NMD-VCellNeuromuscular Virtual Cell Research PlatformResearch Question · Evidence → perturbation → experiment → outcomeNMD = neuromuscular disorders
For Reviewers Evidence Status Claims

Evidence Release 2026.08DMD context observedDMD candidate-conditioned prediction locked

View scientific status
Evidence freeze: 3 August 2026 P59 hold · scorer failure retained · no target score NAR working package · 6 main + 31 supplementary figures Resource: v1.2.0-measured-dmd-evidence Schema: 1.1 Open platform state → Open Trust Center → Open evidence dashboard → Open claim registry → Open release status →

Virtual Cell Query Composer 2.0 · Research OS V201 · production content

Turn one biological question into a bounded research answer.

Compose disease, cell state, perturbation, target, time and endpoint once. The system returns registered evidence, prediction eligibility, domain and OOD state, uncertainty, missing truth, the decisive next experiment and the highest permitted claim.

Structured question

Define the six dimensions that determine the answer boundary.

Open answer schema →
Draft question identityQ2:DMD:PROLIFERATING_MYOBLAST:CRISPRI:ZNF133:48H:FUSION_AND_MOLECULAR_RESPONSE:SINGLE

Does CRISPRi of ZNF133 alter fusion and molecular response after 48 h in proliferating myoblasts from DMD?

Inspect Virtual CellDesign study

Fail-closed rule: only an exact registered question returns its Research Answer Object. Changing any dimension to an unregistered combination produces ABSTAIN_NO_REGISTERED_RESEARCH_ANSWER; the system does not borrow evidence from a nearby query.

Research Answer Object · checksummed canonical example

ABSTAIN — actionable question, prediction not permitted.

The question is scientifically actionable but not eligible for a DMD perturbation prediction. The platform can return bounded evidence and a truth-generating experiment design.

ANSWER:Q2:DMD:PROLIFERATING_MYOBLAST:CRISPRI:ZNF133:48H:FUSION_AND_MOLECULAR_RESPONSE:V1
Evidence

Only versioned objects enter this answer.

  • TARGET CONTEXTGENE:ZNF133:CURRENTOBSERVED SAME CONTEXT PERTURBATION OUTSIDE DMD OUTCOME
    Inspect →
  • GOVERNS ACTIONREGISTRY:CANDIDATE_DECISIONS:CURRENTEXPERIMENT ELIGIBLE NOT THERAPEUTIC TARGET
    Inspect →
  • LIMITSCLAIM:DMD:CANDIDATE_OUTCOME_BOUNDARY:V1MISSING 0 OF 21
    Inspect →
Domain match

Six dimensions remain separate.

  • geneMATCHED GOVERNED CANDIDATECandidate registry and Gene Evidence object
  • cell_statePARTIAL CONTEXT ONLYHuman myogenic context exists; matched perturbation truth is absent
  • diseaseCONTEXT OBSERVED OUTCOME MISSINGDMD context exists; candidate-conditioned outcome is 0/21
  • perturbationSOURCE CONTEXT ONLYPerturbation evidence does not close muscle/DMD transfer
  • timeNOT MATCHEDNo registered DMD candidate-conditioned 48 h outcome
  • endpointNOT MATCHEDNo matched DMD fusion plus transcriptomic outcome
Model

NO ELIGIBLE DMD MODELRUN

Reference runs answer same-context or transfer-diagnostic tasks. They are not calibrated for this DMD question and are not executed by this answer.

Reference runs: MRUN-RIDGE-SAFE-2.3-G0-REPEATED-FOLD · MRUN-TRANSFER-DIAGNOSTIC-1.0-G1
Baselines

Required before comparison

  • zero changeREQUIRED NOT RUN FOR QUERY
  • train meanREQUIRED NOT RUN FOR QUERY
  • ridgeREFERENCE ONLY NOT DMD ELIGIBLE
  • nearest neighbourREGISTERED FOR RELEVANT OOD TASKS NOT RUN FOR QUERY
OOD status

No single “generalizes well” label.

  • geneBOUNDARY UNTESTED
  • cell_stateOUTSIDE VALIDATED PREDICTION DOMAIN
  • diseaseOUTSIDE VALIDATED PREDICTION DOMAIN
  • donorNOT TESTED
  • laboratoryNOT TESTED
  • timeNOT TESTED
Uncertainty

Qualitative and unresolved—not fabricated as a probability.

EpistemicHIGH

No calibrated disease-matched predictive model.

DataHIGH

No candidate-conditioned DMD perturbation outcome.

DomainHIGH

Muscle, disease, donor, laboratory and time transfer remain unvalidated.

Missing truth

The exact layers preventing prediction.

  • Independent human myogenic ZNF133 perturbation response
  • Matched DMD/control or isogenic-corrected functional outcome
  • Target engagement and viability under the selected perturbation
  • Donor-resolved replication
  • Frozen 48 h molecular endpoint and functional fusion endpoint
Best next experiment

Declared human myogenic state with healthy/DMD or isogenic-corrected background

Perturb
CRISPRi with at least two independent reagents where feasible
Primary endpoint
One predeclared functional endpoint, with fusion as the query-selected endpoint
Secondary endpoints
target engagement · viability/toxicity · directional transcriptomic response
Planning n
3–6 independent biological replicates per condition and reagent; final powered n remains open
Decision rule
Advance only if engagement and QC pass, reagents agree, and the phenotype replicates beyond one biological unit. Retain null, discordant and QC-failed results.
Open machine-checkable Experiment OS →Open source Study Card →
Expected information gain

HIGH QUALITATIVE NOT QUANTIFIED

The experiment directly addresses the muscle-context transfer gap and can move the object to a null, supportive, discordant or QC-failure outcome state.

A qualitative priority is not a probability, effect size, therapeutic score or formal value-of-information calculation.
Claim ceiling

EVIDENCE AND EXPERIMENT DESIGN ONLY

ZNF133 is eligible for a bounded human-myogenic feasibility experiment. The current evidence does not support a DMD perturbation-response, efficacy, therapeutic-ranking or clinical claim.

Inspect governing claim →

Export governed object

JSON, YAML and DOI-ready citation metadata share one answer identity.

The object contains a pending DOI state; exporting it does not mint a DOI, register a Study or create a prediction.

Download JSONDownload YAMLJSON SchemaAnswer registry

Connected objects

Carry the answer into the evidence-to-outcome loop.

ClaimsWhich statement governs use?EligibilityWhy must prediction abstain?ExperimentWhich gates block execution?OutcomeWhere will measured truth return?