NAR publication addendum · curated v08.7
Two pairs define a bounded ADAM10 pilot—not general DMD validation.
A deterministic 480-scenario design-sensitivity grid quantifies what the prospective paired design can and cannot conclude before any candidate outcome is accessed.
Evidence-to-experiment bridge · v08.8 → v09.1
From design sensitivity to accountable execution.
The sensitivity analysis now feeds a linked lock, study-amendment, randomization and outcome-ingestion system. One shared evidence base routes 39 governed decisions into 56 independent assignments, 15 prospective checks and three checksum-bound reviewer packets—turning a design result into a portable, auditable handoff.
Experimental unit
Independent model pair for generalization; differentiations are nested support. Wells, guides and cells never inflate n.
Sample-size gate
Two pairs remain Stage A calibration. Stage B pair count requires endpoint-specific variance and effect/margin approval.
Outcome return
Append-only contract retains positive, null, toxic, discordant, failed-QC and inconclusive results.
Supplementary Figure S32
Operating characteristics expose the cost of few independent backgrounds.
Three differentiations and three guides are within-pair support; they are not additional disease backgrounds. Reference assumptions: heterogeneity SD 0.25, guide retention 0.95, equivalence margin 0.50 SD and variance multiplier 4.
One result, four synchronized surfaces
Website, manuscript, figure and attachments resolve to the same evidence object.
| Surface | Primary object | Boundary |
|---|---|---|
| Manuscript | Successor manuscript manifest | Additive over frozen v08.6 |
| Supplement | Supplement/addendum manifest | Figure S32 and Tables S65–S66 |
| Figure | Figure/source manifest | Same frozen probabilities |
| Attachments | Tables S65–S66 manifest | Checksummed public packet |
Version separation
This addendum does not silently relabel another release.
Scientific resourcev1.2.0-measured-dmd-evidence
Unchanged.
Publication addendumcurated v08.7
New design-sensitivity layer over frozen v08.6.
Other manuscript packetScience Advances v09.80
Retained separately and not replaced.
Fail-closed claim boundary
Simulation precision is not biological efficacy.
There are 0 observed DMD candidate response outcomes. Thirty-nine endpoint-, model-pair-, QC- and authority-specific fields remain unresolved. Non-significance is not equivalence; the current object authorizes neither wet-lab execution, public deposit nor submission.
Inspect blocked lock template manifest