Home› Explore› Validation

Biological Validation

Cross-referencing ML-identified candidates against external databases, pathway enrichment, survival analysis, and multi-omics convergence evidence.

⚠️
Validation Status

Biological validation is ongoing. The COSMIC/OncoKB cross-reference provides the strongest evidence, using established cancer gene databases. Other modules show preliminary results that require further data integration and formal statistical testing before definitive conclusions.

—
Total Candidates
—
Known Oncogenes
—
Cox PH Significant
—
Enriched Pathways
Modules: 📊 Survival 🧬 GSEA 📚 COSMIC/OncoKB
📚

COSMIC / OncoKB Cross-Reference

Loading

Candidates cross-referenced against the COSMIC Cancer Gene Census (~110 Tier 1 driver genes) to separate known oncogenes from novel discoveries.

🧬

GSEA Pathway Enrichment

Loading

Enrichr analysis of candidate genes against KEGG and Reactome pathway databases to identify biologically coherent functional themes.

📊

Survival Analysis

Loading

Multivariate Cox proportional hazards regression testing whether candidate gene expression is independently associated with overall survival after adjusting for age, pathologic stage, and PAM50 molecular subtype. BH-corrected p-values control false discovery rate across all tested genes.

ℹ️
Validation Framework

Each module provides independent biological evidence supporting (or challenging) ML-identified candidates. Genes validated across multiple modules carry the strongest biological support.