Overview of dbGIST as a GIST-focused multi-omics resource.

(A) Sample processing of in-house GIST proteomics and single-cell transcriptomics. (B) Public and in-house GIST multi-omics data integration and main dbGIST analytical modules.

Overview of the dbGIST interface and analytical framework.

(A) Homepage and core functional modules of dbGIST. (a) Top navigation bar providing access to Home, Dataset, and Guide pages for platform navigation and user support. (b) LLM-assisted GIST knowledge chatbot, enabling natural-language queries for biological questions and research guidance. (c) GIST gene screening module for candidate gene querying and target exploration. (d) AI interpretation switch that enables automatic explanation of bioinformatics results on the current page. (e) Entry points for multi-omics data analysis, including genomics, transcriptomics, proteomics, phosphoproteomics, single-cell transcriptomics, and non-coding RNA modules. (B) Representative omics analysis page (transcriptomics module). (f) Left-side navigation panel listing available analytical modules, including clinical association analysis, immune infiltration, gene correlation, and functional enrichment. (g) Parameter configuration interface allowing dataset selection, grouping strategy, and statistical threshold customization for dynamic analysis. (h) AI-assisted interpretation panel that provides automated explanation of analytical results and supports interactive user feedback.

Clinical associations of MCM7 based on integrated omics data.

(A) Summary of clinical annotations available for each omics layer. (B) Kaplan-Meier curves showing overall survival (OS) and progression-free survival (PFS) in the proteomics cohort stratified by MCM7 expression. (C) MCM7 expression across NIH risk groups in four independent transcriptomic cohorts. (D) MCM7 expression across NIH risk groups in two proteomic cohorts. (E) MCM7 expression across primary, metastatic, recurrent/metastatic, and non-metastatic comparison groups in four transcriptomic cohorts. (F) Association between MCM7 expression and mitotic count in two proteomic cohorts. (G) MCM7 expression across histologic categories in a transcriptomic cohort. (H) MCM7 expression by imatinib response and ROC analysis for distinguishing imatinib-resistant from imatinib-naive cases in GSE132542. (I) MCM7 expression by imatinib response and ROC analysis in tumor-associated macrophages from GSE51697. Centre line: median; box bounds: 25th and 75th percentiles.

Single-cell distribution and tumor-cell associations of MCM7 in GIST.

(A, D, G) UMAP visualization of annotated cell types in the ZZU-GIST, GSE162115, and GSE254762 cohorts, respectively. (B, E, H) UMAP feature plots showing MCM7 expression in each cohort. (C, F, I) Violin plots showing MCM7 expression across annotated cell types. (J-M) MCM7 expression in tumor cells stratified by NIH risk in GSE162115 and by recurrence, mitotic count, and imatinib resistance in GSE254762. Centre lines indicate medians where shown.

Biological pathway enrichment results of MCM7 in GIST.

(A) Schematic workflow of the integrated bioinformatic analysis pipeline. (B) Gene Set Enrichment Analysis (GSEA) plot of the Hallmark gene set. (C) Bar graph from over-representation analysis (ORA) of the top MCM7-correlated genes. (D) Protein-Protein interaction network of the top MCM7-correlated genes. (E) Bar graph from over-representation analysis (ORA) of the top MCM7-correlated proteins. (F) Protein-Protein interaction network of the top MCM7-correlated proteins.

Immune infiltration and drug-sensitivity analyses of MCM7.

(A) Flowchart of the integrated bioinformatic pipeline used for tumor immunophenotyping and in silico drug prediction across multiple GIST cohorts. (B) Heatmap showing Pearson correlation coefficients between MCM7 expression and activity of the seven-step cancer-immunity cycle. Red indicates positive correlation and blue indicates negative correlation. (C) Circos plot showing correlations between MCM7 expression and immune cell subtype abundance in the tumor microenvironment, as estimated by multiple deconvolution algorithms. (D) Scatter plots showing positive correlations between MCM7 and MEP (megakaryocyte-erythroid progenitor) cell abundance. (E) Heatmap of correlations between MCM7 expression and predicted drug sensitivity or resistance across analyzed cohorts. (F) Scatter plots showing significant negative correlations between MCM7 expression and predicted response to C6-ceramide.

Experimental validation of MCM7 in in vitro GIST models.

(A) Scatter plots showing correlations between MCM7 and MKI67 expression across independent GIST cohorts. (B, D) Western blot validation of siRNA-mediated MCM7 knockdown in GIST-T1 and GIST-882 cells. (C, E) qRT-PCR analysis of MCM7 mRNA expression after siRNA-mediated knockdown in GIST-T1 and GIST-882 cells. (F, I) CCK-8 proliferation assays after MCM7 knockdown in GIST-T1 and GIST-882 cells. (G, J) Quantification of wound-healing assays. (H, K) Representative wound-healing images at 0, 24, and 48 h. Statistical significance: *P < 0.05; **P < 0.01; ***P < 0.001.