wiki: sync eleven pages already committed to the code clone 2026-09-01..09-08 (Study 38's rewrite among them), and the board rows for Study 47 and Run-any-study — byte-identical after the link transform, Study-38 render-checked via the GitHub Markdown API (3 tables, 15 rows, 0 pipe leaks) Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Study 37 — 37,910 validated discoveries, five molecules; zilganersen screened on the real sequence Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Zilganersen / Alexander disease — the safety question made exact: a certification seal for a drug New applied study (The-Safety-Question-Made-Exact) on the first-ever therapy for Alexander disease (a fatal childhood GFAP astrocytopathy). Its one genuinely discrete safety question — where else in the transcriptome an antisense sequence can strike — is exact Watson-Crick complementarity, and a new reproduce demo (aso-offtarget-exact-vs-float.swift) MEASURES it observer-invariant where a floating-point binding-energy screen reclassifies a borderline site on the parameter set it rounds to (ASO_OFFTARGET_EXACT_IS_OBSERVER_INVARIANT). Honest scope: hybridization-dependent off-target only, not the chemistry/aseptic-meningitis risks; the real-drug molecular court is a charter. Names the pharma-reset + licensing value (paid in lives and lost R&D). New animated SVG. Home lead now carries both faces — the CAR-T veto (flagged before Phase 1) and the Zilganersen seal (a re-derivable certification) — both belonging before a first patient is dosed, and useful for a pipeline drug to decide the path forward. Study 14 links both applied deep dives. Harness 158/0. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>