scSignaline

scSignaline is a Windows desktop application for downstream analysis of single-cell RNA-seq data stored as Seurat RDS objects. The English edition provides a guided PySide6 interface and a bundled R analysis environment for CellChat, cell-cycle, pseudotime, reporting, and reproducible export workflows.

Main workflows

• Inspect a Seurat RDS object, metadata, and available UMAP/t-SNE embeddings.
• Infer cell-cell communication with CellChat.
• Compare signaling pathways and ligand-receptor communication between groups.
• Score S/G2M cell-cycle state with Seurat.
• Infer transcriptional-state trajectories with Monocle 3 pseudotime.
• Review figures and tables and export processed/full RDS objects or a complete analysis package.
• Assemble English Word/PDF reports and optionally use an evidence-bounded AI report assistant.

Download and run on Windows

1. Open the repository's Releases page.
2. Download the Windows x64 ZIP archive for the latest release.
3. Extract the entire archive to a normal writable directory. Do not run the executable from inside the ZIP file.
4. Double-click scSignaline.exe.
5. On first launch, activation is required. Copy the displayed device code and obtain a scSignaline activation code from the software provider.
6. The first successful launch deploys the bundled R environment. This normally takes 1–3 minutes and may take longer on slower disks or antivirus-scanned systems.

Keep the complete extracted folder together. scSignaline.exe depends on the adjacent _internal directory and bundled runtime files.

Supported target: 64-bit Windows 10 or Windows 11.

Project input requirements

The input should be a Seurat RDS object produced by scFlow or another compatible workflow. For the complete guided workflow, metadata should include:

• group: experimental group
• sample: biological replicate/sample
• an annotation column such as cell.type, celltype, subtype, sub.type, or sub_type
• a two-dimensional umap or tsne reduction
• an RNA assay with a populated data slot/layer

Analysis output is written to a user-selected project directory. Every run has timestamped parameters, summaries, logs, figures, tables, and R objects for traceability.

Scientific interpretation notes

• CellChat results are computational communication inferences and require experimental validation.
• Cells from the same sample are not independent biological replicates.
• Pseudotime is an inferred ordering of transcriptional states, not elapsed chronological time.
• AI-assisted text must be reviewed by a qualified researcher before scientific, clinical, or publication use.

Run from source

The source tree can use an installed Windows R environment. A release build also requires a tested portable R tree at vendor/R; vendor/ is intentionally not stored in Git.

git clone https://github.com/oligodendrocyte11/scSignaline.git
cd scSignaline
py -3.11 -m venv .venv
.\.venv\Scripts\Activate.ps1
python -m pip install --upgrade pip
python -m pip install -r requirements-windows.txt
python launcher.py

Run source and repository checks:

$env:QT_QPA_PLATFORM = "offscreen"
python -m unittest discover -s tests -v
python scripts/check_windows_source.py
python scripts/check_release_safety.py --strict-english

Build a Windows package

Place the tested portable R distribution in vendor/R, then run:

.\build_windows.ps1

The onedir package is created at:

dist_windows/scSignaline-English-1.0.0/

Do not commit vendor/, dist_windows/, build logs, RDS files, activation files, credentials, or generated results. Release archives belong on the GitHub Releases page, not in Git history.

Privacy and AI services

The standard analysis workflow is local. When a user explicitly enables a cloud AI provider, scSignaline sends only the bounded, de-identified evidence described in the confirmation dialog. It does not send full RDS objects, raw expression matrices, local file paths, logs, or credentials. API keys can remain in process memory or be stored in the operating-system credential store.

Licensing and third-party software

Copyright in the scSignaline application source is reserved; see LICENSE. The application uses third-party components under their own licenses, including Python, PySide6/Qt, R, Seurat, CellChat, and Monocle 3. See THIRD_PARTY_NOTICES.md and the license files included with the packaged runtime.

The offline activation generator and private signing key are not part of this repository or public releases. The desktop application contains only the public verification key.

Security

Do not post activation codes, device codes, API keys, private datasets, or patient information in public issues. See SECURITY.md.