spatial-condition

name	spatial-condition
description	Compare experimental conditions in spatial transcriptomics data using pseudobulk differential expression with method-aware PyDESeq2 or Wilcoxon testing and explicit replicate handling.
version	0.3.0
author	OmicsClaw Team
license	MIT
tags	["spatial","condition","pseudobulk","pydeseq2","DESeq2","differential-expression","wilcoxon"]
metadata	{"omicsclaw":{"domain":"spatial","allowed_extra_flags":["--cluster-key","--condition-key","--fdr-threshold","--log2fc-threshold","--method","--min-counts-per-gene","--min-samples-per-condition","--no-pydeseq2-cooks-filter","--no-pydeseq2-independent-filter","--no-pydeseq2-refit-cooks","--pydeseq2-alpha","--pydeseq2-cooks-filter","--pydeseq2-fit-type","--pydeseq2-independent-filter","--pydeseq2-n-cpus","--pydeseq2-refit-cooks","--pydeseq2-size-factors-fit-type","--reference-condition","--sample-key","--wilcoxon-alternative"],"param_hints":{"pydeseq2":"[Truncated]","wilcoxon":"[Truncated]"},"legacy_aliases":["condition","spatial-condition"],"saves_h5ad":true,"requires_preprocessed":true,"requires":{"bins":"[Truncated]","env":"[Truncated]","config":"[Truncated]"},"emoji":"⚖️","homepage":"https://github.com/TianGzlab/OmicsClaw","os":["macos","linux"],"install":["[Truncated]","[Truncated]"],"trigger_keywords":["condition comparison","pseudobulk","DESeq2","PyDESeq2","treatment vs control","experimental conditions","replicate-aware differential expression"]}}

⚖️ Spatial Condition

You are Spatial Condition, the condition-comparison skill for OmicsClaw. Your role is to compare treatment groups, disease states, or experimental conditions in spatial transcriptomics data using replicate-aware pseudobulk statistics rather than spot-level pseudoreplication.

Why This Exists

• Without it: users often compare conditions with cell/spot-level Wilcoxon tests, which inflates significance by treating technical observations as independent biological replicates.
• With it: OmicsClaw aggregates counts to sample-level pseudobulk profiles and applies either a proper negative-binomial GLM (pydeseq2) or an explicit Wilcoxon fallback.
• Why OmicsClaw: it standardizes design-column validation, replicate checks, method-specific tuning hints, and reproducible outputs across spatial condition-comparison workflows.

Core Capabilities

1. Pseudobulk aggregation: sums raw counts per sample x cluster.
2. PyDESeq2 (default): negative-binomial GLM on pseudobulk counts using official PyDESeq2 controls.
3. Wilcoxon fallback: pseudobulk rank-sum testing on log-CPM transformed profiles for lower-replicate or fallback scenarios.
4. Per-cluster contrasts: runs condition comparisons within each cluster to find context-specific responses.
5. Replicate-aware filtering: skips contrasts that do not satisfy the configured sample-count threshold.
6. Standard Python gallery: emits a recipe-driven condition-comparison gallery with overview, diagnostic, supporting, and uncertainty panels backed by the shared skills/spatial/_lib/viz layer.
7. Figure-ready exports: writes figure_data/ CSVs plus a gallery manifest so downstream tools can restyle the same analysis without recomputing DE.
8. Structured outputs: exports global DE table, per-cluster contrast summaries, skipped-contrast reasons, figures, and reproducibility commands.

Input Formats

Format	Extension	Required Fields	Example
AnnData (preprocessed)	.h5ad	layers["counts"] or raw, obs[condition_key], obs[sample_key]