TNMplot

TNMplot provides comparative analysis of gene expression across normal, tumor, and metastatic tissues by integrating transcriptome and gene chip data for cancer biomarker and target discovery.

Key Features:

• Integrated database: Aggregates 56,938 samples from TCGA, TARGET, GTEx, and NCBI-GEO, including 3,180 gene chip-based studies (453 metastatic, 29,376 tumor, 3,691 normal), 11,010 TCGA samples (394 metastatic, 9,886 tumor, 730 normal), 1,193 TARGET samples (1 metastatic, 1,180 tumor, 12 normal), and 11,215 GTEx normal samples.
• Data types: Includes transcriptome-level and gene chip-based expression data for pan-cancer and tissue-specific analyses.
• Statistical analysis: Compares gene expression using Mann-Whitney or Kruskal-Wallis tests with False Discovery Rate adjustment via the Benjamini-Hochberg method.
• Visualization options: Produces boxplots, violin plots, sensitivity–specificity curves, and stage-specific comparisons.
• Biomarker findings: Identifies differentially expressed genes and reports fold changes, including consistently upregulated TOP2A (FC = 7.8), SPP1 (FC = 7.0), CENPA (FC = 6.03) and downregulated ADH1B (FC = 0.15).
• Validation: Differential expression results were validated using equally sized training and test sets with reported FDR below 10% in breast, colon, and lung cancer analyses.

Scientific Applications:

• Tumor biology investigation: Comparative expression analysis across normal, tumor, and metastatic tissues to study mechanisms of tumor formation and progression.
• Biomarker discovery: Identification and reporting of differentially expressed genes and fold changes to support candidate biomarker selection.
• Therapeutic target exploration: Prioritization of genes with consistent up- or downregulation across cancers for potential therapeutic targeting.

Methodology:

Integrates transcriptome and gene chip data from TCGA, TARGET, GTEx, and NCBI-GEO; analyzes expression differences with Mann-Whitney or Kruskal-Wallis tests, applies Benjamini-Hochberg FDR correction, generates boxplots/violin plots and sensitivity–specificity and stage-specific comparisons, reports fold changes, and validates results using equally sized training and test sets.