Cancer Embeddings Explorer

🚀 Generate 3D Projections 🗑️ Clear Cache

🔍 Number of Neighbors (k): 📏 Distance Metric: Euclidean (L2) - Standard geometric Cosine - Best for embeddings Manhattan (L1) - Robust to outliers Correlation - Linear relationships Chebyshev (L∞) - Max difference Canberra - Weighted for sparse data 🔄 Recalculate Overlap Click two points to compare their k nearest neighbors

🧭 Visualization Technique: PCA (fast, linear) t-SNE (non-linear) UMAP (fast non-linear) 🖼️ Apply Visualization Generate visualization from original embeddings