No-new-wet-lab evaluation | NMD-VCell Research Workbench

Next-stage evaluation · released contract

No-new-wet-lab evaluation pack

A computation-first package for stress-testing, falsifying, downgrading and prioritizing hypotheses with released and public data before a new experiment is started.

Open contract JSON Module registry Scorecard template README

No new wet lab does not mean no new analysis. Public datasets may be added and new computation may be executed, but every run must freeze its eligible universe, split, baseline, metric, uncertainty and claim boundary.

Decision boundary

What the pack can answer now—and what remains locked

Can answer

Computational reliability and compatibility

Release reproducibility, same-context benchmark behaviour, external directional-transfer failure, public myogenic-context compatibility, uncertainty gaps and which hypotheses should be kept, downgraded or deferred.

Cannot answer

DMD efficacy, causality, function or safety

Whether KO or overexpression rescues DMD function, whether a pair is synergistic, whether an effect replicates across DMD donors, or whether a candidate has treatment utility.

Execution matrix

Eleven modules; executed, partial and registered states remain explicit

ID	Module	Independent unit	Current state	Claim impact
NWLE-00	Release and provenance integrity	release object or rendered route	RELEASED_EXECUTED	Supports reproducibility of the released resource; it does not support a biological claim.
NWLE-01	Same-context predictive audit	held-out perturbation target within repeated balanced folds	EXECUTED_BOUNDED_STRICT_GATE_0_OF_16	Supports a small same-context RMSE improvement only; strict direct-head support is not established.
NWLE-02	External aggregate transfer diagnostic	external perturbation target	EXECUTED_NO_DIRECTIONAL_TRANSFER_SUPPORT	Falsifies a broad directional-transfer claim on the current 55-target panel; it is not DMD or muscle truth.
NWLE-03	Public myogenic and DMD context triangulation	public study, target or reference trajectory	EXECUTED_REFERENCE_ONLY	Can downgrade or contextualize hypotheses; cannot create independent DMD candidate-perturbation validation.
NWLE-04	Source-balanced DMD prior multiverse	source-aware gene-context estimate	REGISTERED_NOT_EXECUTED	Will improve disease-prior robustness, but still will not constitute perturbation efficacy.
NWLE-05	Leakage-safe G2 backtest	held-out gene family or pathway	REGISTERED_NOT_EXECUTED	Required before an unseen-family generalization claim.
NWLE-06	In-silico perturbation-mode boundary	perturbation × context × modality	PARTIAL_BOUNDED_SIMULATION_ONLY	Observed same-HepG2 CRISPRi and additive null previews are usable; KO, overexpression, fate and synergy claims remain locked.
NWLE-07	Uncertainty, OOD and abstention audit	target or context-level prediction	PARTIAL_BOOTSTRAP_AVAILABLE_OOD_ABSTENTION_NOT_TRAINED	Sampling intervals are available for selected executed analyses; trained OOD-aware abstention is not yet supported.
NWLE-08	Retrospective selection-enrichment backtest	outcome-blinded candidate selection set	REGISTERED_NOT_EXECUTED_OUTCOME_LEAKAGE_GUARD	Only valid when a pre-outcome model snapshot and eligible outcome universe can be proven.
NWLE-09	Independent computational reproduction	independent analyst execution	REGISTERED_EXTERNAL_REPRODUCTION_NOT_EXECUTED	Independent reproduction can strengthen computational reliability, but not biological validity.
NWLE-10	UniPert cross-domain representation feasibility	candidate gene, held-out target family or compound scaffold	PARTIAL_SOURCE_AND_IDENTIFIER_AUDIT_EXECUTED_MODEL_EVALUATION_PENDING	Supports operational feasibility only; no embedding benefit, cross-domain transfer or DMD therapeutic value has been established.

Already executable evidence

The first scorecard starts with results already present in the release

Same context

0 / 16 strict gates

The repeated HepG2 audit retains a small RMSE advantage but does not pass the strict direct-head family. The historical 16/16 rule remains sensitivity only.

Inspect benchmark

External transfer

No directional transfer support

The executed 55-target aggregate-delta diagnostic does not support broad directional transfer and is not muscle or DMD truth.

Inspect VCC-style diagnostic

Myogenic context

Reference-only triangulation

Public fusion, differentiation and DMD-correction layers can assess coverage and compatibility, but do not perturb the 21 candidates in an independent DMD functional system.

Inspect evidence map

Simulation

Bounded CRISPRi and additive null

Observed same-HepG2 CRISPRi and an additive double-perturbation null are available. KO, overexpression, fate and interaction truth stay locked.

Open Perturbation Lab

UniPert feasibility

21 / 21 identifier preflight

Paper, code surface and candidate identifiers are audited. Encoder runs, embeddings, G2 gains and DMD claims remain zero.

Inspect NWLE-10

Claim ladder

Computation can move C0–C2; C3–C4 require disease-background outcomes

C0Release integrity

Reproducible access to the stated objects and routes.

AVAILABLE

C1Same-context computational support

Bounded performance statements inside the frozen HepG2 evaluation context.

PARTIAL_STRICT_GATE_NOT_PASSED

C2External compatibility

Compatibility or falsification statements on explicitly named public datasets.

PARTIAL_NO_DIRECTIONAL_TRANSFER_SUPPORT

C3DMD perturbation prediction

Disease-background perturbation-response prediction.

LOCKED_DMD_PERTURBATION_TRUTH_MISSING

C4Causal function, safety or treatment utility

Replicated DMD functional, safety or therapeutic claims.

LOCKED_REQUIRES_EXPERIMENTAL_OUTCOME

Decision ledger

Every completed module ends in a bounded action

KEEP_FOR_COMPUTATIONAL_TRIAGEDOWNGRADEABSTAINESCALATE_TO_EXPERIMENT

A computational KEEP decision retains a hypothesis for triage. It does not promote its evidence level.

Hard prohibitions

Seven shortcuts that the pack will not allow

Do not represent the negative of knockdown as overexpression truth.
Do not represent an additive double-perturbation null as synergy prediction.
Do not represent transcriptomic reversal as DMD functional rescue.
Do not represent a public reference layer as independent candidate perturbation validation.
Do not treat cells, images or spots within one biological sample as independent biological replicates.
Do not promote computational compatibility to clinical, safety or treatment utility.
Do not treat a protein or compound identity embedding as perturbation direction, dose, time, mechanism or DMD response.

Machine-readable contract

One frozen module registry

Each module names its unit, baselines, outputs, state, evidence anchors and allowed claim impact.

Open JSON

Run-level output

One Scorecard per execution

Freeze the data universe and split before running; retain estimates, intervals, multiplicity, failure state and artifact URI.

Download template

Boundary. This package adds no wet-lab experiment and no DMD functional outcome. Its highest valid use is auditable computational triage, external compatibility testing, falsification and experiment prioritization.