stereo-seq-spatial-grn-regulon

name: stereo-seq-spatial-grn-regulon description: Use when Stereo-seq or STOmics spatial transcriptomics data needs gene regulatory network, transcription-factor, regulon, pySCENIC/SCENIC, SpaGRN, AUCell, regulon specificity score, or spatial TF activity analysis and paper-quality regulon plots.

Stereo-seq Spatial GRN Regulon

Use This For

• Inferring spatial gene regulatory networks from Stereo-seq expression and coordinates.
• Running or adapting SpaGRN, pySCENIC/SCENIC, AUCell, regulon specificity score, or TF activity workflows.
• Plotting regulon activity heatmaps, top regulons by cluster/domain, or regulon spatial maps.

Default Requirements

• Use bundled article-derived scripts in scripts/ before writing new code or searching external repositories.
• Read source_code.md first for curated GRN/regulon templates; if no curated entry fits, search code_candidates.tsv for additional article-linked repositories and reusable files.
• Inspect local Python/R environments first. Prefer conda run -n stereo-skills-py python ... for pySCENIC/AUCell-style Python scripts and conda run -n stereo-skills-r Rscript ... for R scripts. SpaGRN requires its own env (envs/environment-python-spagrn.yml) because it pins Python 3.8 and older numpy/pandas/scanpy. If a required package or database is missing, stop that step and tell the user exactly what is missing and which GRN step is blocked.
• For SpaGRN/pySCENIC, require explicit motif ranking database, motif annotation table, and TF list; do not invent those resources.
• Keep figures paper-ready: Arial, readable labels, non-overlapping legends, equal-aspect spatial plots, vector PDF when possible.
• In the final response, state the reused paper, DOI, code repository/source file, and dataset-specific edits.

Workflow

1. Confirm input object and keys: .h5ad, obsm["spatial"], cluster/domain label, raw-count layer, and species.
2. Read source_code.md, then adapt the closest bundled script:
   • scripts/spagrn_spatial_regulon_template.py for SpaGRN spatial GRN inference and regulon activity plots.
   • scripts/pyscenic_regulon_activity_plot_template.py for pySCENIC/AUCell activity matrix heatmaps and spatial regulon plots.
   • scripts/pyscenic_full_pipeline_template.py for pySCENIC GRN, motif pruning, and AUCell pipeline execution.
3. Let the current question, available resources, and closest paper-code evidence guide the template choice; do not hard-code a tissue-specific method route.
4. Check external resources before running: TF list, .feather motif rankings, motif annotation .tbl, and optional ligand-receptor/niche table.
5. Validate outputs: number of regulons, TF-target sizes, regulon specificity scores, spatial coherence, and known marker/TF plausibility.
6. Treat GRN edges as hypotheses unless supported by motif evidence, spatial specificity, and external biology.

Output Expectations

• Method, package versions if available, and required databases.
• Regulon table and activity matrix locations.
• Top regulons by cluster/domain and spatial maps.
• Caveats for species, motif database, sparse bins/cellbins, and small domains.
• Reused article code source, paper DOI, repository URL, original file name, and dataset-specific edits.