MDR: Metabolic Divergence Rate — A New Primitive for Early Cancer Detection via DIP Framework and Genesis Engine (v2 — English)

We present MDR (Metabolic Divergence Rate), a new primitive measure for early cancer detection. MDR quantifies the normalised Euclidean distance between a cell's gene expression vector and the reference profile of its healthy cell type. Calculated directly from raw TCGA Pan-Cancer Atlas data (981 samples, 6 Hallmarks of Cancer genes), MDR separates all cancerous subtypes from normal tissue with 96.6% accuracy and AUC=0.71 using threshold MDR_critical=0.414. MDR was identified by the Genesis Engine — a product of the DIP Framework with Constitutional Compiler and KB — in hybrid mode with LLM (Claude, Anthropic). Verifiable data hash: 871963c7c7e7fc5c. Safe Creative n.º 2603114847646. Portuguese version (v1): https://doi.org/10.5281/zenodo.19034027