Gene Evidence Atlas

Find a gene. See the evidence. Identify the missing test.

Search every versioned gene record by approved symbol or identifier. Evidence depth and prediction readiness remain separate: this atlas supports research triage, not therapeutic ranking.

Core records: 17,921
Detailed records: 123
Governed actions: 21
Typed buckets: 26

Released evidence landscape

See where all 17,921 records sit.

Each cell crosses evidence depth with prediction readiness. Select a populated cell to inspect its records; empty cells are shown so absence stays visible.

Descriptive inventory of the frozen gene core. Counts are not target scores, therapeutic ranks, or evidence of efficacy.
Evidence depth ↓	R0evidence annotation only	R1non-DMD perturbation substrate	R2muscle-context bridge evidence
G4muscle-context screen, not DMD replication	2641.47% of coreInspect records →	Not presentAbsent in current releaseEvidence gap, not a score	90.05% of coreInspect records →
G3non-DMD perturbation evidence	Not presentAbsent in current releaseEvidence gap, not a score	120.07% of coreInspect records →	Not presentAbsent in current releaseEvidence gap, not a score
G1DMD-background evidence	17,63598.4% of coreInspect records →	Not presentAbsent in current releaseEvidence gap, not a score	Not presentAbsent in current releaseEvidence gap, not a score
G0stable gene identity only	10.01% of coreInspect records →	Not presentAbsent in current releaseEvidence gap, not a score	Not presentAbsent in current releaseEvidence gap, not a score

How to read it: evidence depth describes what has been observed; readiness describes what the record can support computationally. A higher code is not a better therapeutic target.

Governed subset: all 21 current action records have a detailed audit, but none has a qualifying DMD candidate perturbation outcome.

Governed research set

Start with the decision the evidence can support.

These 21 records are grouped by their current next action. Counts are workflow states—not evidence scores or a target ranking.

2 genesMuscle-context assay

Test whether a source-context response transfers to a controlled human-myogenic system.

Open this governed action →4 genesVerify expression

Confirm RNA or protein detectability before committing perturbation resources.

Open this governed action →7 genesRisk and context review

Resolve dependency, toxicity, or context-specific feasibility before escalation.

Open this governed action →7 genesComputational replication

Reproduce the signal in an independent source with a frozen analysis workflow.

Open this governed action →1 geneHold for external audit

Keep the evidence state unchanged until the external report can be reproduced.

Open this governed action →

Refine released evidence

Filter gene records

Matching order reflects identifier relevance only. It is not a target score, confidence rank, or therapeutic priority.

Filter options

Gene answer

Search one gene to open a research answer

Use an HGNC symbol to see what is known, whether a DMD virtual-cell response is computable and which experiment would change the state.

Known17,921 candidate records need one selected gene

ComputabilitySearch one gene to open computability

Missing evidenceThe decisive gap appears after a gene is selected

Next experimentTry ZNF133, MON1A, ADAM10 or DMD.

Search all resources Ask in DMD Design experiment Machine record

Versioned records

Browse the atlas

Showing 3,721–3,744 of 17,921 matching records.

Mini answer cards for visible gene records

Mini answer cardDDX19B

G4R2

Known: muscle-context screen, not DMD replication · DMD context: Sources disagree

Computability: muscle-context bridge evidence

Missing: Independent computational replication is absent
Next: Computational replication

Evidence Ask in Virtual Cell Planner Study Card

Mini answer cardDDX20

G1R0

Known: DMD-background evidence · DMD context: Sources disagree

Computability: evidence annotation only