Sample name	Group	Replicate
NPCTRL1	NPCTRL → Control	1
NPCTRL2	NPCTRL → Control	2
NPCTRL3	NPCTRL → Control	3

Sample name	Group	Time	Exposure	Replicate
NP124H50UG1	NP1 → Nanomaterial A	24H	50UG	1
NP124H50UG2	NP1 → Nanomaterial A	24H	50UG	2
NP124H50UG3	NP1 → Nanomaterial A	24H	50UG	3
NP172H50UG1	NP1 → Nanomaterial A	72H	50UG	1
NP172H50UG2	NP1 → Nanomaterial A	72H	50UG	2

Research record Public release Controlled scientific workflow

Document ID: GB-MARSS-001
Record class: Scientific workflow
Public status: Feature record
Revision: 2026-09
Scope: Dataset inspection → row identity → expression-matrix canonicalization

Subject / GeneBean MARSS

GeneBean MARSS — governed scientific data workflow

GeneBean MARSS helps researchers inspect supported tabular datasets, establish reviewable feature identity, and prepare the intended expression matrix before later controlled analysis. The public workflow keeps those preparation steps visible so the researcher can review what the dataset represents before progressing.

Public workflow record GB-MARSS / PUB-01

01 Inspect data
02 Resolve identity
03 Prepare matrix

Public GeneBean workflow coverage ends with Expression Matrix Canonicalization. Later MARSS stages remain private and controlled.

Open GeneBean Read the RNA-seq workflow

01 Public feature record

Three reviewable preparation features

The public GeneBean record describes what the researcher can inspect and prepare. Internal methods, implementation details, private documentation, and unpublished research remain outside this page.

GB-MARSS / 01

File Reader

Public preparation feature

Opens supported tabular expression data and establishes the dataset context before identity or matrix preparation begins.

Purpose: Bring a new upload or an available public dataset into a visible, reviewable starting state.
Researcher review: File type, dataset name and shape, a paginated data preview, and provenance/licence information when an available catalogue dataset provides it.
Public result: A structurally inspected dataset ready to move into biological row identity review.

GB-MARSS / 02

Row Namer

Public identity feature

Establishes which column identifies each biological feature and makes that choice explicit before expression measurements are treated as a numeric matrix.

Purpose: Detect candidate identity columns, let the researcher choose the biological identifier, and preserve a stable row reference.
Researcher review: Candidate identity fields, duplicate or missing identifier handling, and non-numeric metadata columns that should not remain inside the expression-measurement matrix.
Public result: A row-identified numeric expression matrix that can be saved for the next public preparation stage or downloaded for independent review.

GB-MARSS / 03

Expression Matrix Canonicalization

Final public MARSS stage

Turns the row-identified dataset into a researcher-reviewed study matrix while keeping sample selection, naming, identifier state, and the final matrix boundary visible.

Purpose: Review dataset structure and row identity, resolve identifiers when required, and define the sample columns that belong to the intended analysis.
Researcher review: Input structure, canonical identifier result, included or excluded sample columns, sample order, group naming, study annotations, sample-level preview, and the final matrix summary before confirmation.
Public result: A confirmed study matrix that represents the final public GeneBean/MARSS preparation state. A cleaned table can also be downloaded for independent review.

Evidence specimen / synthetic

Illustrative expression-matrix record

SPEC-EMX-01

Illustrative synthetic values only — not user data. The specimen shows the kind of row identity, sample columns, and reviewable matrix boundary that the public GeneBean workflow prepares.

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

TRACE / 01 Dataset inspected Shape + preview visible

TRACE / 02 Identity resolved Feature axis explicit

TRACE / 03 Matrix prepared Public endpoint: EMC

Public illustrative specimen for documentation only. Values are synthetic; no private dataset, internal algorithm, or unpublished method is represented.

02 Scientific controls

Preparation remains inspectable before progression

Control	Public feature description
Dataset inspection	The researcher can establish what kind of tabular dataset has been provided before moving into later preparation.
Feature identity	Row identity is made explicit before the dataset is treated as an expression matrix.
Matrix boundary	The intended sample columns and resulting study matrix remain visible for review rather than being presented as an unexplained result.
Researcher review	Public preparation features are presented as reviewable scientific work, with later controlled analysis kept outside the public workflow.

03 Public workflow boundary

Public through Expression Matrix Canonicalization

Public endpoint Expression Matrix Canonicalization

Public MARSS access ends here File Reader → Row Namer → Expression Matrix Canonicalization

The confirmed EMC study matrix is the final public GeneBean/MARSS preparation state. Normalization & Quality Control and later MARSS stages continue as private, controlled functionality and are not public workflow entry points.

Controlled continuation: Normalization & Quality Control
Later MARSS stages: Private / controlled
Internal details: Withheld from public documentation

¹ Zomniverse is a modular knowledge and computational universe designed, developed, and maintained by Wilder Ruiz. The system organizes documentation, tools, learning systems, and executable analytical components into structured domains known as knowledge cities.

² Each city within Zomniverse functions as a self-contained thematic environment focused on domains such as bioinformatics, open science, and AI-assisted research.

³ Zomniverse is designed as a modular architecture enabling independent evolution of cities, modules, and knowledge systems.

⁴ Built with PHP, JavaScript, Node.js, Plumber, and R.

⁵ Enables integration of learning, reasoning, and execution into one system.

⁶ Outputs require independent verification by the user.

⁷ Internal architecture maintained in private research repository.

Module

Provenance

File Metadata

Structure

Frozen gene-symbol authority human_gene_symbols.rds

Gene-identity inspection result

Dataset state after gene-identity inspection

Provenance

File Metadata

Structure

Production resolution policy

Identifier resolution result

Canonical dataset state

Provenance

File Metadata

Structure

Study matrix configuration

Sample-name interpretation

Detected sample naming grammar

Final matrix summary

Configuration summary

EMC complete — ready for Normalization & QC

Select and validate input

Governed EMC Sample Design

Configure filtering and normalization

Low-expression filtering

Library normalization ?

QC explorer

Review and commit

Candidate run ?

Run annotation ?

Reference-signature scoring preparation

Validate input

Select reference signature

Evaluate signature coverage

Component coverage ?

Configure scoring

Governed method settings

Researcher summary option ?

Run governed scoring

Execution review

GB-Tox Explorer

Review and governed reports

Immutable source contract

Approved figures

Current report contents

Component evidence

Exact sample results

File Reader

Row Namer

Expression Matrix Canonicalization

Illustrative expression-matrix record

Frozen gene-symbol authority `human_gene_symbols.rds`

Library normalization

Candidate run

Run annotation

Component coverage

Researcher summary option