Cancer Research: Cross-Literature Analysis and Ferroptosis Simulation

Open-source cancer research combining cross-literature analysis of 4,830 PubMed-indexed research articles with Monte Carlo biochemical simulations of ferroptosis sensitivity across cell phenotypes, drug combinations, and tumor microenvironment features. This archive contains the complete analysis pipeline (Python), simulation engine (Rust, ferroptosis-core library with 10 binaries and 31 unit tests), a 112-page book-format manuscript with 20 figures, and corpus metadata for reconstruction from PubMed. The simulations explore how physical ROS-generating modalities (sonodynamic therapy, photodynamic therapy) compare to pharmacologic ferroptosis inducers (RSL3) under realistic tumor conditions including hypoxia, stromal shielding, acidic pH, and immune coupling. Key findings include: (1) RSL3 + FSP1 inhibitor produces 1.99x Bliss synergy through dual-pathway depletion, (2) the tumor microenvironment selectively penalizes pharmacologic ferroptosis inducers through three resistance mechanisms while leaving physical modalities unaffected, and (3) tissue-specific drug penetration creates a substantial in-vitro-to-in-vivo translation gap. All findings are model-predicted with documented assumptions, sensitivity analyses, and proposed experimental validations. Full-text corpus articles are excluded from this deposit for licensing reasons (61 of 4,830 have unconfirmed open-access status). The corpus metadata index (INDEX.jsonl) provides PMIDs, mechanisms, cancer types, and evidence tiers for reconstruction via PubMed. Content classes: MIT-licensed code, author-owned manuscript, non-copyrightable abstracts and metadata. See PROVENANCE.yaml for full asset licensing details.