Abstract
Background
CH25H encodes an interferon-inducible hydroxylase that converts cholesterol to 25-hydroxycholesterol (25-HC), linking lipid homeostasis to immune regulation. Its pan-cancer relevance remains incompletely defined.
Methods
We integrated TCGA/GTEx transcriptomes with TCGA copy-number, mutation and DNA-methylation data, and immune infiltration estimates (TIMER/MCP-COUNTER/ssGSEA). For each cancer type, overall survival was modeled by an identical multivariable Cox regression (high vs. low CH25H by within-cancer median) adjusting for age, sex and pathologic stage; proportional hazards were assessed.
Results
CH25H was dysregulated across multiple malignancies with tumor-type-specific directions. Copy-number losses and promoter hypermethylation jointly associated with reduced mRNA levels. Survival analyses revealed a context-dependent pattern: higher CH25H aligned with favorable outcomes in several immune-active tumors, but with poorer prognosis in selected gastrointestinal cancers. CH25H positively correlated with effector immune infiltration and IFN-γ-dominant/inflammatory immune subtypes. These directions were concordant with multivariable Cox estimates; full per-cancer results are provided in Supplementary Table S1.
Conclusions
CH25H exhibits a context-dependent, biphasic role in cancer: on one hand, it may exert tumor-suppressive effects by enhancing anti-tumor immunity; on the other, it can promote tumor progression in specific cancer types or stages. Its expression is co-regulated by CNV and promoter methylation. CH25H and its enzymatic product, 25-HC, hold promise as both prognostic biomarkers and therapeutic targets in cancer immunotherapy.
Supplementary Information
The online version contains supplementary material available at 10.1007/s12672-026-04464-9.
Introduction
Tumor cells exhibit a marked dependency on de novo lipid synthesis and homeostasis to support rapid proliferation and invasion, and cholesterol metabolism reprogramming has emerged as a critical driver of tumorigenesis, metastasis, and immune evasion [1, 2]. Cholesterol-25‐hydroxylase (CH25H), an interferon‐stimulated gene product localized to the endoplasmic reticulum, catalyzes the conversion of cholesterol into 25‐hydroxycholesterol (25‐HC) [3]. 25‐HC exerts feedback inhibition on HMG‐CoA reductase to limit cholesterol biosynthesis and activates liver X receptors (LXRs) to promote cholesterol efflux and regulate inflammatory gene transcription [4], thereby serving as a pivotal link between metabolic homeostasis and immune defense [5]. In addition, 25‐HC participates in antiviral immunity by blocking enveloped virus membrane fusion and modulating autophagy pathways [6], thereby serving as a pivotal link between metabolic homeostasis and immune defense.
Recent in vitro and in vivo studies in individual cancer models have revealed a context-dependent, “biphasic” role for the CH25H-25‐HC axis in oncology [7]. In pancreatic ductal adenocarcinoma and melanoma, upregulation of CH25H or exogenous administration of 25‐HC depletes tumor cell cholesterol, activates the LXR-ABCA1 pathway, and enhances CD8⁺ T-cell‐mediated antitumor immunity, resulting in suppressed tumor growth [8]. Conversely, in inflammation‐dependent tumors such as colorectal cancer, 25‐HC promotes epithelial-mesenchymal transition and metastasis via TLR2-NF-κB or GPR183-cAMP signaling cascades, upregulating proinflammatory cytokines (e.g., IL-6, CXCL8) [9]. These findings underscore that the biological functions of the CH25H-25-HC axis are highly contingent on the metabolic and inflammatory milieu of the tumor microenvironment.
Despite these insights, comprehensive pan-cancer analyses that integrate transcriptomic, genomic, epigenetic, and immune infiltration multi-omics data to systematically delineate CH25H expression patterns, genomic and methylation-mediated regulation, prognostic significance, and immune associations are lacking [10]. Moreover, the prevalence of CH25H copy number alterations, functional mutations, and promoter methylation in clinical cohorts—and their mechanistic impacts on 25-HC biosynthesis and tumor immunity—remain poorly defined [11]. To date, no pan-cancer study has systematically integrated transcriptome, copy number, mutation, and promoter methylation with immune infiltration to delineate the CH25H–25-HC axis. Our multi-omics design explicitly addresses this gap by linking genomic/epigenetic alterations to expression, prognosis, and immunologic context across 33 cancers under a uniform analytical framework.
To address these gaps, we integrated RNA-seq data from The Cancer Genome Atlas (TCGA) and the Genotype-Tissue Expression (GTEx) project, genomic alteration profiles from cBioPortal, DNA methylation data from TCGA, and immune infiltration estimates generated by TIMER-2.0, MCP-COUNTER, and ssGSEA [12]. Our study aims to (i) characterize CH25H expression dysregulation and molecular subtype differences across 33 cancer types; (ii) elucidate the combined regulatory influences of copy-number variation and epigenetic modification on CH25H expression; (iii) assess the prognostic value of CH25H for overall survival using a uniform within-cancer median split; and (iv) explore the coupling between the CH25H–25-HC axis and the tumor immune microenvironment.
Materials and methods
Data sources
Multi-omics datasets covering 33 TCGA cancer types were retrieved from public repositories. Transcriptomic data were obtained from the TCGA-Toil re-computed matrix (10,854 tumor samples and 730 matched normal tissues) [13] and the GTEx v8 normal tissue RNA-seq dataset (8,587 samples) [14]. Copy number alteration (CNA), somatic mutation (MAF), and structural variation (SV) files were batch-downloaded from the cBioPortal April 2024 snapshot, comprising 10,711 samples. DNA methylation β-value matrices (Level-3) were sourced from the TCGA Illumina HumanMethylation450K platform (9,663 samples).Clinical follow-up information was drawn from the TCGA Pan-Cancer Clinical Data Resource (PCDC/TCGA Pan-Cancer Clinical Matrix). Immune infiltration and tumor purity estimates were generated using TIMER-2.0, MCP-COUNTER (via immunedeconv v2.0), single-sample gene set enrichment analysis (ssGSEA) based on the Bindea 28-gene signature (GSVA v1.50), and ESTIMATE v1.0.13. All raw datasets are accessible via the official database portals and our GitHub repository.
Data preprocessing
Transcriptome: TPM matrices downloaded from UCSC Xena were merged, and CH25H expression (Ensembl ID ENSG00000160206) was extracted. Values were log₂-transformed as log₂(TPM + 1).
Copy Number Alteration: GISTIC2 thresholds were applied to CNA data: “−2” denoted deep deletion and “+2” denoted high-level amplification.
DNA Methylation: We selected Illumina 450 K probes annotated to the CH25H promoter region (TSS1500, TSS200, and first exon). For each sample, promoter methylation was summarized as the per-sample arithmetic mean β-value across all retained promoter probes (i.e., averaging within sample across probes; not averaging across samples). Probes with missing β were excluded on a per-sample basis; if < 2 promoter probes remained, the sample was excluded from promoter-methylation analyses.
Survival Data: Overall survival (OS) was defined as the primary endpoint. If days_to_death was missing, days_to_last_followup was used for right-censoring.We performed complete-case analyses for the Cox models and reported cohort-wise sample sizes and event counts in Supplementary Table S1.
Immune Infiltration: Scores from TIMER-2.0, MCP-COUNTER, and ssGSEA were Z-score normalized across cancer types to facilitate comparative analyses [15–17].
Statistical analysis
For each TCGA cancer type, CH25H expression was dichotomized at the within-cohort median (High vs. Low). Overall survival (OS) was modeled using multivariable Cox proportional hazards with covariates age (continuous), sex (sex-specific tumors excluded), and pathologic stage (I-IV; where missing, available T/N/M were used). The proportional hazards assumption was assessed by Schoenfeld residuals for each cohort. Group comparisons used Wilcoxon rank-sum (unpaired), paired Wilcoxon (paired tumor-normal), or Kruskal-Wallis (multi-group) tests as appropriate. Multiple testing was controlled by the Benjamini-Hochberg false discovery rate (FDR).For subtype-level and immune-subtype comparisons, subgroups with n < 10 were excluded from statistical testing; when either HM-SNV or HM-indel subgroup had n < 10, the two hypermutated categories were merged a priori for robustness and reported as ‘HM (combined)’. Exact per-subgroup sample sizes are annotated on plots.
Software environment and reproducibility
Analyses were performed in R (v4.3) and Python (3.11) with standard scientific libraries; exact package versions and scripts are provided in the GitHub repository.
Results
Transcriptional heterogeneity of CH25H across tumor types. Pan-cancer analysis of 33 TCGA malignancies revealed significant CH25H expression differences between tumor and paired normal tissues in 16 cancer types (FDR-q < 0.05; Fig. 1A-C). CH25H was markedly downregulated in breast carcinoma (BRCA), bladder urothelial carcinoma (BLCA), and thyroid carcinoma (THCA), whereas it was upregulated in colorectal adenocarcinoma (COAD), liver hepatocellular carcinoma (LIHC), and ovarian serous cystadenocarcinoma (OV). Paired-sample Wilcoxon tests confirmed the robustness of these alterations (Fig. 1B). Subtype-specific analysis further demonstrated that the Basal subtype of BRCA exhibited the lowest CH25H expression—1.21 log₂(TPM + 1) lower than Luminal A (q = 3.5 * 10⁻⁶)—while LIHC iCluster 3 showed the highest expression, 0.86 log₂(TPM + 1) above iCluster 1 (q = 1.1 * 10⁻⁴; Fig. 1D), highlighting fine-tuned regulation by tumor molecular context.
Fig. 1.
Pan-cancer transcriptional landscape of CH25H. A Boxplots of CH25H mRNA expression (log₂(TPM + 1)) in 33 cancer types, combining TCGA tumor samples (green) and corresponding normal tissues from TCGA/GTEx (blue). Asterisks denote statistical significance by unpaired Wilcoxon rank-sum test: *P < 0.05; P < 0.01; *P < 0.001; ns, not significant.Sample sizes are displayed for each box/bar (tumor vs. normal) per cancer type. B Paired tumor vs. adjacent normal in 15 cancers (paired Wilcoxon). Per-group sample sizes (n) are labeled. C Bar plots of mean CH25H (log₂(TPM + 1)) in tumor (red) vs. normal (black). Per-group sample sizes (n) are labeled. D Violin plots across subtypes in four cancers: BRCA (Basal, HER2, Luminal A, Luminal B, Normal-like); COAD (CIN, GS, HM-SNV, HM-indel); LIHC (iCluster 1/2/3); STAD (CIN, EBV, GS, HM-SNV, HM-indel). Subtype comparisons: Kruskal–Wallis; P values and n are indicated
Genomic and epigenetic regulation of CH25H. Overall CH25H copy-number alteration (CNA) frequency was low (median 1.7%), with the highest rate observed in BLCA (6.2%), predominantly deep deletions and low-level gains (Fig. 2A). The somatic mutation rate was below 0.5%, and recurrent missense variants (e.g., R265Q/W, I202T) localized to the FA_hydroxylase catalytic domain (Fig. 2B), suggesting potential impacts on enzymatic activity. Promoter methylation analysis showed that cholangiocarcinoma (CHOL) and COAD tumors exhibited mean β-value increases of 0.11 and 0.18, respectively, relative to normal tissues, and methylation levels correlated inversely with CH25H mRNA (ρ = −0.46, P < 0.001; Fig. 2D). These data suggest that copy-number loss and promoter hypermethylation are each associated with, and in combination may reinforce, reduced CH25H mRNA, although causal relationships require orthogonal validation.
Fig. 2.
Genomic Alterations and Promoter Methylation of CH25H in Pan-Cancer. A Summary of CH25H genomic alterations across 33 TCGA cancer types from cBioPortal (April 2024). Bar height indicates the total alteration frequency (%); colors denote mutation (green), amplification (red), and deep deletion (blue). The “+/-” matrix below indicates the availability of structural variation (SV), somatic mutation (Mut), and copy-number alteration (CNA) data for each cohort. B Lollipop plot of CH25H protein (272 amino acids), showing the distribution and recurrence of missense mutations. The height of each green lollipop corresponds to the number of samples harboring that variant; the FA_Hydroxylase catalytic domain and recurrent sites (e.g., R265Q/W, I202T) are annotated. C Heatmap of Spearman correlation coefficients (ρ), rows and columns hierarchically clustered using 1 − ρ (distance) and average linkage; dendrograms shown, cancer types ordered accordingly. D Boxplots comparing promoter-region DNA methylation (mean β-value) of CH25H in cholangiocarcinoma (CHOL; normal n = 9, tumor n = 36) and colorectal adenocarcinoma (COAD; normal n = 37, tumor n = 313). Paired Wilcoxon signed-rank tests yielded P = 1.58 × 10⁻³ for CHOL and P < 1 × 10⁻¹² for COAD
Prognostic significance of CH25H. Survival analyses revealed cancer-type-dependent effects of CH25H expression. In BRCA and non-small cell lung cancer (NSCLC; LUAD + LUSC), high CH25H expression was associated with improved overall survival (HR = 0.68 and 0.78, respectively; Fig. 3A, D). Conversely, elevated CH25H predicted poorer outcomes in COAD (HR = 1.40) and stomach adenocarcinoma (STAD; HR = 1.29; Fig. 3B, C). Analysis by tumor stage showed that CH25H expression increased with advancing pathological stage in THCA but was lowest at Stage IV in BLCA (Fig. 3E–F). In contrast, no significant differences were observed across T stages in BRCA (Fig. 3G) or COAD (Fig. 3H).An identical multivariable Cox model (High vs. Low by within-cancer median; covariates: age [continuous], sex [sex-specific tumors excluded], and pathologic stage) was applied consistently across all cancer types. Supplementary Table S1 reports, for each cancer: N, events, HR (High vs. Low), 95% CI, Wald P, BH-FDR q, PH test P (Schoenfeld), covariates used, and notes (e.g., insufficient events).
Fig. 3.
Prognostic Value of CH25H Expression and Its Association with Clinical Stage. A–D Kaplan-Meier overall survival curves stratified by the median CH25H expression level (high = red; low = black) in four cohorts: (A) breast invasive carcinoma (BRCA), (B) colon adenocarcinoma (COAD; COLON cohort), (C) stomach adenocarcinoma (STAD; GASTRIC cohort), and (D) non-small cell lung cancer (LUAD + LUSC combined). Hazard ratios (HRs) with 95% confidence intervals and log-rank P values are indicated on each plot. The number of patients at risk at each time point is shown below the curves.Hazard ratios (HRs) and 95% CIs are from the same multivariable Cox model (age, sex, stage); full per-cancer estimates are reported in Supplementary Table S1. Log-rank P values are from KM analyses. E–H CH25H expression across clinical stages: (E) Pathologic stage I-IV in thyroid carcinoma (THCA). F Pathologic stage I-IV in bladder urothelial carcinoma (BLCA). G T stage (T1-T4) in breast invasive carcinoma (BRCA). H Pathologic stage I-IV in colon adenocarcinoma (COAD). Boxplots depict the median, interquartile range, and whiskers extending to 1.5× the interquartile range. Statistical comparisons were performed using the Kruskal-Wallis test; significance annotations follow the conventions described in Figure 1.
Association with the immune microenvironment. CH25H expression positively correlated with infiltration of CD8⁺ T cells, NK cells, and M1 macrophages (mean ρ ≈ 0.30), and negatively with immunosuppressive populations such as Tregs and MDSCs (Fig. 4A). In LIHC and BLCA, high CH25H levels were accompanied by lower tumor purity (ρ ≈ −0.25) and increased CD8⁺ T-cell infiltration (ρ ≈ 0.35, P < 0.001; Fig. 4B). Immune subtype analysis confirmed that CH25H expression was highest in the IFN-γ-dominant (C2) and inflammatory (C3) subtypes, and lowest in the immune-silent (C5) subtype (q < 0.001; Fig. 4C), consistent with its classification as an interferon-stimulated gene.
Fig. 4.
Interaction between CH25H Expression and the Tumor Immune Microenvironment. A Heatmap of Spearman ρ, with hierarchical clustering (1 − ρ, average linkage) applied to both axes; dendrograms and clustered ordering are displayed. B Representative correlation scatter plots illustrating the relationship between CH25H expression and tumor purity (ESTIMATE) as well as CD8⁺ T-cell infiltration (TIMER) in four cancer types: bladder urothelial carcinoma (BLCA) and liver hepatocellular carcinoma (LIHC) in the upper panels; breast invasive carcinoma (BRCA) and head and neck squamous cell carcinoma (HNSC) in the lower panels. Shaded areas represent 95% confidence intervals; Spearman ρ and P-values are indicated in the upper right of each plot. C Violin plots of CH25H expression across six immune molecular subtypes (C1–C6, per Thorsson et al., Immunity 2018: wound healing, IFN-γ dominant, inflammatory, lymphocyte depleted, immunologically quiet, TGF-β dominant)
Discussion
In this study, we present the first pan-cancer, multi-dimensional characterization of the CH25H-25-HC axis by integrating transcriptomic, genomic, epigenetic, and immune infiltration data across 33 tumor types [12]. Collectively, these observations support a context-dependent role. All associations reported here are correlative rather than causal; the therapeutic propositions herein should be regarded as hypothesis-generating. Our findings demonstrate that CH25H expression is significantly dysregulated in nearly half of the analyzed cancers, with downregulation in BRCA, BLCA, and THCA, and upregulation in COAD, LIHC, and OV. Subtype-level analyses further revealed that CH25H expression is fine-tuned by the tumor’s inflammatory and metabolic context, being lowest in subtypes lacking active interferon signaling and highest where interferon and lipid metabolism programs are enriched [18].
Genomic and epigenetic analyses uncovered a dual-hit suppression mechanism for CH25H. Although overall CNA and mutation rates were low, BLCA displayed a 6.2% deep deletion frequency [19], and missense mutations within the FA_hydroxylase domain (e.g., R265Q/W, I202T) may compromise 25-HC synthesis [20]. Concurrently, promoter hypermethylation in CHOL and COAD strongly correlated with reduced mRNA expression (ρ = −0.46), supporting an ATF3-DNMT1-mediated silencing model [21]. This “genome + epigenome” convergence not only stabilizes CH25H repression but also provides a molecular basis for tumor evasion of cholesterol-mediated metabolic surveillance and immune control [22].Notably, the FA_hydroxylase catalytic core encompasses the HxH motif and nearby residues required for di-iron coordination. Substitutions such as R265Q/W and I202T fall proximal to conserved elements and are predicted to perturb Fe²⁺ binding or substrate positioning, potentially lowering 25-HC catalytic efficiency; future enzymatic assays and structural modeling are warranted.
Consistent with recent findings identifying disulfidptosis-related lncRNAs as prognostic indicators for immunotherapy response and chemotherapy sensitivity in colon cancer [23], the prognostic impact of CH25H was cancer-dependent: high expression conferred survival benefits in BRCA and NSCLC but predicted adverse outcomes in COAD and STAD. Integration with immune-infiltration data revealed that elevated CH25H often coincides with enrichment of effector populations (CD8⁺ T cells, NK cells, M1 macrophages) and reduced tumor purity, suggesting that 25-HC may potentiate antitumor immunity by depleting cellular cholesterol and activating LXR-ABCA1-mediated cholesterol efflux, thereby augmenting the efficacy of PD-1/PD-L1 blockade [24]. In contrast, in chronic-inflammation-driven settings, excess 25-HC can engage GPR183 or TLR2-NF-κB pathways to upregulate CXCL8 and IL-6, promoting gastrointestinal tumor progression and metastasis.
Collectively, these observations support a context-dependent role of the CH25H-25-HC axis. The therapeutic propositions herein should be regarded as hypothesis-generating; rigorous functional validation—for example, quantitative 25-HC assays and CRISPR-based perturbations—is required to test causality.
This study has limitations: (1) reliance on bulk RNA-seq precludes distinction between tumor and stromal/immune cell CH25H sources; single-cell or spatial transcriptomics are needed for cellular resolution; (2) limited normal sample sizes in certain cancers may underestimate expression differences; (3) our analyses are correlative—functional validation via CRISPR-KO/KI models and quantitative 25-HC mass spectrometry is required to establish causal links between CH25H alterations and downstream immune regulation.
Conclusion
The CH25H-25-HC metabolic pathway functions as a context-dependent nexus of tumor metabolism and immunity, exhibiting both tumor-suppressive and tumor-promoting roles across distinct cancer types. Its expression is coordinately regulated by genomic and epigenetic mechanisms and tightly coupled to the immune microenvironment. Precision modulation of this axis—either activation or inhibition—tailored to tumor type and microenvironmental context, holds promise for improving the efficacy of combined metabolic-immunotherapeutic strategies. Rigorous functional studies (e.g., quantitative 25-HC assays and CRISPR-based perturbations) are required to test causality.
Supplementary Information
Author contributions
CRediT authorship contribution statementZhongqi Diao : Writing – Original Draft; VisualizationYi Chen : Data Curation; ConceptualizationWenjia Guo : Investigation; Formal Analysis; Funding Acquisition; Supervision; Conceptualization.
Funding
This work was supported by the Scientific Research Innovation Team Project of Xinjiang Medical University (XYD-2024C09) and the Key R&D Task Special Project of the Autonomous Region (2022B03019-4).
Data availability
All datasets used in this study are publicly available from their official portals (e.g., UCSC Xena for TCGA/GTEx; cBioPortal for somatic alterations; TIMER/MCP-COUNTER for immune estimates). Accession identifiers, access dates, and processing steps are listed in Methods and Supplementary Table S1. No controlled-access individual-level data were generated in this work.
Declarations
Ethics approval and consent to participate
The data used is from public dataset, and the original data has obtained ethical approval. Not applicable.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Nong S, Han X, Xiang Y, et al. Metabolic reprogramming in cancer: mechanisms and therapeutics. MedComm. 2023;4(2): e218. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Chen X, Chi H, Zhao X, et al. Role of exosomes in immune microenvironment of hepatocellular carcinoma. J Oncol. 2022;2022:2521025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Seo YS, Cho IA, Kim TH, et al. Oxysterol 25-hydroxycholesterol as a metabolic pathophysiological factors of osteoarthritis induces apoptosis in primary rat chondrocytes. Korean J Physiol Pharmacol. 2020;24(3):249–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Hayashi T, Kotani H, Yamaguchi T, et al. Endothelial cellular senescence is inhibited by liver X receptor activation with an additional mechanism for its atheroprotection in diabetes. Proc Natl Acad Sci U S A. 2014;111(3):1168–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Song B, Wu P, Liang Z, et al. A novel Necroptosis-Related gene signature in skin cutaneous melanoma prognosis and tumor microenvironment. Front Genet. 2022;13:917007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Weinhofer I, Buda A, Kunze M, et al. Peroxisomal very long-chain fatty acid transport is targeted by herpesviruses and the antiviral host response. Commun Biol. 2022;5(1):944. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Bao S, Zhou T. A blood-based MiRNA signature for early non-invasive diagnosis of preeclampsia. BMC Med. 2022;20(1):303. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Zheng Y, Xu R, Chen X, et al. Metabolic gatekeepers: harnessing tumor-derived metabolites to optimize T cell-based immunotherapy efficacy in the tumor microenvironment [J]. Cell Death Dis. 2024;15(10):775. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Engevik MA, Danhof HA, Ruan W, et al. Fusobacterium nucleatum secretes outer membrane vesicles and promotes intestinal inflammation. mBio. 2021;12(2):1–17. 10.1128/mBio.02706-20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.He K, Xie M, Li J, et al. CENPO is associated with immune cell infiltration and is a potential diagnostic and prognostic marker for hepatocellular carcinoma. Int J Gen Med. 2022;15:7493–510. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Chen C, Gao D, Huo J, et al. Multiomics analysis reveals CT83 is the most specific gene for triple negative breast cancer and its hypomethylation is oncogenic in breast cancer. Sci Rep. 2021;11(1):12172. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Wang X, Lin R, Sun C, et al. Identification of PPFIA3 as a potential prognostic and immune biomarker through pan-cancer analysis. Organ Med. 2025;2(2):103–17. [Google Scholar]
- 13.Vivian J, Rao AA, Nothaft FA, et al. Toil enables reproducible, open source, big biomedical data analyses. Nat Biotechnol. 2017;35(4):314–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.GTEx Consortium, Aguet F. The GTEx consortium atlas of genetic regulatory effects across human tissues. Science. 2020;369(6509):1318–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Chèneby J, Ménétrier Z, Mestdagh M, et al. ReMap 2020: a database of regulatory regions from an integrative analysis of human and Arabidopsis DNA-binding sequencing experiments. Nucl Acid Res. 2019;48(D1):D180-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Barbie DA, Tamayo P, Boehm JS, et al. Systematic RNA interference reveals that oncogenic KRAS-driven cancers require TBK1. Nature. 2009;462(7269):108–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Hagman J. Plasma cells for hire: prior experience required. Immunity. 2013;39(1):89–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Thorsson V, Gibbs DL, Brown SD, et al. Immune landsc cancer. Immun. 2018;48(4):812-e3014. [Google Scholar]
- 19.Gao J, Aksoy BA, Dogrusoz U, et al. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal. 2013;6(269):pl1-pl. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.El-Gebali S, Mistry J, Bateman A, et al. The Pfam protein families database in 2019. Nucleic Acids Res. 2018;47(D1):D427–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Feber A, Guilhamon P, Lechner M, et al. Using high-density DNA methylation arrays to profile copy number alterations. Genome Biol. 2014;15(2):R30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Saito Y, Shultz LD, Ishikawa F. Understanding normal and malignant human hematopoiesis using next-generation humanized mice. Trends Immunol. 2020;41(8):706–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Chi H, Huang J, Yan Y, et al. Unraveling the role of disulfidptosis-related LncRNAs in colon cancer: a prognostic indicator for immunotherapy response, chemotherapy sensitivity, and insights into cell death mechanisms. Front Mol Biosci. 2023;10:1254232. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Monteleone G, Pallone F, Macdonald TT. Interleukin-21: a critical regulator of the balance between effector and regulatory T-cell responses. Trends Immunol. 2008;29(6):290–4. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All datasets used in this study are publicly available from their official portals (e.g., UCSC Xena for TCGA/GTEx; cBioPortal for somatic alterations; TIMER/MCP-COUNTER for immune estimates). Accession identifiers, access dates, and processing steps are listed in Methods and Supplementary Table S1. No controlled-access individual-level data were generated in this work.