NMD-VCell Neuromuscular Virtual Cell Research Platform Module: Models · Evidence → perturbation → experiment → outcome NMD = neuromuscular disorders

Release 2026.08DMD context observedDMD candidate-conditioned prediction not yet eligible

View scientific status
Evidence freeze: 3 August 2026 Resource: v1.2.0-measured-dmd-evidence Schema: 1.1 Open release status →

← ModelRun & Audit Registry

ModelRun Release · immutable execution history

TxPert public-STRING config-gat

This page publishes the frozen task, result, baseline decision, provenance receipt and prohibited uses for one completed run—including failed and no-go results.

01Frozen dataNMDVCELL-HEPG2-CRISPRI-TXPERT-ADAPTER-V26
02Task & splitG0 · txpert-v26-frozen-seed-20260712
03Baseline comparisonPREREGISTERED_ADVANCED_COMPARATOR
04FAILADVANCEMENT_GATE_FAILED

Frozen task receipt

What was evaluated

Model
EXTERNAL-TXPERT-CONFIG-GAT-PUBLIC-STRING · source-08d82ee
Biological context
Processed HepG2 CRISPRi perturbation-response substrate
Task
G0
Data release
NMDVCELL-HEPG2-CRISPRI-TXPERT-ADAPTER-V26
Split
txpert-v26-frozen-seed-20260712
Execution status
COMPLETED

Decision receipt

FAIL

Median delta cosine 0.346196 was below the train-mean control at 0.373307.

Decision state
ADVANCEMENT GATE FAILED
Baseline result
Improved RMSE and retrieval over simple controls but missed the preregistered median-delta-cosine gate against train mean.
Promoted
No
Next action
Do not run the remaining confirmatory seeds without a new preregistration and untouched evaluation set.

Released result

Metrics retained in this ModelRun

Independent recomputation passed 130/130 checks; the advancement gate remained failed.

MetricValue
held out conditions534
global delta rmse0.115092
median delta rmse0.097833
median delta cosine0.346196
top1 retrieval fraction0.02809
top5 retrieval fraction0.09176
train mean median delta cosine0.373307

Quickstart · reproduction receipt

PARTIAL REPRODUCTION RECEIPT

This is an audit-first quickstart. It exposes the exact JSON, known prerequisites and missing receipts; it does not invent a runnable command when code, container, data or seeds are absent.

Source artifacts

Associated execution evidence

  • No public source-artifact route is registered for this run.

Runtime note: Official-width config-gat with public STRING; 20 epochs; V100 recovery evaluation made zero parameter updates

Responsible use

Prohibited claims

  • No calibrated DMD prediction.
  • No DMD disease-response prediction.
  • No therapeutic or clinical use.
  • Do not inherit performance from the model family or an external publication.

Associated resources

Follow the result into the next decision.

Claim boundary

Completed does not mean clinically or biologically qualified.

Do not run the remaining confirmatory seeds without a new preregistration and untouched evaluation set.