- selecting validation questions
- choosing an assay or analysis
- balancing a pilot perturbation panel
- creating a Study Card with explicit controls
Candidate evidence workbench · DMD hypotheses
Compare evidence vectors, then decide what to test next.
Use candidates to decide what to test next. A candidate is useful only when it resolves a named uncertainty. The default matrix keeps direct observation, muscle context, source agreement, model gate and missing replication separate—without a target score or winner.
- a proven therapeutic target
- clinical benefit or safety
- a calibrated DMD response prediction
- superiority over another candidate
Shared reading grammar
Evidence metadata fields—never one confidence score.
Every state uses text plus a shape or border. Missing evidence and a equivalent-null result are different states.
Origin● Observed△ Computational□ Missing
Context relationD Target contextM Same tissueL Locus correction↗ External cell
Unit structuren Donor / linec Culturep Pathway? Not estimable
Lifecycle◇ Draft◆ Frozen▣ Registered● Released
Outcome / absence↑ Supportive0 Measured null? Inconclusive□ Not measured
Candidate evidence vector matrix
All 21 hypotheses, compared without ranking.
Each cell is one frozen evidence dimension—not a component of a hidden score. Text and symbols distinguish available, contextual, conflicted, below-threshold and missing evidence.
7Computational replication7Risk/context review4Verify expression2Muscle-context assay1Provisional hold · external report not yet audited
●available/supporting
◐partial/context
↕conflicted
!below expression threshold
■missing/locked
| Gene | Current action | L2 obs. | Muscle expr. | Source agree | Counteralignment range | Strict gate | L3b rep. |
|---|---|---|---|---|---|---|---|
| CPEB1 | Verify expression | ●yes | !low | ◐0.67 | ●0.01…0.12 | ■not passed | ■pending |
| DNAAF3 | Verify expression | ●yes | !low | ●1 | ↕-0.04…0.06 | ■not passed | ■pending |
| EPS8L1 | Verify expression | ●yes | !low | ●1 | ↕-0…0.11 | ■not passed | ■pending |
| ZFP69B | Verify expression | ●yes | !low | ◐0.67 | ↕-0.05…0.06 | ■not passed | ■pending |
| INTS13 | Risk/context review | ●yes | ●TPM>1 | ●1 | ↕-0.03…0.06 | ■not passed | ■pending |
| NAGLU | Risk/context review | ●yes | ●TPM>1 | ●0.75 | ↕-0.04…0.05 | ■not passed | ■pending |
| RAC3 | Risk/context review | ●yes | !low | ●0.75 | ●0.01…0.09 | ■not passed | ■pending |
| RNASEH2C | Risk/context review | ●yes | ●TPM>1 | ●1 | ↕-0.05…0.04 | ■not passed | ■pending |
| RNF8 | Risk/context review | ●yes | ●TPM>1 | ↕0.5 | ↕-0.03…0.04 | ■not passed | ■pending |
| WDR4 | Risk/context review | ●yes | ●TPM>1 | ●1 | ↕-0.02…0.06 | ■not passed | ■pending |
| ZNF236 | Risk/context review | ●yes | ●TPM>1 | ↕0.5 | ↕-0.07…0.09 | ■not passed | ■pending |
| MON1A | Muscle-context assay | ●yes | ●TPM>1 | ↕0.5 | ↕-0.07…0.05 | ■not passed | ■pending |
| ZNF133 | Muscle-context assay | ●yes | ●TPM>1 | ●0.75 | ↕-0.04…0.07 | ■not passed | ■pending |
| ADAM10 | Computational replication | ●yes | ●TPM>1 | ↕0.5 | ↕-0.01…0.08 | ■not passed | ■pending |
| CALR | Computational replication | ●yes | ●TPM>1 | ●0.75 | ↕-0.01…0.07 | ■not passed | ■pending |
| DDX19B | Computational replication | ●yes | ●TPM>1 | ↕0.5 | ↕-0.05…0.02 | ■not passed | ■pending |
| DNM1 | Computational replication | ●yes | ●TPM>1 | ↕0.5 | ↕-0.07…0.02 | ■not passed | ■pending |
| EHMT2 | Computational replication | ●yes | ●TPM>1 | ●0.75 | ↕-0.03…0.03 | ■not passed | ■pending |
| LMO2 | Computational replication | ●yes | ●TPM>1 | ●0.75 | ↕-0.08…0.02 | ■not passed | ■pending |
| MPHOSPH6 | Computational replication | ●yes | ●TPM>1 | ●1 | ↕-0.04…0.08 | ■not passed | ■pending |
| GFOD2 | Provisional hold · external report not yet audited | ●yes | ●TPM>1 | ↕0.5 | ↕-0.05…0.08 | ■not passed | ■pending |
Candidate Decision Layer · 21 governed hypotheses
What is each candidate actually for?
Each card binds one gene to one research question, one decision-blocking gap and one result rule. The cards are grouped by current action and alphabetized within each group; they are not ranked.
Candidate a governed hypothesisUse the uncertainty it can resolveNext action the smallest valid handoffResult rule what changes our mind
2 candidates
Test in human myogenic cells
Muscle-context assay
MON1A
GAP-02Does the MON1A response observed outside muscle transfer to a controlled human myogenic system?
- Why this candidate is useful
- Use this candidate to test context transfer and decide whether a matched DMD/control study is justified.
- Decision available now
- Eligible for a preregistered human-myogenic feasibility experiment; not established as a therapeutic target.
- What blocks the decision
- Independent muscle-context perturbation is absent
- Minimum next action
- Perturb MON1A with at least two independent reagents in a declared human myogenic state and freeze one functional primary endpoint.
How would the result change the decision?
SupportiveIf engagement and QC pass, both reagents agree, and the primary phenotype replicates beyond one biological unit, advance to matched DMD/control validation.
NullIf engagement and QC pass but the primary phenotype is null, record a null muscle-context outcome and stop treating the non-muscle response as transferable.
Inconclusive / QC failureIf reagents, donors or batches disagree, classify the result as inconclusive and resolve heterogeneity before any escalation.
Muscle-context assay
ZNF133
GAP-02Does the ZNF133 response observed outside muscle transfer to a controlled human myogenic system?
- Why this candidate is useful
- Use this candidate to test context transfer and decide whether a matched DMD/control study is justified.
- Decision available now
- Eligible for a preregistered human-myogenic feasibility experiment; not established as a therapeutic target.
- What blocks the decision
- Independent muscle-context perturbation is absent
- Minimum next action
- Perturb ZNF133 with at least two independent reagents in a declared human myogenic state and freeze one functional primary endpoint.
How would the result change the decision?
SupportiveIf engagement and QC pass, both reagents agree, and the primary phenotype replicates beyond one biological unit, advance to matched DMD/control validation.
NullIf engagement and QC pass but the primary phenotype is null, record a null muscle-context outcome and stop treating the non-muscle response as transferable.
Inconclusive / QC failureIf reagents, donors or batches disagree, classify the result as inconclusive and resolve heterogeneity before any escalation.
4 candidates
Confirm the gene is detectable
Verify expression
CPEB1
GAP-06Is CPEB1 detectably expressed at RNA and protein level in the intended human muscle model?
- Why this candidate is useful
- Use this candidate to decide whether a perturbation experiment would be technically interpretable.
- Decision available now
- Expression feasibility must be resolved before perturbation resources are committed.
- What blocks the decision
- Expression feasibility is not verified in the intended model
- Minimum next action
- Measure CPEB1 RNA and, where a validated reagent exists, protein in the intended myoblast or myotube state.
How would the result change the decision?
SupportiveIf the predeclared detectability threshold is met in independent biological replicates, move to a bounded perturbation design.
NullIf the target stays below the frozen detectability threshold, stop or redesign the assay; this is a feasibility result, not evidence of no biological role.
Inconclusive / QC failureIf RNA and protein or biological replicates disagree, retain the expression gate and resolve assay specificity or state dependence.
Verify expression
DNAAF3
GAP-06Is DNAAF3 detectably expressed at RNA and protein level in the intended human muscle model?
- Why this candidate is useful
- Use this candidate to decide whether a perturbation experiment would be technically interpretable.
- Decision available now
- Expression feasibility must be resolved before perturbation resources are committed.
- What blocks the decision
- Expression feasibility is not verified in the intended model
- Minimum next action
- Measure DNAAF3 RNA and, where a validated reagent exists, protein in the intended myoblast or myotube state.
How would the result change the decision?
SupportiveIf the predeclared detectability threshold is met in independent biological replicates, move to a bounded perturbation design.
NullIf the target stays below the frozen detectability threshold, stop or redesign the assay; this is a feasibility result, not evidence of no biological role.
Inconclusive / QC failureIf RNA and protein or biological replicates disagree, retain the expression gate and resolve assay specificity or state dependence.
Verify expression
EPS8L1
GAP-06Is EPS8L1 detectably expressed at RNA and protein level in the intended human muscle model?
- Why this candidate is useful
- Use this candidate to decide whether a perturbation experiment would be technically interpretable.
- Decision available now
- Expression feasibility must be resolved before perturbation resources are committed.
- What blocks the decision
- Expression feasibility is not verified in the intended model
- Minimum next action
- Measure EPS8L1 RNA and, where a validated reagent exists, protein in the intended myoblast or myotube state.
How would the result change the decision?
SupportiveIf the predeclared detectability threshold is met in independent biological replicates, move to a bounded perturbation design.
NullIf the target stays below the frozen detectability threshold, stop or redesign the assay; this is a feasibility result, not evidence of no biological role.
Inconclusive / QC failureIf RNA and protein or biological replicates disagree, retain the expression gate and resolve assay specificity or state dependence.
Verify expression
ZFP69B
GAP-06Is ZFP69B detectably expressed at RNA and protein level in the intended human muscle model?
- Why this candidate is useful
- Use this candidate to decide whether a perturbation experiment would be technically interpretable.
- Decision available now
- Expression feasibility must be resolved before perturbation resources are committed.
- What blocks the decision
- Expression feasibility is not verified in the intended model
- Minimum next action
- Measure ZFP69B RNA and, where a validated reagent exists, protein in the intended myoblast or myotube state.
How would the result change the decision?
SupportiveIf the predeclared detectability threshold is met in independent biological replicates, move to a bounded perturbation design.
NullIf the target stays below the frozen detectability threshold, stop or redesign the assay; this is a feasibility result, not evidence of no biological role.
Inconclusive / QC failureIf RNA and protein or biological replicates disagree, retain the expression gate and resolve assay specificity or state dependence.
7 candidates
Review safety and biological context
Risk/context review
INTS13
GAP-08Does INTS13 have a dependency, toxicity or context-specific safety profile that blocks escalation?
- Why this candidate is useful
- Use this candidate to decide whether a usable perturbation window exists before testing disease relevance.
- Decision available now
- Safety and dependency context require review before a larger perturbation experiment.
- What blocks the decision
- Safety and dependency context blocks escalation
- Minimum next action
- Run a context-matched dose/time and viability review for INTS13 with engagement, toxicity and stop rules declared in advance.
How would the result change the decision?
SupportiveIf a reproducible perturbation window meets engagement and viability gates across biological units, advance to a functional muscle assay.
NullIf no acceptable window remains after engagement and QC, stop escalation and retain the negative feasibility evidence.
Inconclusive / QC failureIf the window is donor-, dose- or reagent-specific, classify the result as context dependent and redesign before escalation.
Risk/context review
NAGLU
GAP-08Does NAGLU have a dependency, toxicity or context-specific safety profile that blocks escalation?
- Why this candidate is useful
- Use this candidate to decide whether a usable perturbation window exists before testing disease relevance.
- Decision available now
- Safety and dependency context require review before a larger perturbation experiment.
- What blocks the decision
- Safety and dependency context blocks escalation
- Minimum next action
- Run a context-matched dose/time and viability review for NAGLU with engagement, toxicity and stop rules declared in advance.
How would the result change the decision?
SupportiveIf a reproducible perturbation window meets engagement and viability gates across biological units, advance to a functional muscle assay.
NullIf no acceptable window remains after engagement and QC, stop escalation and retain the negative feasibility evidence.
Inconclusive / QC failureIf the window is donor-, dose- or reagent-specific, classify the result as context dependent and redesign before escalation.
Risk/context review
RAC3
GAP-08Does RAC3 have a dependency, toxicity or context-specific safety profile that blocks escalation?
- Why this candidate is useful
- Use this candidate to decide whether a usable perturbation window exists before testing disease relevance.
- Decision available now
- Safety and dependency context require review before a larger perturbation experiment.
- What blocks the decision
- Safety and dependency context blocks escalation
- Minimum next action
- Run a context-matched dose/time and viability review for RAC3 with engagement, toxicity and stop rules declared in advance.
How would the result change the decision?
SupportiveIf a reproducible perturbation window meets engagement and viability gates across biological units, advance to a functional muscle assay.
NullIf no acceptable window remains after engagement and QC, stop escalation and retain the negative feasibility evidence.
Inconclusive / QC failureIf the window is donor-, dose- or reagent-specific, classify the result as context dependent and redesign before escalation.
Risk/context review
RNASEH2C
GAP-08Does RNASEH2C have a dependency, toxicity or context-specific safety profile that blocks escalation?
- Why this candidate is useful
- Use this candidate to decide whether a usable perturbation window exists before testing disease relevance.
- Decision available now
- Safety and dependency context require review before a larger perturbation experiment.
- What blocks the decision
- Safety and dependency context blocks escalation
- Minimum next action
- Run a context-matched dose/time and viability review for RNASEH2C with engagement, toxicity and stop rules declared in advance.
How would the result change the decision?
SupportiveIf a reproducible perturbation window meets engagement and viability gates across biological units, advance to a functional muscle assay.
NullIf no acceptable window remains after engagement and QC, stop escalation and retain the negative feasibility evidence.
Inconclusive / QC failureIf the window is donor-, dose- or reagent-specific, classify the result as context dependent and redesign before escalation.
Risk/context review
RNF8
GAP-08Does RNF8 have a dependency, toxicity or context-specific safety profile that blocks escalation?
- Why this candidate is useful
- Use this candidate to decide whether a usable perturbation window exists before testing disease relevance.
- Decision available now
- Safety and dependency context require review before a larger perturbation experiment.
- What blocks the decision
- Safety and dependency context blocks escalation
- Minimum next action
- Run a context-matched dose/time and viability review for RNF8 with engagement, toxicity and stop rules declared in advance.
How would the result change the decision?
SupportiveIf a reproducible perturbation window meets engagement and viability gates across biological units, advance to a functional muscle assay.
NullIf no acceptable window remains after engagement and QC, stop escalation and retain the negative feasibility evidence.
Inconclusive / QC failureIf the window is donor-, dose- or reagent-specific, classify the result as context dependent and redesign before escalation.
Risk/context review
WDR4
GAP-08Does WDR4 have a dependency, toxicity or context-specific safety profile that blocks escalation?
- Why this candidate is useful
- Use this candidate to decide whether a usable perturbation window exists before testing disease relevance.
- Decision available now
- Safety and dependency context require review before a larger perturbation experiment.
- What blocks the decision
- Safety and dependency context blocks escalation
- Minimum next action
- Run a context-matched dose/time and viability review for WDR4 with engagement, toxicity and stop rules declared in advance.
How would the result change the decision?
SupportiveIf a reproducible perturbation window meets engagement and viability gates across biological units, advance to a functional muscle assay.
NullIf no acceptable window remains after engagement and QC, stop escalation and retain the negative feasibility evidence.
Inconclusive / QC failureIf the window is donor-, dose- or reagent-specific, classify the result as context dependent and redesign before escalation.
Risk/context review
ZNF236
GAP-08Does ZNF236 have a dependency, toxicity or context-specific safety profile that blocks escalation?
- Why this candidate is useful
- Use this candidate to decide whether a usable perturbation window exists before testing disease relevance.
- Decision available now
- Safety and dependency context require review before a larger perturbation experiment.
- What blocks the decision
- Safety and dependency context blocks escalation
- Minimum next action
- Run a context-matched dose/time and viability review for ZNF236 with engagement, toxicity and stop rules declared in advance.
How would the result change the decision?
SupportiveIf a reproducible perturbation window meets engagement and viability gates across biological units, advance to a functional muscle assay.
NullIf no acceptable window remains after engagement and QC, stop escalation and retain the negative feasibility evidence.
Inconclusive / QC failureIf the window is donor-, dose- or reagent-specific, classify the result as context dependent and redesign before escalation.
7 candidates
Reproduce the result independently
Computational replication
ADAM10
GAP-07Does the ADAM10 evidence pattern survive an independent dataset and frozen analysis workflow?
- Why this candidate is useful
- Use this candidate to decide whether the current computational signal is reproducible enough to justify experimental escalation.
- Decision available now
- Experimental escalation remains provisional until the signal is independently reproduced.
- What blocks the decision
- Independent computational replication is absent
- Minimum next action
- Repeat the ADAM10 analysis in an independent dataset with frozen mapping, contrasts, baselines, null controls and uncertainty reporting.
How would the result change the decision?
SupportiveIf direction and uncertainty reproduce under the frozen plan, retain the candidate and move to its next context-specific gate.
NullIf the signal does not reproduce despite adequate coverage and QC, retain the negative result and do not escalate from the original analysis alone.
Inconclusive / QC failureIf coverage, mapping or source design prevents a fair comparison, mark the replication non-evaluable rather than positive or negative.
Computational replication
CALR
GAP-07Does the CALR evidence pattern survive an independent dataset and frozen analysis workflow?
- Why this candidate is useful
- Use this candidate to decide whether the current computational signal is reproducible enough to justify experimental escalation.
- Decision available now
- Experimental escalation remains provisional until the signal is independently reproduced.
- What blocks the decision
- Independent computational replication is absent
- Minimum next action
- Repeat the CALR analysis in an independent dataset with frozen mapping, contrasts, baselines, null controls and uncertainty reporting.
How would the result change the decision?
SupportiveIf direction and uncertainty reproduce under the frozen plan, retain the candidate and move to its next context-specific gate.
NullIf the signal does not reproduce despite adequate coverage and QC, retain the negative result and do not escalate from the original analysis alone.
Inconclusive / QC failureIf coverage, mapping or source design prevents a fair comparison, mark the replication non-evaluable rather than positive or negative.
Computational replication
DDX19B
GAP-07Does the DDX19B evidence pattern survive an independent dataset and frozen analysis workflow?
- Why this candidate is useful
- Use this candidate to decide whether the current computational signal is reproducible enough to justify experimental escalation.
- Decision available now
- Experimental escalation remains provisional until the signal is independently reproduced.
- What blocks the decision
- Independent computational replication is absent
- Minimum next action
- Repeat the DDX19B analysis in an independent dataset with frozen mapping, contrasts, baselines, null controls and uncertainty reporting.
How would the result change the decision?
SupportiveIf direction and uncertainty reproduce under the frozen plan, retain the candidate and move to its next context-specific gate.
NullIf the signal does not reproduce despite adequate coverage and QC, retain the negative result and do not escalate from the original analysis alone.
Inconclusive / QC failureIf coverage, mapping or source design prevents a fair comparison, mark the replication non-evaluable rather than positive or negative.
Computational replication
DNM1
GAP-07Does the DNM1 evidence pattern survive an independent dataset and frozen analysis workflow?
- Why this candidate is useful
- Use this candidate to decide whether the current computational signal is reproducible enough to justify experimental escalation.
- Decision available now
- Experimental escalation remains provisional until the signal is independently reproduced.
- What blocks the decision
- Independent computational replication is absent
- Minimum next action
- Repeat the DNM1 analysis in an independent dataset with frozen mapping, contrasts, baselines, null controls and uncertainty reporting.
How would the result change the decision?
SupportiveIf direction and uncertainty reproduce under the frozen plan, retain the candidate and move to its next context-specific gate.
NullIf the signal does not reproduce despite adequate coverage and QC, retain the negative result and do not escalate from the original analysis alone.
Inconclusive / QC failureIf coverage, mapping or source design prevents a fair comparison, mark the replication non-evaluable rather than positive or negative.
Computational replication
EHMT2
GAP-07Does the EHMT2 evidence pattern survive an independent dataset and frozen analysis workflow?
- Why this candidate is useful
- Use this candidate to decide whether the current computational signal is reproducible enough to justify experimental escalation.
- Decision available now
- Experimental escalation remains provisional until the signal is independently reproduced.
- What blocks the decision
- Independent computational replication is absent
- Minimum next action
- Repeat the EHMT2 analysis in an independent dataset with frozen mapping, contrasts, baselines, null controls and uncertainty reporting.
How would the result change the decision?
SupportiveIf direction and uncertainty reproduce under the frozen plan, retain the candidate and move to its next context-specific gate.
NullIf the signal does not reproduce despite adequate coverage and QC, retain the negative result and do not escalate from the original analysis alone.
Inconclusive / QC failureIf coverage, mapping or source design prevents a fair comparison, mark the replication non-evaluable rather than positive or negative.
Computational replication
LMO2
GAP-07Does the LMO2 evidence pattern survive an independent dataset and frozen analysis workflow?
- Why this candidate is useful
- Use this candidate to decide whether the current computational signal is reproducible enough to justify experimental escalation.
- Decision available now
- Experimental escalation remains provisional until the signal is independently reproduced.
- What blocks the decision
- Independent computational replication is absent
- Minimum next action
- Repeat the LMO2 analysis in an independent dataset with frozen mapping, contrasts, baselines, null controls and uncertainty reporting.
How would the result change the decision?
SupportiveIf direction and uncertainty reproduce under the frozen plan, retain the candidate and move to its next context-specific gate.
NullIf the signal does not reproduce despite adequate coverage and QC, retain the negative result and do not escalate from the original analysis alone.
Inconclusive / QC failureIf coverage, mapping or source design prevents a fair comparison, mark the replication non-evaluable rather than positive or negative.
Computational replication
MPHOSPH6
GAP-07Does the MPHOSPH6 evidence pattern survive an independent dataset and frozen analysis workflow?
- Why this candidate is useful
- Use this candidate to decide whether the current computational signal is reproducible enough to justify experimental escalation.
- Decision available now
- Experimental escalation remains provisional until the signal is independently reproduced.
- What blocks the decision
- Independent computational replication is absent
- Minimum next action
- Repeat the MPHOSPH6 analysis in an independent dataset with frozen mapping, contrasts, baselines, null controls and uncertainty reporting.
How would the result change the decision?
SupportiveIf direction and uncertainty reproduce under the frozen plan, retain the candidate and move to its next context-specific gate.
NullIf the signal does not reproduce despite adequate coverage and QC, retain the negative result and do not escalate from the original analysis alone.
Inconclusive / QC failureIf coverage, mapping or source design prevents a fair comparison, mark the replication non-evaluable rather than positive or negative.
1 candidate
Pause until external evidence is audited
Provisional hold · external report not yet audited
GFOD2
GAP-09Can the reported external GFOD2 evidence be reproduced from traceable quantitative data and provenance?
- Why this candidate is useful
- Use this candidate to decide whether an external report is strong enough to change the governed evidence state.
- Decision available now
- Keep the current status unchanged until the external result is imported and audited.
- What blocks the decision
- External quantitative evidence has not been imported and audited
- Minimum next action
- Import the GFOD2 effect estimates, denominators, analysis rules and provenance, then reproduce the reported conclusion.
How would the result change the decision?
SupportiveIf the quantitative result and provenance reproduce, reassign the candidate through the normal evidence-transition rules.
NullIf the report cannot be reproduced or supported by its source data, keep the hold and retain the failed audit as evidence.
Inconclusive / QC failureIf essential files or metadata are unavailable, keep the report unassessed and do not translate absence into a negative biological result.
Claim boundary: a Candidate Decision Card authorizes a question and next study, never therapeutic efficacy, safety, DMD benefit or superiority over another gene.
Choose by next action
Start from the uncertainty you can resolve.
These five groups say what should happen before a gene advances. Group size is workload, not importance; genes inside a group are alphabetical, not ranked.
2 genes
Ready for a muscle-cell test
These genes have enough current evidence to justify a controlled human muscle-cell experiment.
Next: Test in matched DMD and control muscle cells.
4 genes
Confirm the gene is detectable first
Very low or uncertain muscle expression could make a perturbation result hard to interpret.
Next: Measure target RNA or protein before escalating.
7 genes
Review safety or biological context
A dependency, toxicity or context concern should be resolved before a larger experiment.
Next: Run the stated safety or context check first.
7 genes
Reproduce the analysis independently
The current result should be repeated in a separate frozen workflow before experimental escalation.
Next: Repeat the analysis and compare the direction and uncertainty.
1 gene
Pause until external evidence is reviewed
A reported external result has not yet passed the platform’s evidence review.
Next: Import and audit the evidence before changing the plan.Reading rule: “ready for a muscle-cell test” means ready for a controlled experiment—not proven to benefit DMD.
How to use one record
Evidence → interpretation → gap → experiment
- 1Open a geneCheck identifiers and source-linked observations.
- 2Read by contextSeparate DMD, muscle, healthy-myoblast and benchmark evidence.
- 3Find the blocking gapDo not treat “not assessed” as “no effect”.
- 4Open the Study CardFreeze the minimum experiment that can change the conclusion.
Advanced: how candidates enter the prospective pilot panel
Prospective selection flow
How 21 hypotheses become a balanced 24-slot panel.
- 1Freeze the eligible universe21 governed candidate recordsHistorical rank is excluded from the current selection rule.
- 2Assign candidate strata14 provisional candidate slots6 muscle-context, 4 high-uncertainty and 4 mechanism-diversity slots.
- 3Add calibration controls10 provisional control slotsKnown, negative/random and toxicity controls remain distinct strata.
- 4Checksum-freeze the panelSignoff pendingImmutable replacement rules must be approved before registration.
- 5Register the truth-generation studyAwaiting experimental authorityRequires one endpoint, effect thresholds, approvals and an immutable Study object. Prediction is a separate validation-track requirement.
Selection boundary: occupying a provisional slot means “chosen for a balanced pilot design,” not “ranked as a therapeutic target.”
Expert view: inspect all evidence dimensions and filters
Plan experiments from evidence gaps
Filter the 21 governed hypotheses
All 21 have L2-bounded target-level HepG2 evidence. This research route does not load historical ranks or scores; use the separate historical audit route when provenance fields are required.
Candidate visual dashboard
One screen for action, agreement and missing evidence
Click a bar or tile to filter the table. The tile order follows current action and gene name; it is not clustered and not ranked.
21L2 hypotheses
9L3a assessed
16Muscle TPM ≥ 1
7Dependency caution
Conflicted
Partial agreement
Full agreement
Muscle expression
L3a assessed
Dependency caution
Gene evidence card
Click any tile to inspect one candidate
The card converts the selected row into a compact evidence route, current gap and minimum next experiment. It is descriptive and reuses frozen fields only.
Gene evidence card
CPEB1
Verify expression · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 0.731 TPM · confirm expression
- DMD source direction Conflicted · 2/3 sources
- Muscle screen Not assessed
- Dependency context No moderate-dependency flag
- Source agreement
- 2/3 sources
- Legacy integrated DMD prior
- up
- Variant stability
- same sign across variants
- Observed cells
- 1
GAP-06
Expression feasibility is not verified in the intended model Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
DNAAF3
Verify expression · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 0.076 TPM · confirm expression
- DMD source direction Direction-consistent up · 1/1 sources
- Muscle screen Not assessed
- Dependency context No moderate-dependency flag
- Source agreement
- 1/1 sources
- Legacy integrated DMD prior
- up
- Variant stability
- sign varied across variants
- Observed cells
- 115
GAP-06
Expression feasibility is not verified in the intended model Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
EPS8L1
Verify expression · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 0.236 TPM · confirm expression
- DMD source direction Direction-consistent up · 2/2 sources
- Muscle screen Not assessed
- Dependency context No moderate-dependency flag
- Source agreement
- 2/2 sources
- Legacy integrated DMD prior
- up
- Variant stability
- sign varied across variants
- Observed cells
- 269
GAP-06
Expression feasibility is not verified in the intended model Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
ZFP69B
Verify expression · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 0.155 TPM · confirm expression
- DMD source direction Conflicted · 2/3 sources
- Muscle screen Not assessed
- Dependency context No moderate-dependency flag
- Source agreement
- 2/3 sources
- Legacy integrated DMD prior
- up
- Variant stability
- sign varied across variants
- Observed cells
- 81
GAP-06
Expression feasibility is not verified in the intended model Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
INTS13
Risk/context review · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 10.991 TPM
- DMD source direction Direction-consistent up · 4/4 sources
- Muscle screen Not assessed
- Dependency context Moderate-dependency caution
- Source agreement
- 4/4 sources
- Legacy integrated DMD prior
- up
- Variant stability
- sign varied across variants
- Observed cells
- 157
GAP-08
Safety and dependency context blocks escalation Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
NAGLU
Risk/context review · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 8.664 TPM
- DMD source direction Conflicted · 3/4 sources
- Muscle screen Assessed · no current hit
- Dependency context Moderate-dependency caution
- Source agreement
- 3/4 sources
- Legacy integrated DMD prior
- up
- Variant stability
- sign varied across variants
- Observed cells
- 88
GAP-08
Safety and dependency context blocks escalation Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
RAC3
Risk/context review · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 0.339 TPM · confirm expression
- DMD source direction Conflicted · 3/4 sources
- Muscle screen Not assessed
- Dependency context Moderate-dependency caution
- Source agreement
- 3/4 sources
- Legacy integrated DMD prior
- up
- Variant stability
- same sign across variants
- Observed cells
- 92
GAP-08
Safety and dependency context blocks escalation Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
RNASEH2C
Risk/context review · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 9.526 TPM
- DMD source direction Direction-consistent up · 4/4 sources
- Muscle screen Assessed · no current hit
- Dependency context Moderate-dependency caution
- Source agreement
- 4/4 sources
- Legacy integrated DMD prior
- up
- Variant stability
- sign varied across variants
- Observed cells
- 132
GAP-08
Safety and dependency context blocks escalation Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
RNF8
Risk/context review · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 4.105 TPM
- DMD source direction Conflicted · 2/4 sources
- Muscle screen Not assessed
- Dependency context Moderate-dependency caution
- Source agreement
- 2/4 sources
- Legacy integrated DMD prior
- up
- Variant stability
- sign varied across variants
- Observed cells
- 122
GAP-08
Safety and dependency context blocks escalation Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
WDR4
Risk/context review · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 4.159 TPM
- DMD source direction Direction-consistent up · 4/4 sources
- Muscle screen Assessed · no current hit
- Dependency context Moderate-dependency caution
- Source agreement
- 4/4 sources
- Legacy integrated DMD prior
- up
- Variant stability
- sign varied across variants
- Observed cells
- 108
GAP-08
Safety and dependency context blocks escalation Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
ZNF236
Risk/context review · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 2.007 TPM
- DMD source direction Conflicted · 2/4 sources
- Muscle screen Not assessed
- Dependency context Moderate-dependency caution
- Source agreement
- 2/4 sources
- Legacy integrated DMD prior
- up
- Variant stability
- sign varied across variants
- Observed cells
- 202
GAP-08
Safety and dependency context blocks escalation Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
MON1A
Muscle-context assay · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 8.057 TPM
- DMD source direction Conflicted · 2/4 sources
- Muscle screen Assessed · no current hit
- Dependency context No moderate-dependency flag
- Source agreement
- 2/4 sources
- Legacy integrated DMD prior
- up
- Variant stability
- sign varied across variants
- Observed cells
- 94
GAP-02
Independent muscle-context perturbation is absent Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-03 DMD-relevant functional validation; GAP-07 independent replication.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
ZNF133
Muscle-context assay · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 4.558 TPM
- DMD source direction Conflicted · 3/4 sources
- Muscle screen Assessed · no current hit
- Dependency context No moderate-dependency flag
- Source agreement
- 3/4 sources
- Legacy integrated DMD prior
- up
- Variant stability
- sign varied across variants
- Observed cells
- 84
GAP-02
Independent muscle-context perturbation is absent Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-03 DMD-relevant functional validation; GAP-07 independent replication.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
ADAM10
Computational replication · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 3.116 TPM
- DMD source direction Conflicted · 2/4 sources
- Muscle screen Assessed · no current hit
- Dependency context No moderate-dependency flag
- Source agreement
- 2/4 sources
- Legacy integrated DMD prior
- up
- Variant stability
- sign varied across variants
- Observed cells
- 161
GAP-07
Independent computational replication is absent Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
CALR
Computational replication · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 88.631 TPM
- DMD source direction Conflicted · 3/4 sources
- Muscle screen Assessed · no current hit
- Dependency context No moderate-dependency flag
- Source agreement
- 3/4 sources
- Legacy integrated DMD prior
- up
- Variant stability
- sign varied across variants
- Observed cells
- 86
GAP-07
Independent computational replication is absent Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
DDX19B
Computational replication · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 7.229 TPM
- DMD source direction Conflicted · 1/2 sources
- Muscle screen Assessed · no current hit
- Dependency context No moderate-dependency flag
- Source agreement
- 1/2 sources
- Legacy integrated DMD prior
- up
- Variant stability
- sign varied across variants
- Observed cells
- 145
GAP-07
Independent computational replication is absent Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
DNM1
Computational replication · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 1.4 TPM
- DMD source direction Conflicted · 2/4 sources
- Muscle screen Not assessed
- Dependency context No moderate-dependency flag
- Source agreement
- 2/4 sources
- Legacy integrated DMD prior
- up
- Variant stability
- sign varied across variants
- Observed cells
- 116
GAP-07
Independent computational replication is absent Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
EHMT2
Computational replication · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 6.406 TPM
- DMD source direction Conflicted · 3/4 sources
- Muscle screen Not assessed
- Dependency context No moderate-dependency flag
- Source agreement
- 3/4 sources
- Legacy integrated DMD prior
- up
- Variant stability
- sign varied across variants
- Observed cells
- 101
GAP-07
Independent computational replication is absent Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
LMO2
Computational replication · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 15.237 TPM
- DMD source direction Conflicted · 3/4 sources
- Muscle screen Not assessed
- Dependency context No moderate-dependency flag
- Source agreement
- 3/4 sources
- Legacy integrated DMD prior
- up
- Variant stability
- sign varied across variants
- Observed cells
- 96
GAP-07
Independent computational replication is absent Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
MPHOSPH6
Computational replication · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 4.946 TPM
- DMD source direction Direction-consistent up · 4/4 sources
- Muscle screen Assessed · no current hit
- Dependency context No moderate-dependency flag
- Source agreement
- 4/4 sources
- Legacy integrated DMD prior
- up
- Variant stability
- sign varied across variants
- Observed cells
- 83
GAP-07
Independent computational replication is absent Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Gene evidence card
GFOD2
Provisional hold · external report not yet audited · current research action, not a target rank.
Evidence route
- L2 perturbation Observed same-context perturbation
- Expression feasibility 2.947 TPM
- DMD source direction Conflicted · 2/4 sources
- Muscle screen Not assessed
- Dependency context No moderate-dependency flag
- Source agreement
- 2/4 sources
- Legacy integrated DMD prior
- up
- Variant stability
- sign varied across variants
- Observed cells
- 127
GAP-09
External quantitative evidence has not been imported and audited Highest missing layer: L4 dmd functional validation. Secondary gaps: GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication.
No new score or ranking is generated; this card reorganizes frozen current-status and evidence-gap fields for inspection.
Interpretation boundary: this view is a triage dashboard. It does not create a target score, rank, cluster, perturbation outcome label or prediction claim.
| Gene | Current action | Highest supported level | Integrated direction | Source agreement | Source state | Prior-variant stability | Evidence modules | Statistical support | Muscle context | External screen (L3a) | Decision-blocking gap | Highest missing layer | Secondary gaps |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| CPEB1 | Verify expression | L2 · observed same-context perturbation | up | 2/3 | Conflicted | same sign across variants | direction conflicted | direction only no significant context rows | 0.731 TPM · confirm expression | Not assessed | GAP-06 Expression feasibility is not verified in the intended model |
L4 dmd functional validation | GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication |
| DNAAF3 | Verify expression | L2 · observed same-context perturbation | up | 1/1 | Direction-consistent up | sign varied across variants | source limited | direction only no significant context rows | 0.076 TPM · confirm expression | Not assessed | GAP-06 Expression feasibility is not verified in the intended model |
L4 dmd functional validation | GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication |
| EPS8L1 | Verify expression | L2 · observed same-context perturbation | up | 2/2 | Direction-consistent up | sign varied across variants | multisource direction consistent | direction only no significant context rows | 0.236 TPM · confirm expression | Not assessed | GAP-06 Expression feasibility is not verified in the intended model |
L4 dmd functional validation | GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication |
| ZFP69B | Verify expression | L2 · observed same-context perturbation | up | 2/3 | Conflicted | sign varied across variants | direction conflicted | direction only no significant context rows | 0.155 TPM · confirm expression | Not assessed | GAP-06 Expression feasibility is not verified in the intended model |
L4 dmd functional validation | GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication |
| INTS13 | Risk/context review | L2 · observed same-context perturbation | up | 4/4 | Direction-consistent up | sign varied across variants | exploratory multisource core | direction only no significant context rows | 10.991 TPM | Not assessed | GAP-08 Safety and dependency context blocks escalation |
L4 dmd functional validation | GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication |
| NAGLU | Risk/context review | L2 · observed same-context perturbation | up | 3/4 | Conflicted | sign varied across variants | direction conflicted | direction only no significant context rows | 8.664 TPM | Assessed · no current hit | GAP-08 Safety and dependency context blocks escalation |
L4 dmd functional validation | GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication |
| RAC3 | Risk/context review | L2 · observed same-context perturbation | up | 3/4 | Conflicted | same sign across variants | direction conflicted | direction only no significant context rows | 0.339 TPM · confirm expression | Not assessed | GAP-08 Safety and dependency context blocks escalation |
L4 dmd functional validation | GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication |
| RNASEH2C | Risk/context review | L2 · observed same-context perturbation | up | 4/4 | Direction-consistent up | sign varied across variants | exploratory multisource core | direction only no significant context rows | 9.526 TPM | Assessed · no current hit | GAP-08 Safety and dependency context blocks escalation |
L4 dmd functional validation | GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication |
| RNF8 | Risk/context review | L2 · observed same-context perturbation | up | 2/4 | Conflicted | sign varied across variants | direction conflicted | direction only no significant context rows | 4.105 TPM | Not assessed | GAP-08 Safety and dependency context blocks escalation |
L4 dmd functional validation | GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication |
| WDR4 | Risk/context review | L2 · observed same-context perturbation | up | 4/4 | Direction-consistent up | sign varied across variants | exploratory multisource core | direction only no significant context rows | 4.159 TPM | Assessed · no current hit | GAP-08 Safety and dependency context blocks escalation |
L4 dmd functional validation | GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication |
| ZNF236 | Risk/context review | L2 · observed same-context perturbation | up | 2/4 | Conflicted | sign varied across variants | direction conflicted | direction only no significant context rows | 2.007 TPM | Not assessed | GAP-08 Safety and dependency context blocks escalation |
L4 dmd functional validation | GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication |
| MON1A | Muscle-context assay | L2 · observed same-context perturbation | up | 2/4 | Conflicted | sign varied across variants | direction conflicted | direction only no significant context rows | 8.057 TPM | Assessed · no current hit | GAP-02 Independent muscle-context perturbation is absent |
L4 dmd functional validation | GAP-03 DMD-relevant functional validation; GAP-07 independent replication |
| ZNF133 | Muscle-context assay | L2 · observed same-context perturbation | up | 3/4 | Conflicted | sign varied across variants | direction conflicted | direction only no significant context rows | 4.558 TPM | Assessed · no current hit | GAP-02 Independent muscle-context perturbation is absent |
L4 dmd functional validation | GAP-03 DMD-relevant functional validation; GAP-07 independent replication |
| ADAM10 | Computational replication | L2 · observed same-context perturbation | up | 2/4 | Conflicted | sign varied across variants | direction conflicted | direction only no significant context rows | 3.116 TPM | Assessed · no current hit | GAP-07 Independent computational replication is absent |
L4 dmd functional validation | GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation |
| CALR | Computational replication | L2 · observed same-context perturbation | up | 3/4 | Conflicted | sign varied across variants | direction conflicted | direction only no significant context rows | 88.631 TPM | Assessed · no current hit | GAP-07 Independent computational replication is absent |
L4 dmd functional validation | GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation |
| DDX19B | Computational replication | L2 · observed same-context perturbation | up | 1/2 | Conflicted | sign varied across variants | direction conflicted | direction only no significant context rows | 7.229 TPM | Assessed · no current hit | GAP-07 Independent computational replication is absent |
L4 dmd functional validation | GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation |
| DNM1 | Computational replication | L2 · observed same-context perturbation | up | 2/4 | Conflicted | sign varied across variants | direction conflicted | direction only no significant context rows | 1.4 TPM | Not assessed | GAP-07 Independent computational replication is absent |
L4 dmd functional validation | GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation |
| EHMT2 | Computational replication | L2 · observed same-context perturbation | up | 3/4 | Conflicted | sign varied across variants | direction conflicted | direction only no significant context rows | 6.406 TPM | Not assessed | GAP-07 Independent computational replication is absent |
L4 dmd functional validation | GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation |
| LMO2 | Computational replication | L2 · observed same-context perturbation | up | 3/4 | Conflicted | sign varied across variants | direction conflicted | direction only no significant context rows | 15.237 TPM | Not assessed | GAP-07 Independent computational replication is absent |
L4 dmd functional validation | GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation |
| MPHOSPH6 | Computational replication | L2 · observed same-context perturbation | up | 4/4 | Direction-consistent up | sign varied across variants | exploratory multisource core | direction only no significant context rows | 4.946 TPM | Assessed · no current hit | GAP-07 Independent computational replication is absent |
L4 dmd functional validation | GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation |
| GFOD2 | Provisional hold · external report not yet audited | L2 · observed same-context perturbation | up | 2/4 | Conflicted | sign varied across variants | direction conflicted | direction only no significant context rows | 2.947 TPM | Not assessed | GAP-09 External quantitative evidence has not been imported and audited |
L4 dmd functional validation | GAP-02 independent muscle-context perturbation; GAP-03 DMD-relevant functional validation; GAP-07 independent replication |
Gap ontology
GAP-01 no independent perturbation · GAP-02 no muscle-context perturbation · GAP-03 no DMD-derived model · GAP-04 no functional phenotype · GAP-05 no dose/time response · GAP-06 no protein/expression validation · GAP-07 no donor or dataset replication · GAP-08 no safety assessment · GAP-09 external evidence not imported.