SPATA2 Spatial Transcriptomics Analysis Skill

Comprehensive assistance with SPATA2 spatial transcriptomics analysis, generated from official documentation covering 414 core files including API references, tutorials, and spatial analysis methods.

When to Use This Skill

This skill should be triggered when:

Core Spatial Analysis Tasks:

• Analyzing spatial transcriptomics data from Visium, MERFISH, or other platforms
• Identifying tissue outlines and spatial boundaries
• Computing spatial gradients and expression patterns
• Performing spatial annotation screening (SAS)
• Working with SPATA2 objects and SpatialData classes

Data Processing & Manipulation:

• Loading and preprocessing spatial transcriptomics data
• Adding molecular variables and image data to SPATA2 objects
• Platform compatibility checks (Visium, VisiumHD, MERFISH)
• Spatial segmentation and tissue outline identification

Visualization & Interpretation:

• Creating surface plots and spatial visualizations
• Setting up SI unit axes for spatial plots
• Visualizing spatial annotations and gradients
• Interpreting spatial expression patterns

Advanced Analysis:

• Spatial trajectory analysis
• Differential expression in spatial context
• Gene set enrichment analysis for spatial data
• Copy number variation analysis in spatial data

Quick Reference

Core SPATA2 Workflow

1. Load and Initialize SPATA2 Object

# Load example data
library(SPATA2)
data("example_data")
object <- loadExampleObject("UKF275T")

# Or create from individual data
object <- spata2_initiation_seurat(seurat_object)

2. Identify Tissue Outline

# Method 1: Based on observations (recommended for most platforms)
object <- identifyTissueOutline(
  object,
  method = "obs",
  eps = getCCD(object, unit = "px") * 1.25,
  minPts = 3
)

# Method 2: Based on image (requires pixel content identification)
object <- identifyPixelContent(object, img_name = "image1")
object <- identifyTissueOutline(object, method = "image", img_name = "image1")

3. Create Spatial Annotations

# Create numeric annotations (e.g., hypoxia areas)
object <- createNumericAnnotations(
  object = object,
  variable = "HM_HYPOXIA",
  threshold = "kmeans_high",
  id = "hypoxia_ann",
  inner_borders = FALSE,
  force1 = TRUE
)

4. Spatial Annotation Screening (SAS)

# Pre-filter genes with SPARKX (recommended)
object <- runSPARKX(object)
sparkx_genes <- getSparkxGenes(object, threshold_pval = 0.05)

# Run SAS analysis
sas_out <- spatialAnnotationScreening(
  object = object,
  ids = "hypoxia_ann",
  variables = sparkx_genes,
  core = FALSE,
  distance = "dte"  # distance to edge
)

5. Visualization

# Basic surface plot
plotSurface(object, color_by = "METRN")

# Add spatial annotation outline
plotSurface(object, color_by = "bayes_space") +
  ggpLayerSpatAnnOutline(object, ids = "hypoxia_ann")

# Add SI unit axes
plotSurface(object, color_by = "METRN") +
  ggpLayerThemeCoords() +
  ggpLayerAxesSI(object, unit = "mm")

Essential Utility Functions

Get Center-to-Center Distance

# Get CCD in different units
ccd_px <- getCCD(object, unit = "px")
ccd_mm <- getCCD(object, unit = "mm")

Compute Capture Area

# Automatically calculates capture area for the spatial method
object <- computeCaptureArea(object)

Spatial Annotation Barplot

# Plot changes in grouping proportion vs distance to annotation
plotSasBarplot(
  object,
  grouping = "bayes_space",
  id = "hypoxia_ann",
  distance = distToEdge(object, "hypoxia_ann"),
  unit = getDefaultUnit(object)
)

Platform-Specific Examples

Visium Data Processing

# Load Visium data
object <- loadExampleObject("UKF275T", process = TRUE, meta = TRUE)

# Visium-specific tissue outline parameters
object <- identifyTissueOutline(
  object,
  method = "obs",
  eps = getCCD(object, unit = "px") * 1.25,
  minPts = 3
)

MERFISH Data Processing

# For non-grid platforms (MERFISH, SlideSeq, etc.)
object <- identifyTissueOutline(
  object,
  method = "obs",
  eps = average_min_distance * 10,
  minPts = 25
)

Reference Files

This skill includes comprehensive documentation organized in references/:

api.md (385 pages)

Complete API reference covering:

• Spatial Methods: createSpatialMethod(), computeCaptureArea(), getCCD()
• Visualization: plotSurface(), plotSasBarplot(), ggpLayerAxesSI()
• Spatial Analysis: identifyTissueOutline(), spatialAnnotationScreening()
• Data Manipulation: Functions for adding data, platform compatibility
• Core Classes: SPATA2, SpatialData, SpatialMethod object methods

tutorials.md (27 pages)

Step-by-step tutorials including:

• Spatial Annotation Screening: Complete workflow for SAS analysis
• Data Processing: Platform-specific preprocessing guides
• Visualization Techniques: Advanced plotting and customization
• Analysis Workflows: End-to-end analysis examples

spatial_analysis.md (1 page)

Overview of SPATA2 capabilities:

• Introduction: What SPATA2 offers for spatial analysis
• Platform Support: Visium, MERFISH, and other technologies
• Key Features: Spatial gradients, annotations, trajectories

other.md (2 pages)

Additional information:

• Installation and setup
• Authors and citation information
• License details

Working with This Skill

For Beginners

1. Start with tutorials.md - Read the spatial annotation screening tutorial for a complete workflow example
2. Understand the basics - Learn about SPATA2 objects, tissue identification, and basic visualization
3. Follow the Core Workflow - Use the quick reference steps as your foundation
4. Practice with example data - Use loadExampleObject() to explore SPATA2 functionality

For Intermediate Users

1. Explore api.md - Dive into specific functions for your analysis needs
2. Platform-specific optimization - Check platform compatibility and adjust parameters accordingly
3. Advanced visualization - Master SI units, custom color palettes, and multi-layer plots
4. Spatial annotation screening - Implement SAS for your specific research questions

For Advanced Users

1. Custom spatial methods - Use createSpatialMethod() for new platforms
2. Advanced analysis - Combine multiple techniques like spatial trajectories with SAS
3. Performance optimization - Fine-tune parameters for large datasets
4. Integration with other tools - Combine SPATA2 with Seurat, SingleCellExperiment, etc.

Navigation Tips

• Search by function name - Use the API reference to find specific functions
• Check platform compatibility - Always verify your platform is supported
• Follow the tutorials - The spatial annotation screening tutorial covers most use cases
• Use examples - Copy-paste examples from the documentation as starting points

Key Concepts

Core Objects

SPATA2 Object: The main data structure containing spatial transcriptomics data, coordinates, images, and analysis results.

SpatialData: Contains spatial coordinates, image data, and method-specific information.

SpatialMethod: Defines the spatial technology platform (Visium, MERFISH, etc.) and its parameters.

Spatial Analysis Concepts

Tissue Outline Identification: Using DBSCAN clustering to identify tissue boundaries and sections. Critical for proper spatial analysis.

Spatial Annotation Screening (SAS): Hypothesis-driven screening for genes showing non-random spatial patterns relative to reference features.

Center-to-Center Distance (CCD): The physical distance between adjacent spots/barcodes, crucial for distance-based analyses.

Distance Measures: SPATA2 supports multiple distance units (pixels, mm, μm) with automatic conversion capabilities.

Platform-Specific Considerations

Visium: Fixed grid layout with known CCD. Use eps = CCD * 1.25 and minPts = 3.

MERFISH/SlideSeq: Variable spot locations. Use average minimum distance * 10 for eps and minPts = 25.

Image-based Analysis: Requires identifyPixelContent() before using image-based tissue outline identification.

Resources

references/

Complete documentation extracted from official SPATA2 sources:

• Detailed function documentation with parameters and examples
• Step-by-step tutorials for common analyses
• Platform-specific guidance and best practices
• Code examples with proper R syntax highlighting

scripts/

Add your custom SPATA2 analysis scripts and automation workflows here.

assets/

Store reference images, example datasets, and visualization templates.

Notes

• This skill covers SPATA2 version 3.1.0+ with all modern features
• Documentation includes 414 core files with comprehensive coverage
• All code examples are extracted from official documentation
• Platform-specific guidance is included for major spatial technologies
• Spatial annotation screening (SAS) is a key feature for hypothesis-driven analysis

Updating

To refresh this skill with updated SPATA2 documentation:

1. Re-run the scraper with updated configuration
2. The skill will be rebuilt with the latest API changes and features
3. Check version-specific notes for compatibility updates