03). Our results provide the first integrative study showing that hsa-miR-1246 is consistently overexpressed from cancer stem cells in different tumor cell lines, suggesting a direct link between its oncogenic activity and the CSC phenotype. Since our analysis demonstrates that miR-1246 overexpression is highly specific to CSCs rather than differentiated tumor cells, its detection in patients could serve as an indirect indicator of CSC abundance and tumor aggressiveness. This provides new biological insight into the cellular origin of this miRNA and supports its potential use as a biomarker of stemness and therapeutic target in cancer. Additional studies using in vivo models and functional knockdown experiments will be important to validate these findings and better define the role of hsa-miR-1246 in CSC regulation.

Sample size calculations and power analyses are essential components of experimental design in modern biomedical research. Designs that account for sample correlation, multiple testing, and other sources of variability inherent to specific studies are routinely employed for identifying differential expressions. Despite recent advances in methodologies and software tools for power analysis, there remains a lack of statistical packages capable of accommodating these complex designs in differential expression studies. To fill this gap, we provide the R package depower, which implements the simulation-based framework presented in our recent publications. This unified framework covers both independent and dependent group comparisons and controls false positive rates by employing a simulation-based approach to calculate the empirical null distribution of test statistics.

Objectives: Autophagy-dependent cell death (ADCD) plays a pivotal role in solid tumors, ultimately influencing immunotherapeutic efficacy and cancer prognosis. However, its significance in hepatocellular carcinoma (HCC) remains underexplored.

Methods: Through integrated analysis of single-cell and bulk transcriptomic data, this research systematically identified ADCD-associated genes in LIHC. This was achieved by applying AddModuleScore, ssGSEA, and WGCNA for robust gene screening. A prognostic model was developed for LIHC grounded in The Cancer Genome Atlas (TCGA) dataset. Its validity was confirmed through internal validation with an independent TCGA cohort and external validation using GEO datasets. Immune characteristics were assessed by adopting CIBERSORT and ESTIMATE algorithms. Through LASSO-Cox regression analysis, this research established a 9-gene ADCD signature and derived the ADCD-related risk score system (ADCDRS).

Results: The ADCDRS demonstrated superior prognostic performance. Aside from that, this unique system was significantly associated with clinical features, immune infiltration patterns, and the tumor's local environment. To improve clinical applicability, this research constructed a nomogram incorporating the ADCDRS. Additionally, potential therapeutic agents targeting specific risk subgroups were identified.

Conclusion: This study highlights the prognostic and therapeutic potential of ADCD-related biomarkers in LIHC.

Purpose: Identifying appropriate clinical trials for cancer patients with specific gene mutations remains a significant challenge, largely due to limitations in current search tools like ClinicalTrials.gov, which at times return irrelevant or misleading results. This diagnostic accuracy study investigates the efficacy of 2 large language models (LLMs), GPT-4.0 and Gemini 2.0, in evaluating the eligibility of patients with specific cancer-related gene mutations for clinical trials.

Methods: The study prompts GPT 4.0 and Gemini 2.0 with trial details from ClinicalTrials.gov and a particular cancer mutation. We then assess model performance against physician-curated benchmarks across 6 gene mutations (ALK, BRAF, EGFR, ERBB2, KIT, and KRAS).

Results: The results demonstrate good F1-scores for both LLMs-averaging 64% for GPT-4.0 and 70% for Gemini 2.0-highlighting their potential to streamline clinical trial matching. Furthermore, decision trees provided interpretability by identifying key textual indicators that LLMs use.

Conclusion: This work demonstrates the feasibility of using proprietary LLMs such as GPT 4.0 and Gemini 2.0 "off the shelf" with both limited LLM fine-tuning and limited patient information to evaluate clinical trial eligibility.

Objective: This research aims to develop and evaluate a clinically deployable multimodal deep learning framework for breast cancer diagnosis that maintains robustness, even when clinical data are asynchronous, unpaired, or incomplete, effectively addressing real-world challenges related to data heterogeneity and fragmented clinical workflows.

Methods: In this retrospective study, a multimodal deep learning architecture was developed that integrates histopathological images with structured clinical risk factors. Custom models were developed and independently trained for each modality, and late fusion was achieved via a dynamically reweighted Sinkhorn-based fusion layer. Model performance was evaluated using precision-recall Area Under Curve (PR-AUC), recall, F1 score, and Brier score under complete and partial modality availability scenarios. Robustness and clinical utility were further assessed through statistical significance testing and decision curve analysis (DCA). Additionally, we employed a Sinkhorn cost matrix to enhance interpretability.

Results: The proposed Sinkhorn fusion model outperformed all baseline methods, achieving the highest recall (0.96), PR-AUC (0.775), F1 score (0.828), and the best calibration (Brier score ≈ 0.19). Notably, it maintained perfect recall (1.00) under a 50% simulated modality dropout, despite a significant drop in PR-AUC (20% vs 0%: t = -20.35, P < .0001; 50% vs 0%: t = 88.60, P < .0001), portraying a strong overall robustness to information missingness. Under internally controlled conditions, DCA demonstrated superior clinical utility across thresholds of 0.2 to 0.7.

Conclusions: The model's ability to accommodate unpaired and incomplete clinical inputs while maintaining both calibration and sensitivity makes it particularly well-suited for deployment in asynchronous and resource-constrained settings. Its consistent performance under clinical uncertainty and minimal preprocessing requirements represents a significant advancement toward equitable, reliable, and scalable AI-assisted breast cancer screening. To our knowledge, this is the first paper to model breast cancer late fusion as an optimal transport problem.

Background & aim: Oral squamous cell carcinoma (OSCC) is a devastating disease with poor prognosis and low survival rates, despite advancements in diagnosis and treatment. Early detection and identification of molecular targets are crucial for improving patient outcomes. This study aims to identify differentially expressed genes (DEGs) and key molecular pathways involved in the OSCC. This study's findings will contribute to the development of effective targeted therapies, ultimately improving the prognosis and survival rates of OSCC patients.

Materials & methods: Three gene expression profiles (GSE37991, GSE30784, and GSE107591) from the GEO database were analyzed for differentially expressed genes using EnrichR. Subsequent downstream analyses of the selected module genes were conducted using various bioinformatics tools including STRING, Cytoscape, GEPIA, cBioPortal, NetworkAnalyst, MirWalk, and a bipartite miRNA-mRNA correlation network.

Result: The reanalysis indicated that the Toll-like receptor (TLR) signaling pathway plays a significant role in the development of oral SCC and CXCL8, CCL5, CXCL10, STAT1, IL1B, and TLR2 genes were up-regulated and enriched significantly in the signaling pathways' interactions in oral SCC. Genetic mutation analysis of hub genes in OSCC revealed that STAT1 have 2.5% mutation rate and 0% for other genes. It was revealed that the development and prediction of OSCC may be affected by hsa-mir-146a-5 and hsa-mir-155-5p.

Conclusion: Novel potential biomarkers and signaling pathways associated with OSCC have been identified, which may be important in the transformation of OSCC adenocarcinoma and may serve as therapeutic targets for OSCC.

Objectives: C1QBP is a multi-compartmental protein implicated in diverse cellular processes. However, its clinical predictive value, particularly its association with immune cell infiltration, in lung adenocarcinoma (LUAD) remains unelucidated. Thus, the present study aimed to comprehensively evaluate C1QBP expression patterns, prognostic significance, and its correlation with the tumor immune microenvironment (TIME) in LUAD.

Methods: We first assessed C1QBP expression levels and prognostic relevance in LUAD using multiple bioinformatics platforms. Subsequently, we analyzed the associations of C1QBP expression with immune cell infiltration and immunotherapeutic response, and identified signaling pathways linked to C1QBP expression via Gene Set Enrichment Analysis (GSEA). Finally, enzyme-linked immunosorbent assay (ELISA) was employed to validate the correlation between serum C1QBP concentration and prognosis in non-small cell lung cancer (NSCLC) patients receiving immunotherapy.

Results: C1QBP was highly expressed in LUAD tissues, and this high expression was significantly associated with advanced tumor stage. Moreover, high C1QBP expression emerged as an independent risk factor for overall survival (OS) in LUAD patients. Bioinformatics analyses revealed that C1QBP expression was negatively correlated with the infiltration levels of multiple immune cell subsets (including T cells, B cells, and dendritic cells) in LUAD, while patients with low C1QBP expression exhibited higher Immunophenoscore (IPS). GSEA further demonstrated that high C1QBP expression was positively correlated with pathways regulating the tumor cell cycle, but negatively correlated with immune-related signaling pathways. Finally, in NSCLC patients treated with immune checkpoint inhibitors (ICIs), those with higher serum C1QBP concentrations had significantly shorter OS and progression-free survival (PFS).

Conclusions: Our study identifies C1QBP as a potential oncogene that is closely associated with the TIME in LUAD. Collectively, these findings suggest that C1QBP holds promise as a novel indicator of poor prognosis in LUAD patients.

Background: The THBS gene family plays key functions in various diseases; however, its specific roles in colorectal cancer (CRC) have not been systematically characterized.

Methods: Multi-omics data and online databases were used to analyze the mRNA expression levels of the THBS gene family in CRC and their correlations with clinicopathological features and survival. This analysis identified THBS3 as a potential oncogene closely linked with CRC progression. Then, the relationship between THBS3 expression and the immune landscape was assessed. Single-cell RNA sequencing analyzed THBS3 distribution in CRC subtypes. Additionally, GO, KEGG, and GSEA enrichment analyses investigated the mechanisms of THBS3 in CRC. Molecular docking identified anticancer compounds with high affinity for THBS3. Lastly, in vitro experiments examined THBS3's function in CRC.

Results: THBS3 was significantly upregulated in CRC and correlated with poor prognosis. Elevated THBS3 correlated with increased infiltration of M2 macrophages and regulatory T cells (Treg cells), as well as higher expression of immune checkpoint molecules, suggesting its role in shaping an immunosuppressive microenvironment. THBS3 promoted CRC cell proliferation and metastasis, through activation of the PI3K-AKT and EMT pathways.

Conclusion: THBS3 facilitates the progression of CRC and may serve as a novel prognostic biomarker and therapeutic target.

Objectives: Breast cancer is a heterogeneous disease driven by dysregulated cellular processes, including altered metabolic pathways. The oncogenic microRNA miR-526b influences several cancer hallmark phenotypes and holds promise as a plasma biomarker. Given miR-526b's role in metabolic regulation, we have decided LDHA, PDHA1, ATP5A1, and TIGAR that may help to identify additional biomarkers for breast cancer detection.

Methods: We analyzed mRNA expression of these 4 metabolic markers in breast cancer tissue biopsies and plasma samples from patients and disease-free controls, using publicly available datasets and RT-qPCR validation. Diagnostic performance was evaluated using univariate and multivariate logistic regression and LASSO regression modeling. The potential of combining ATP5A1 with pri-miR-526b expression to improve plasma biomarker accuracy was also assessed.

Results: Individually, none of the metabolic markers demonstrated sufficient sensitivity or specificity as plasma biomarkers. However, combining markers via logistic and LASSO regression improved classification performance. ATP5A1 showed strong biomarker potential in biopsy tissue samples but limited utility in blood plasma. The combination of ATP5A1 with pri-miR-526b significantly enhanced plasma-based diagnostic accuracy, highlighting the value of integrated biomarker panels.

Conclusions: Our study validates the potential of miR-526b-regulated metabolic genes as complementary breast cancer biomarkers. While ATP5A1 shows promise in tissue, plasma-based screening benefits from combining multiple markers, including pri-miR-526b. Further research is needed to refine plasma biomarker panels for effective early detection of breast cancer.

Background and objectives: Next-generation sequencing (NGS) is transforming clinical diagnostics by enabling the detection of genetic variation with unprecedented precision. However, successful implementation of NGS workflows necessitates stringent quality control. This study introduces Molecular Genetics Dashboard (MGDB), a novel bioinformatics tool designed to enhance quality control in clinical NGS workflows.

Methods: Using the Python dash framework for visualizations and MySQL databases, we have developed a novel tool for variant-level monitoring of clinical NGS sequencing runs. MGDB uses a docker-compose containerization for improved portability and can flexibly include or exclude samples from accumulated statistics with notes from interpreters.

Results: MGDB facilitates variant-level run-to-run monitoring, ensuring the consistency of variant detection across sequencing cycles. The tool provides an interactive platform for visualizing and assessing variant data, identifying potential inconsistencies or outliers and improving data management and interpretation compared to traditional methods. MGDB was tested using samples sequenced with Oncomine Focus/Comprehensive Plus assays on S5 sequencers and analyzed via IonReporter software.

Conclusions: MGDB offers a robust and user-friendly solution for enhancing quality control in clinical NGS workflows, contributing to greater accuracy and reliability in variant detection. The tool is freely available on GitHub: https://github.com/acri-nb/GeneticVariantsDB.