Zomniverse
GENEBEAN
Select a scientific workflow to begin.
ZOMNIVERSE / PUBLIC STATUS
Scientific capability, implementation, access and limitations, without private infrastructure details.
Loading public systems registry…
MARSS MODULE
Include the controls and experimental samples relevant to the comparison you intend to analyse. Remove unrelated experiments before loading the matrix into MARSS.
| gene_id | CTRL_1 | CTRL_2 | CTRL_3 | TREAT_1 | TREAT_2 | TREAT_3 |
|---|---|---|---|---|---|---|
| GENE_A | 124 | 131 | 127 | 242 | 251 | 246 |
| GENE_B | 87 | 92 | 89 | 174 | 168 | 179 |
| GENE_C | 305 | 297 | 311 | 492 | 510 | 503 |
| GENE_D | 61 | 58 | 64 | 121 | 117 | 126 |
Example only — replicate number and group structure should follow the actual experimental design.
Select and freeze the gene identity column.
Govern gene identity and study design before creating the immutable input for Normalization & QC.
Verify the current Row Namer browser-state handoff and dataset authority before preprocessing.
Validate row identifiers against the frozen production symbol authority.
human_gene_symbols.rds
Validate current row identifiers against the frozen offline
human_gene_symbols.rds reference. This provides a
deterministic, reproducible assessment of symbol validity,
duplication, identifier mode, and whether canonicalization is
required. Inspection is read-only and does not modify the active dataset.
Inspection does not replace the dataset. This snapshot shows the current dataset together with the identifier assessment established by the frozen reference.
Review and apply production cleaning/conversion, duplicate, and unresolved-identifier policy.
Apply the existing deterministic EMC normalization/conversion route when inspection identifies row names requiring canonicalization. Changed identifiers, removed rows, duplicates, canonical artifacts, and the resulting dataset structure remain reviewable in this stage.
This snapshot reflects the canonical output produced by identifier resolution. Compare it with the original input before proceeding to Study Matrix configuration.
Select, order, annotate, and preview the samples included in the governed study matrix.
Select the samples to include, identify the control group, and optionally standardize sample names. Your original dataset will remain unchanged.
Define which dataset columns enter the MARSS analysis before configuring the study matrix. Columns that do not match below are excluded, not modified.
Provide optional group-renaming evidence by uploading a text/code file or entering mapping information below. ZAR AI will use this information during the next Suggest analysis to associate detected sample groups with candidate standardized names. Uploaded code is treated strictly as text and is never executed. Suggested renaming remains provisional until applied through the existing study-matrix controls.
Complete group renaming to unlock time and exposure annotation.
Review deterministic naming candidates and approve the study-design interpretation.
Compare local deterministic analysis with optional, validated ZAR suggestions. Neither path writes learning memory.
00:00.000Group/material prefix → time segment → exposure number → replicate suffix
| Original | Display | Group | Time | Exposure | Replicate | Status |
|---|
Verify the production contract, then explicitly create and register the immutable EMC study matrix.
Awaiting confirmed Study Matrix configuration.
The governed emc_study_matrix artifact was materialized and registered successfully.
Transform a validated EMC raw-count artifact into governed, analysis-ready outputs.
not resolved ·
checksum verified by the artifact registry
logCPM is already normalized; NQC requires the governed EMC raw_counts artifact.—Sample membership, order, approved names and scientific dimensions were reviewed in EMC and cannot be changed here.
| Original | Approved name | Group | Time | Exposure | Replicate |
|---|
Retain genes with sufficient expression for the confirmed group structure.
Correct compositional differences while retaining filtered raw counts.
Sample separation may overlap with an acquisition or processing batch. Review the design metadata before downstream modeling.
One analysis group contains fewer biological replicates than the recommended review threshold. Confirm whether the group should be retained for downstream modeling.
Commit this candidate to make its normalized log2 CPM artifact available to GB-Tox from the MARSS dock.
Inputs are recorded at commit, but the Differential Expression module is not yet implemented.
Governed toxicogenomic reference-signature scoring from a committed NQC normalized logCPM artifact.
Confirm the NQC handoff and prepare the gene identity contract for governed validation.
Complete and commit Normalization & QC to make normalized_log_cpm.csv available.
Choose the governed toxicogenomic reference signature used for this analysis.
Stage 1 verification is required.
Confirm how the selected governed reference signature is represented in the verified normalized expression matrix.
Stage 1 verification and a server-verified signature are required.
Coverage is preserved for each gene set used by GSVA/ssGSEA.
| Component | Matched | Missing | Coverage |
|---|
Review the governed ssGSEA contract and select only the supported rule-based summary mode.
A current deterministic coverage evaluation is required.
In fixed mode, the researcher-selected cutoff is applied after scoring to produce the rule-based High/Low summary. Changing it changes the classification contract and configuration fingerprint, not the underlying ssGSEA calculation; it is not a toxicity probability, clinical-risk threshold, or independent biological validation.
Review the frozen input, reference signature, coverage and scoring configuration before starting the real ssGSEA execution.
A current server-verified Stage 4 configuration is required.
No score has been calculated yet. Starting the run executes the server-bound contract shown above exactly once.
— · Complete · immutable · — · —
Inspect governed score, component, coverage, fingerprint, and artifact views.
Open Stage 6 to verify and load the completed result.
Sample ranking, median, fingerprint drivers, and pinned differences are descriptive views of the immutable governed result. They are not significance tests, treatment effects, toxicity conclusions, or mechanism inferences.
Raw values remain the authoritative secondary data object. Row z-scores are display coordinates only.
Exact sample scores, ranks, labels, and governed Sample Design context.
Inspect the governed signature genes and their exact case-sensitive representation in this run.
Open this tab to load verified signature membership.
No receipt loaded.
Verify immutable evidence, choose optional report sections, and generate a canonical static report.
Report generation never changes the completed result. Every report remains bound to its source artifact and exact checksums.
Check an approved snapshot to include and preview it; uncheck it to remove the preview. Approved figures remain immutable.
Open Stage 7 to load approved figures.
Every visible figure is rendered from its exact approved specification.
Generated reports are private, owner-scoped artifacts registered on the server for this governed result. Signing out or closing the browser does not remove them.
No reports generated for this run.
Zomniverse · Governance
How Zomniverse handles identity, research data, analysis workflows, storage, and connected services.
Loading Privacy Policy…
ZAR AI