Omics A Journal of Integrative Biology最新文献

Neuroblastoma (NB) is the most common extracranial solid malignancy of children, and MYCN amplification defines a high-risk subtype with poor outcomes. Although widely used in preclinical drug discovery, NB cell lines are often selected based on availability rather than the molecular characteristics of patient-derived tumors, leading to a critical translational gap between experimental outcomes and clinical relevance. To address this, we developed a rank-based transcriptomic correlation framework to assess the concordance between patient-derived tumors (n = 642; combined from the SEQC/MAQC-III and TARGET cohorts) and publicly available NB cell lines (n = 39). This system-level analysis enabled the identification of cell model representatives (CMRs) that closely recapitulate the gene expression landscapes of clinical tumors. COG-N-557, SMS-KAN, and NB-SD emerged as the top CMRs for MYCN-amplified tumors, whereas COG-N-549, FELIX, and SK-N-SH were identified for MYCN-nonamplified tumors. Pathway enrichment analyses indicated that MYCN-amplified CMRs retain key transcriptional programs involved in neuronal development and tumor proliferation, supporting their biological relevance. Leveraging these models, we integrated pharmacogenomic connectivity mapping and drug-gene network analyses to uncover kinase inhibitors and epigenetic modulators as promising therapeutic candidates capable of targeting MYCN-driven transcriptional programs, despite MYCN being an undruggable oncogene. In conclusion, this study addresses a fundamental systems biology and translational research gap by establishing a data-driven framework for selecting NB cell lines that accurately reflect patient-derived tumor biology with direct implications for prioritizing therapeutically relevant drug candidates. Future studies should prioritize the top CMRs as in vitro models to enhance translational relevance and accelerate precision drug discovery in high-risk pediatric NB.

Colon adenocarcinoma (COAD) is a heterogeneous malignancy whose molecular complexity limits effective therapy. Existing transcriptome-based classifications capture only part of this diversity. To refine COAD stratification, we integrated genomic, epigenomic, and transcriptomic data from 297 The Cancer Genome Atlas patients. Ten complementary clustering algorithms were combined through a consensus ensemble framework to ensure robust and unbiased subtype discovery. The resulting molecular subtypes were characterized by genomic alterations, signaling pathways, tumor microenvironment features, and predicted therapeutic responses. As a result, four reproducible molecular subtypes (CS1-CS4) were identified. CS1 displayed enrichment of extracellular matrix organization and epithelial-mesenchymal transition signatures, suggesting invasive potential. CS2 exhibited transcriptional similarity to PD-1 responders, indicating potential benefit from immune checkpoint blockade. CS3 represented a mutation-driven subtype with frequent APC, TP53, and KRAS alterations and extensive copy number gains. CS4 showed the highest immune infiltration, elevated tumor mutational burden, and enhanced sensitivity to 5-fluorouracil and cetuximab. Validation across four independent cohorts confirmed the reproducibility of these subtypes. This integrative multi-omics framework refines the molecular taxonomy of COAD, revealing immunologically active and therapeutically distinct subgroups. The classification not only bridges genomic, epigenomic, and transcriptomic regulation but also provides a practical roadmap for precision oncology by linking molecular features to potential treatment strategies.

Mesial temporal lobe epilepsy (MTLE) is the most common form of drug-resistant epilepsy in adults, yet its molecular pathogenesis remains elusive. While iron dysregulation has been implicated in MTLE, transcriptome-level regulation of iron-related genes in MTLE brain, including regional, subcellular, and pathology-specific patterns, remains largely unexplored. We analyzed publicly available nuclear and cytoplasmic RNA-sequencing data from hippocampal and cortical tissues of patients with MTLE with and without hippocampal sclerosis and controls. We identified differential expression among 562 curated iron-related genes, which constituted 1.46-2.95% of all differentially expressed genes across regions and compartments. These genes showed region- and compartment-specific expression profiles, with recurrent upregulation of CH25H, TAL1, BTG2, TNF, and PTGIS and consistent downregulation of OGFOD3. Protein-protein interaction and hub gene network analysis identified SLC40A1, CH25H, HBB, PTGIS, and CYP2C19 as central hubs linking iron transport, lipid metabolism, oxidative stress, and neuroprotection. Upstream regulatory analysis revealed enrichment of seizure-responsive immediate early genes (EGR2, ATF3, JUN) and neurogenic transcription factors (NEUROD1, ASCL1), with the former upregulated and the latter downregulated, indicating seizure-driven transcriptional reprogramming. Our analyses suggest potential regulatory links connecting iron homeostasis with apoptosis, osmotic balance, cholesterol metabolism, and pH/CO₂ buffering. Exploratory analysis showed a negative association between several iron-related genes, including CYP26B1, and seizure frequency in MTLE. Collectively, these findings reveal complex transcriptional programs governing iron dysregulation in MTLE. The results underscored coordinated regulation of inflammatory and metabolic pathways converging on iron homeostasis and neuronal stress responses in MTLE pathophysiology, providing a systems-level framework for potential prognosis and therapeutic targeting.

Antimicrobial resistance (AMR) is a growing threat in planetary health and demands innovative systems biology strategies for rapid and accurate detection of AMR and attendant resistance phenotypes. Chief among the AMR cases is Pseudomonas aeruginosa that exhibits remarkable genomic adaptability and contributes to multidrug resistance. This study aimed to evaluate the potential of transcriptome-based machine learning (ML) models to predict AMR in P. aeruginosa and attendant gene expression signatures. We integrated transcriptomic profiles of clinical isolates (n = 414) with ML algorithms to predict resistance to four antibiotics: ceftazidime, ciprofloxacin, meropenem, and tobramycin. ML models achieved high predictive accuracy, with the tobramycin model attaining 98.8% accuracy and 100% sensitivity. Each of the four antibiotics yielded distinct transcriptomic signatures enriched in pathways such as biofilm formation, membrane transport, virulence, and amino acid metabolism. Importantly, 10 gene signatures were identified across all four antibiotics, implicating them in core resistance mechanisms including oxidative stress response and iron acquisition. We further identified a core set of 10 mRNAs that are consistently deregulated in resistant isolates across all four drugs, pointing to a shared transcriptional program underpinning multidrug resistance. In conclusion, the transcriptome-based signatures reported herein (1) provide promising candidates for translational research toward development of mechanism-guided diagnostic assays for AMR in P. aeruginosa, and (2) attest to the potential of transcriptome-based ML models to predict AMR. Further studies and validation in independent cohorts are called for.

Regulatory backstage of cancer involves complex multifactorial mechanisms. Among these, the posttranscriptional modulation between microRNA (miRNAs) and mRNAs is implicated as major regulatory mechanisms in different cancers. The tissue- and disease-specific regulation of miRNAs by transcription factors (TFs) further adds to the complexity of this system potentially impacting cancer pathogenesis. To uncover the major TFs impacting miRNA transcription in cancer, the differentially expressed miRNAs and mRNAs in six different cancer types, namely, liver hepatocellular carcinoma, lung adenocarcinoma, prostate adenocarcinoma, stomach adenocarcinoma, breast invasive carcinoma, and colon adenocarcinoma, were compiled from The Cancer Genome Atlas. Using bioinformatics approaches and the TransmiR database, we assembled the 374 TFs that transcriptionally regulate miRNAs through repression mechanisms, and the negatively correlating TF-miR pairs in each cancer type. Importantly, we found that E2F1 and EZH2, primarily linked with cell proliferation and cell cycle regulation, are potential regulators of miRNA transcription in cancers. The comprehensive TF-miR pairs in each cancer type uncovered in this study represent the unique and shared constituents that could functionally affect the cancer pathology. Understanding the mechanisms that regulate miRNA transcription in different cancers could help understand the pathology of cancer from a novel perspective with shared TFs constitutive to multiple cancers. This may also open up new avenues for cancer therapeutic innovation, in which miRNA-based interventions might be able to target and be relevant to multiple cancers at once in the future.

Integrating Big Data, such as large-scale transcriptomic datasets across diseases, continues to be a major challenge. This is in part due to inconsistent preprocessing and the lack of a standardized, reproducible analytical framework. Existing pipelines often rely on manual parameter tuning and fragmented scripts, which limits cross-dataset comparability and downstream interpretability. We developed disint (disease integration and clustering toolkit), an open-source Python framework for standardized cross-dataset expression integration, embedding, and clustering. The pipeline implements housekeeping gene-based normalization, disease-specific log₂ fold-change computation, automated Uniform Manifold Approximation and Projection hyperparameter optimization, and adaptive K-means clustering. Building on its outputs, we further implemented a prototype downstream module, disease reposition, which extracts disease-specific gene signatures, evaluates their shared components, and explores potential drug repositioning candidates. The framework was validated on 28 transcriptomic datasets encompassing 34 disease categories and 386 samples, including 255 patient and 131 healthy control samples, covering 194,182 genes in total. These results highlight the reproducibility, scalability, and translational versatility of our proposed framework.

SLC6A15, (sodium-dependent neutral amino acid transporter B(0)AT2), plays a crucial role in amino acid homeostasis and neuronal signaling, and it has been genetically and transcriptionally associated with major depressive disorder (MDD). However, the post-translational regulation of SLC6A15, particularly through phosphorylation, remains poorly understood. To address this knowledge gap, we report here curation of 3825 global phosphoproteomic datasets and layered statistical and bioinformatic analyses to uncover predominant phosphosites, co-phospho-regulated proteins, upstream kinases, and binary interactors of SLC6A15. Importantly, after stringent filtration, three sites, pS701, pS699, and pS687, were identified as the predominant sites of phosphorylation. Predicted upstream kinases included mitogen-activated protein kinase 3 (MAPK3) and cyclin-dependent kinase 12 (CDK12), suggesting regulatory control at these phosphorylation sites. Binary interactors such as tyrosine-protein kinase Lyn, epidermal growth factor receptor, and calnexin were found to have direct associations with stress-related pathways, indicating potential roles in stress-responsive signaling mechanisms. Pathway enrichment analysis of the high-confidence phosphosites in other proteins revealed significant enrichment of the ErbB signaling pathway, which is frequently dysregulated in MDD. Collectively, this study presents a comprehensive co-phospho-regulation-based catalog of SLC6A15, systematically mapping its key phosphorylation sites, regulatory kinases, and interaction partners. Identification of upstream kinases such as MAPK3 and CDK12 and enrichment of the ErbB signaling axis indicate a potential role of SLC6A15 in synaptic plasticity, neuronal signaling, and stress response mechanisms associated with depression. These findings uncover novel protein-protein relationships and phosphorylation-driven interactions, offering new insights into transporter regulation in neuropsychiatric disorders and potential therapeutic targets for MDD.

Checkpoint kinase 1 (CHEK1/CHK1) is a serine/threonine kinase that is pivotal in maintaining genomic stability by regulating DNA replication, mitotic progression, and DNA damage response (DDR). Phosphorylation at distinct regulatory sites of CHK1 serves as a central signaling switch that tightly coordinates checkpoint control and DNA repair pathways. However, the broad phosphorylation network associated with the DNA repair pathway and CHK1 phosphorylation remains relatively underexplored, representing an untapped avenue with profound therapeutic potential. To bridge this discovery gap, we systematically analyzed global phosphoproteome datasets to visualize CHK1 phosphosites and their coregulated protein phosphosites, providing new insights into the functional networks governed by DDR signaling. The integrative analysis of 577 qualitative and 120 quantitative cellular phosphoproteomic datasets identified signatures of the CHK1 phosphorylation landscape. Our study visualized a strong co-occurrence of DDR-associated phosphosites in proteins, particularly with S280, and the Ataxia telangiectasia and Rad3-related protein-dependent S317 phosphosites in CHK1 located outside its kinase domain. Their coregulation analysis across CHK1 substrates, kinase regulators, and protein interactions uncovered connectivity between CHK1 phosphosites and DDR regulators. Collectively, our phosphosite-concordance approach reported here provides a regulatory map of CHK1 phosphorylation patterns, uncovering unexplored regulatory layers, and highlights new opportunities to explore mechanistic insights into CHK1 phosphoregulation as a target for therapeutic interventions in cancer.

Glioblastoma (GBM) remains one of the most aggressive human malignancies, and its biological diversity is largely shaped by the isocitrate dehydrogenase (IDH) mutation status. Although apoptosis, autophagy, and ferroptosis have each been implicated in GBM pathophysiology, their coordinated regulation through the immunoproteasome (i-PSM) axis has not been systematically explored. Here, we integrated transcriptomic datasets from The Cancer Genome Atlas and The Chinese Glioma Genome Atlas with immune deconvolution and network analyses to map the IDH-specific crosstalk between i-PSM components and cell death programs. We identified a nine-gene signature (i.e., CD14, HMOX1, CTSB, CTSS, RRAS, BAK1, FTH1, PRKCD, and CYBB) that captures IDH-dependent immune and metabolic divergence. IDH-mutant (IDHmt) astrocytomas exhibited coordinated upregulation of HMOX1 and CD14, suggesting an actively maintained, immunosuppressive niche rather than a merely "immune-cold" state. Receiver operating characteristic analyses demonstrated strong discrimination of IDH status (area under the curve ≥ 0.92 for key genes), whereas structure-guided docking nominated rational, multipathway combinations such as temozolomide + bortezomib, luteolin, or lovastatin. Collectively, this integrative omics approach reframes IDHmt GBM as a redox- and myeloid-driven suppressive ecosystem and provides a systems-level rationale for personalized, IDH-informed therapeutic strategies.

Glioma remains a major clinical challenge due to its molecular heterogeneity and limited therapeutic options. While numerous biomarker and drug discovery efforts exist, most are restricted by small sample sizes, subtype-agnostic analyses, or limited integration of computational strategies. Here, we present an integrative machine learning-based systems pipeline for the identification of subtype-specific biomarkers and repurposed therapeutics for glioblastoma (GBM) and low-grade glioma (LGG). We report high-confidence, subtype-specific biomarker candidates by harnessing publicly available gene expression datasets and systematic analyses with oversampling strategies to balance class distributions, followed by feature selection algorithms. Specifically, 10 candidate genes with strong diagnostic potential were identified, including RAB11FIP4, TYRO3, THEM5, SST, SMIM32, MIGA1, ARFGEF3, and ANK3 for GBM and GUCA1A and CES4A for LGG. Repurposed drug candidates were then predicted via signature-based prioritization and evaluated using molecular docking simulations, revealing six promising compounds for GBM (vandetanib, capecitabine, melatonin, agomelatine, ramelteon, and tasimelteon) and one for LGG (ambroxol). This study demonstrates the utility of combining class-balancing, feature selection, and drug repurposing pipelines to uncover clinically relevant glioma biomarkers and therapeutic candidates, thus providing a computational foundation for future experimental and translational validation in these brain cancers and neuro-oncology.