v0.1.0 — evidence-audit cycle complete

v0 cycle complete. Gain is frozen here as a public-data evidence audit + substrate-gap diagnosis for the canonical lung developmental regulators (NKX2-1, SOX2, SOX9).

Read first:

• notes/project_summary.md — full narrative
• notes/key_findings.md — concise findings
• notes/limitations_and_next_methods.md — method-level futures
• notes/next_project_decision.md — decision memo for the successor project (CELLxGENE Census expression route picked)

What this release establishes:

• Substrate-gap diagnosis: no primary human lung tissue ChIP-seq for NKX2-1 / SOX2 / SOX9 / CTNNB1 / SMAD1 / GLI2 in ChIP-Atlas hg38 (verified via NCBI GEO + SRA E-utilities external search).
• 5 NKX2-1 chromatin-supported standing distal candidate loci (FOXA2 chr20:22,716,796 strongest; FOXA2 ×2, SCGB1A1 ×2, ABCA3 ×1).
• 18 stdlib-only audit scripts implementing a 5-class evidence model + peak intersection + sensitivity sweep + distal-candidate tiering + histone-support filter + motif + accessibility scan + quantitative refinement.
• Negative finding: bulk fetal lung motif + accessibility at v0 settings cannot discriminate canonical SOX2 targets from controls.

Standing biological output: the 5 NKX2-1 candidate loci. Candidates, not validated regulatory elements. Become testable when primary lung developmental ChIP-seq is published or when Hi-C / promoter-capture data is brought into scope.