Vol. 01 · No. 1 · Beta Edition
Oreumicsbeta
Bulk RNA-seq

featureCounts to PyDESeq2 to GSEA — no R session required.

Load a count matrix, a featureCounts or STAR ReadsPerGene table, or Kallisto / Salmon quantifications; attach the sample sheet with Merge Cell Meta; run sample-level QC, cpm / tmm / deseq2 / vst normalisation, PyDESeq2 differential expression, GSEA or ORA and PROGENy / Hallmark / CollecTRI activity — then volcano, MA, heatmap, sample-distance and library-size plots.

Download for WindowsSee the pipeline

In the Bulk RNA-seq view

Nodes in this view
19

8 pipeline stages

Loaders
4

Count Matrix · featureCounts · Abundance (Kallisto/Salmon) · H5AD

Templates
3

1 need a Resource Center install

Plot types
7

across the visualization nodes in this view

Counted from the app’s v0.25 Node Library view, not typed by hand.

Oreumics — airway_deseq2.bioflowRun
Load featureCounts

airway_featurecounts.txt

samples × genes

Merge Cell Meta

airway_metadata.csv

dex column attached

Bulk QC

library size · IQR

outliers labelled

Bulk Normalize

deseq2 size factors

counts kept in layers

Differential Expr.

PyDESeq2 · trt vs untrt

padj · log2FC

GSEA / ORA

GO_Biological_Process_2023

pathways ranked

Bulk Visualization

volcano · α 0.05

figure ready

Building workflow…

A Bulk RNA-seq pipeline on the Oreumics canvas — drag, connect, run. Illustration; parameters are examples.

What it produces

Result capture

Oreumics — Bulk RNA-seq result
Bulk Visualization, volcano plot type — log2 fold-change against −log10 adjusted p-value from PyDESeq2, with genes passing the FDR and fold-change thresholds highlighted.

Bulk Visualization, volcano plot type — log2 fold-change against −log10 adjusted p-value from PyDESeq2, with genes passing the FDR and fold-change thresholds highlighted.

Supported input formats

What you can drag in.

featureCounts.txt

Subread featureCounts output; comment and header lines are handled automatically.

Load featureCounts
STAR ReadsPerGene.tab

ReadsPerGene.out.tab; strand_column picks unstranded (1), forward (2) or reverse (3) counts.

Load featureCounts
Kallisto abundance.tsv

abundance.tsv, format auto-detected. One file per sample; choose TPM or estimated counts with value_type.

Load Abundance (Kallisto/Salmon)
Salmon quant.sf

quant.sf, format auto-detected. One file per sample, joined afterwards with Merge Data.

Load Abundance (Kallisto/Salmon)
Count matrix.csv

CSV / TSV / TXT count table, genes × samples by default (Transpose is on); an optional metadata file joins on sample name.

Load Count Matrix
AnnData.h5ad

A samples × genes AnnData built earlier, e.g. saved from a previous Oreumics run.

Load H5AD
Sample sheet.csv

Sample × condition table attached with Merge Cell Meta (auto-matched, deterministic) or built with Define Groups.

Merge Cell Meta

PyDESeq2 needs raw integer counts — set value_type = counts for Kallisto / Salmon. TPM is fine for the plots.

4 loader nodes in this viewLoad Count MatrixLoad featureCountsLoad Abundance (Kallisto/Salmon)Load H5AD
A typical run

A canonical Bulk RNA-seq pipeline.

The steps below are the story; the full stage-by-stage node list follows in section 04.

  1. 01

    Load counts

    Pick the loader that matches your quantifier. For Kallisto / Salmon, load each sample and concatenate with Merge Data.

    Nodes:Load featureCountsLoad Abundance (Kallisto/Salmon)Load Count MatrixMerge Data
  2. 02

    Attach the design

    Merge Cell Meta joins the sample sheet onto obs with a live match-rate preview; Define Groups builds condition labels from rules when there is no sheet.

    Nodes:Merge Cell MetaDefine Groups
  3. 03

    Sample QC and gene filtering

    Bulk QC reports library size, detected genes and IQR-based outliers — label first, filter once you have seen the plots. Filter Low Expressed drops genes below CPM 1 in fewer than 2 samples (defaults).

    Nodes:Bulk QCFilter Low Expressed
  4. 04

    Normalize

    CPM, TMM, DESeq2 size factors or VST. Raw counts are always preserved in layers["counts"], so differential expression still sees integers.

    Nodes:Bulk Normalize
  5. 05

    Differential expression

    PyDESeq2 between a reference and a test group from any obs column (t-test fallback), with FDR and absolute log2 fold-change thresholds for the significant set.

    Nodes:Differential Expression
  6. 06

    Pathways and activity

    GSEA / ORA runs gseapy preranked GSEA on all genes or ORA on the significant ones against an Enrichr library — Enrichr is the one online service, enabled in the Resource Center. PROGENy, Hallmark and CollecTRI score pathway and TF activity per sample.

    Nodes:GSEA / ORAPROGENy ActivityHallmark ActivityCollecTRI TF Activity
  7. 07

    Plots and export

    Bulk Visualization draws volcano, MA, DEG heatmap, sample distance and library size; Export CSV writes the DE and GSEA tables, and any figure’s underlying data can be exported from its Visualization Workspace.

    Nodes:Bulk VisualizationExport CSV
The Node Library, Bulk RNA-seq view

Every node, in pipeline order.

The app’s Node Library is modality-first: pick bulk rna-seq and it shows these 19 nodes grouped into the 8 stages below. Shared nodes such as Quality Control or Merge Cell Meta appear in more than one view.

Stage 01

Load Data

4 nodes

Load Count MatrixLoad featureCountsLoad Abundance (Kallisto/Salmon)Load H5AD

Stage 02

Sample Sheet & Groups

2 nodes

Merge Cell MetaDefine Groups

Stage 03

QC & Filtering

2 nodes

Bulk QCFilter Low Expressed

Stage 04

Normalization

1 nodes

Bulk Normalize

Stage 05

Differential Expression

1 nodes

Differential Expression

Stage 06

Pathways & Activity

4 nodes

GSEA / ORAPROGENy ActivityHallmark ActivityCollecTRI TF Activity

Stage 07

Plots

1 nodes

Bulk Visualization

Also available

Utilities & Export

4 nodes · collapsed by default in the app

Subset DataObs InspectorSave H5ADExport CSV
Reference:Input / OutputFunctional ActivityBulk RNA-seq
3 templates shipped with the app

Pre-wired pipelines — open, point at data, run.

Templates open from the Templates dialog with sensible defaults already set. Those marked below include a node whose database, model or manifest must be installed in the Resource Center first — every other node runs, and the gated one stops until it is installed. A run never downloads.

Bulk RNA-seq Differential Expression

7 nodes

Load featureCounts output, run QC, normalization, DE, and volcano plot

Load featureCounts → Bulk QC → Filter Low Expressed → Bulk Normalize → Differential Expression → Bulk Visualization

Bulk RNA-seq with GSEA

9 nodes

Load Airway featureCounts, QC, filter, VST, DE, pathway enrichment (ORA), and heatmap

Load featureCounts → Bulk QC → Filter Low Expressed → Bulk Normalize → Differential Expression → GSEA / ORA → Bulk Visualization

Needs Resource Center install · GSEA / ORA

STAR ReadsPerGene — Single-sample Load Demo

3 nodes

Load featureCounts → Bulk QC → Bulk Visualization

All templates, with node chains How the Resource Center works

Sample data the templates point at

Paths are relative to the app’s sampledata/ folder.

  • sampledata/bulk/airway_featurecounts.txtAirway smooth-muscle featureCounts table (dexamethasone treated vs untreated).Used by: Bulk RNA-seq Differential Expression · Bulk RNA-seq with GSEA
  • sampledata/bulk/airway_metadata.csvSample sheet for the airway dataset, for Merge Cell Meta.Used by: Bulk RNA-seq Differential Expression · Bulk RNA-seq with GSEA
  • sampledata/bulk/star_ReadsPerGene.out.tabA single-sample STAR ReadsPerGene output for the strand-column demo.Used by: STAR ReadsPerGene — Single-sample Load Demo
All sample datasets, with sources

What sets it apart

  • DESeq2 without R

    Differential Expression runs PyDESeq2, a Python implementation of the DESeq2 method, from a parameter panel: pick the condition column, reference and test group, and read padj and log2FC from the node.

  • Every quantifier, one canvas

    featureCounts, STAR ReadsPerGene, Kallisto, Salmon and plain matrices come in through 4 loaders that share the same downstream nodes, so switching quantifier does not mean switching pipeline.

  • Activity, not only enrichment

    Alongside GSEA / ORA, the PROGENy (14 pathways), Hallmark (50 programs) and CollecTRI (TF regulons) nodes give per-sample activity scores through decoupler, with priors installed and checksummed in the Resource Center.

  • Local by design

    Counts never leave your machine. Every run records its parameters and a Methods paragraph is generated for the report.

Figures from this pipeline

3 captures

MA plot — fold-change against mean expression across all genes.
Fig. 1MA plot — fold-change against mean expression across all genes.
Sample distance — clustered heatmap of sample similarity.
Fig. 2Sample distance — clustered heatmap of sample similarity.
Library size — per-sample depth, the first QC glance.
Fig. 3Library size — per-sample depth, the first QC glance.
Tutorials and reference

Walk it step by step.

Tutorial

Bulk RNA-seq analysis

featureCounts to DE to volcano on the airway sample, then GSEA.

Tutorial

Pathways and cell–cell interaction

How Enrichr libraries and ORA / GSEA fit together across modalities.

Node reference

End matter

Start with Bulk RNA-seq today.

Free during the beta with an admin-issued Beta Access ticket. Windows 10/11, 64-bit. Runs locally; sign in at each launch.

Create an account Download app